The best batsman across years and formats!
Nirvanam
He leads the number 2 by around 10% points!
'nuff said!
jzfredricks
discussed at length not too long ago
Howzzat
> http://blogs.cricinfo.com/itfigures/archives/2010/01/_sachin_tendulka...
>
> He leads the number 2 by around 10% points!
>
> 'nuff said!
Glad that it makes you feel good. Me too actually since I am fellow
Sachin admirer.
However, it doesn't take a genius to figure out right away that the
author started with the hypothesis that Tendulkar was the best, and
then come up with the appropriate criteria to make that happen. Just
one example is enough - Viv Richards has a better average and strike
rate than Tendulkar in ODIs, yet is given 10% less points. The
author's bias is clear when he overweights the aggregate runs -
Tendulkar was always going to win this one.
I am not suggesting that Tendulkar shouldn't be near the top. He is
probably #1 or #2 in ODIs, but among the names considered, would
barely make my top 10 in tests. Loads of batsmen have played more
significant innings in tests than Tendulkar. And in a separate
analysis on Cricinfo yesterday, Tendulkar was shown at #23 (yes -
twenty third) in batting averages of the last 10 years (against top 8
sides) - 'nuff said!
Nirvanam
> On Jan 5, 9:01 am, Nirvanam <viz.sha...@gmail.com> wrote:
>
> >http://blogs.cricinfo.com/itfigures/archives/2010/01/_sachin_tendulka...
>
> > He leads the number 2 by around 10% points!
>
> > 'nuff said!
>
> Glad that it makes you feel good. Me too actually since I am fellow
> Sachin admirer.
>
> However, it doesn't take a genius to figure out right away that the
> author started with the hypothesis that Tendulkar was the best, and
> then come up with the appropriate criteria to make that happen. Just
> one example is enough - Viv Richards has a better average and strike
> rate than Tendulkar in ODIs, yet is given 10% less points. The
> author's bias is clear when he overweights the aggregate runs -
> Tendulkar was always going to win this one.
I think you got it wrong there. For one, if you are a regular reader
of It Figures blog you'll know how and why this particular article
came up. The author has not started with a hypothesis that Sachin is
the best. Secondly, you may not know this.....this is the same guy who
did not pick even one of Sachin's centuries in the Wisden 100 best
centuries. Yes, he is the same guy who ranked those centuries and he
is a well known Dravid-liker than a Sachin praiser.
> I am not suggesting that Tendulkar shouldn't be near the top. He is
> probably #1 or #2 in ODIs, but among the names considered, would
> barely make my top 10 in tests. Loads of batsmen have played more
> significant innings in tests than Tendulkar. And in a separate
> analysis on Cricinfo yesterday, Tendulkar was shown at #23 (yes -
> twenty third) in batting averages of the last 10 years (against top 8
> sides) - 'nuff said!
You are referring to a list of averages in a particular period. Same
thing for the previous decade shows Sachin as the clear topper with
all except one of those other 22 not even above 50. Come on dude,
understand the difference between a career analysis and a temporal
analysis. I mean seriously which one of those batsmen in the top 10
list can you say is not correctly placed...leave alone the topper?
Please differentiate between a Sachin fanatic claiming things that
Sachin has not done and a person who is sharing his glee in an
objective analysis of all batsmen who played tests and odi's. If you
are going to view each one of my posts on Sachin as some kinda a
'clueless' sachin-fanatic's rant then you are neither giving the point
I am making due consideration nor are you allowing a fair open-minded
assessment of my viewpoints
alvey
Yawn.
It doesn't matter how many criteria and weighting factors the Sachinots
devise, he'll always be in the 2nd estate of batsmen.
alvey
Howzzat
Look - I never called you a clueless Sachin-fanatic. You are as
entitled to your opinion as I am to mine. And anyway, this was not a
response to your post per se - I was actually pointing out flaws in
that analyst's methodology. I would have responded similarly if
someone else had posted that analysis. So no reason to take this
personally.
Regarding the points you raised:
1. I am aware of the Wisden 100 list (and the subsequent debate it
generated), but didn't know that it was compiled by the same person
who did this analysis. Regardless, my point about the flaw in his
methodology still stands. He has overweighted the "aggregate" runs -
category in which Sachin is the clear leader. Do you seriously buy
that? You could tweak some parameters easily to make any of the other
greats come out on top. The only (mild) criticism I have of you is
your glee in treating this analysis as gospel, just because the
outcome suits your pov.
2. Your point about career vs temporal analysis is laughable. The
"temporal" analysis is the last TEN years - the entire decade -,
during which Tendulkar played 89 of his 162 tests (55%). This is not
selective, unlike the "Ponting vs Tendulkar" post yesterday showing
averages from the last 2 years.
We can debate about the lists and rankings forever. You obviously
consider SRT to be the best test batsman of the lot, so the analysis
appears objective to you. Fair enough. But to me, SRT had done too
much damage in the last 8-10 years to be considered worthy of that
crown. ODIs are a different story though - there isn't too much to
argue about his status as #1 or #2 (behind IVAR).
Howzzat
> On Jan 5, 9:52 pm, Howzzat <shgupt...@yahoo.com> wrote:
>
>
>
> > On Jan 5, 9:01 am, Nirvanam <viz.sha...@gmail.com> wrote:
>
> > >http://blogs.cricinfo.com/itfigures/archives/2010/01/_sachin_tendulka...
>
> > > He leads the number 2 by around 10% points!
>
> > > 'nuff said!
>
> > Glad that it makes you feel good. Me too actually since I am fellow
> > Sachin admirer.
>
> > However, it doesn't take a genius to figure out right away that the
> > author started with the hypothesis that Tendulkar was the best, and
> > then come up with the appropriate criteria to make that happen. Just
> > one example is enough - Viv Richards has a better average and strike
> > rate than Tendulkar in ODIs, yet is given 10% less points. The
> > author's bias is clear when he overweights the aggregate runs -
> > Tendulkar was always going to win this one.
>
> I think you got it wrong there. For one, if you are a regular reader
> of It Figures blog you'll know how and why this particular article
> came up. The author has not started with a hypothesis that Sachin is
> the best. Secondly, you may not know this.....this is the same guy who
> did not pick even one of Sachin's centuries in the Wisden 100 best
> centuries. Yes, he is the same guy who ranked those centuries and he
> is a well known Dravid-liker than a Sachin praiser.
One more thing. If you are going to take these "objective" analyses at
face value, then I presume you also agree with Sachin's non-inclusion
in the the top 100 centuries of all time? Would you then seriously
consider Sachin as the greatest test bat of the modern era - when
several of his contemporaries appear in that list multiple times? I
mean - does the ability to play a special knock that significantly
influences a match not figure in your criteria of determining great
batsman at all?
FWIW, I am not suggesting that Sachin has played no such knock. IMO
his 136 Chennai knock against Pak was a great one, better that his
103* against England, (despite the unfavorable result), and worthy of
inclusion in any "list". But for me he hasn't produced enough of
these, thus my reluctance in placing him as high as you do. But as I
said earlier, to each his own.
Nirvanam
> entitled to your opinion as I am to mine. And anyway, this was not a
> response to your post per se - I was actually pointing out flaws in
> that analyst's methodology. I would have responded similarly if
> someone else had posted that analysis. So no reason to take this
> personally.
Nah...I didn't take it personally dude.
> Regarding the points you raised:
> 1. I am aware of the Wisden 100 list (and the subsequent debate it
> generated), but didn't know that it was compiled by the same person
> who did this analysis. Regardless, my point about the flaw in his
> methodology still stands. He has overweighted the "aggregate" runs -
> category in which Sachin is the clear leader. Do you seriously buy
> that? You could tweak some parameters easily to make any of the other
> greats come out on top. The only (mild) criticism I have of you is
> your glee in treating this analysis as gospel, just because the
> outcome suits your pov.
Nope I haven't treated this as gospel at all.
Also, just on the methodology you might want to consider that over-all
aggregate bit a lil more...if you read the article Anat has actually
normalized the aggregate values. You are only looking at the table
displayed...read the whole article and you'll see he has normalized
even the aggregate.
> 2. Your point about career vs temporal analysis is laughable. The
> "temporal" analysis is the last TEN years - the entire decade -,
> during which Tendulkar played 89 of his 162 tests (55%). This is not
> selective, unlike the "Ponting vs Tendulkar" post yesterday showing
> averages from the last 2 years.
It is temporal. There was a period between 2004 and 2007 when Ponting
scored heavily and Sachin was seriously injured (every other batsman
of any note averaged more than Sachin during those 3 yrs). Now, if you
remove those 3 yrs you'll see Sachin averages more than Ponting in all
other years of the decade. If not more in each and every year, at
least not significantly behind Ponting (not even 5 runs behind Is'd
think).
> We can debate about the lists and rankings forever. You obviously
> consider SRT to be the best test batsman of the lot, so the analysis
> appears objective to you. Fair enough. But to me, SRT had done too
> much damage in the last 8-10 years to be considered worthy of that
> crown. ODIs are a different story though - there isn't too much to
> argue about his status as #1 or #2 (behind IVAR).
Well, you have your opinion that is fair enough. But if I may, would
you like to share the reasons of why you think Sachin is not the best
batsman in Test cricket ever (according to me, no one is better than
him in the history of the game considering subjective and objective
criteria...he along with Don and Lara are in a class apart). And who
would your best batsman be if we consider only the last 40 yrs? How
does that batsman do better than Sachin....if you'd like to share your
reasons pls do so...else ignore it. If you consider only the last 10
yrs, definitely Ponting has better numbers than Sachin...that I
maintain (to be specific Ponting, and not only him but 20 others would
have been better than Sachin in the mid noughties)
Nirvanam
Howzatt, some numbers for you:
Year Sachin Ponting
1999 68.00 63.07
2000 63.88 63.60
2001 62.68 38.60
2002 55.68 70.93
2003 17.00 100.20
2004 91.50 41.00
2005 44.40 67.13
2006 24.27 88.86
2007 55.42 38.40
2008 48.31 47.28
2009 67.62 38.77
What do these numbers tell you? These numbers pretty clearly tell me
that during the mid noughties Ponting capitalized heavily while Sachin
was injured, but since then Ponting has become rather mortal (not even
averaging 50 in a batting heaven period for 3 yrs continuously)
whereas Sachin has reinvented himself. The early noughties both of
them are good.
Nirvanam
> face value, then I presume you also agree with Sachin's non-inclusion
> in the the top 100 centuries of all time? Would you then seriously
> consider Sachin as the greatest test bat of the modern era - when
> several of his contemporaries appear in that list multiple times? I
> mean - does the ability to play a special knock that significantly
> influences a match not figure in your criteria of determining great
> batsman at all?
>
> FWIW, I am not suggesting that Sachin has played no such knock. IMO
> his 136 Chennai knock against Pak was a great one, better that his
> 103* against England, (despite the unfavorable result), and worthy of
> inclusion in any "list". But for me he hasn't produced enough of
> these, thus my reluctance in placing him as high as you do. But as I
> said earlier, to each his own.
On this same thing I argued with Anant on how when he could include
Kapil's knock of 129 v/s SA in 93 in a losing cause in the top 100,
how then could he not include Sachin's 136. This despite him / Wisden
stating that had India won that game Sachin's 136 would've made it to
number 2 or number 3 or some such high number. And he basically did
not respond to the specific case.
And I find it rather amusing that none of Sachin's masterpieces are in
that list (I mean not even 1 from a total of 20+ by then)...for ex the
155 at Mumbai when he demolished Warne in the 3rd innings on a dust
bowl to setup a victory, the 119 at Perth, the century at Manchester
to save the game, etc, etc
alvey
> But if I may, would
> you like to share the reasons of why you think Sachin is not the best
> batsman in Test cricket ever (according to me, no one is better than
> him in the history of the game considering subjective and objective
> criteria...
I'd be absolutely rivetted to know what "...subjective and objective
criteria" exist that can overcome a 45 run difference in batting average.
Number of syllables in the name?
Scrabble values applied to name?
Losing a run off your average for every year elapsed?
Tell! tell!
Incredulous minds want to know!
alvey
alvey
> On Jan 6, 12:53�am, Howzzat <shgupt...@yahoo.com> wrote:
> Howzatt, some numbers for you:
>
> Year Sachin Ponting
> 1999 68.00 63.07
> 2000 63.88 63.60
> 2001 62.68 38.60
> 2002 55.68 70.93
> 2003 17.00 100.20
> 2004 91.50 41.00
> 2005 44.40 67.13
> 2006 24.27 88.86
> 2007 55.42 38.40
> 2008 48.31 47.28
> 2009 67.62 38.77
>
> What do these numbers tell you? These numbers pretty clearly tell me
> that during the mid noughties Ponting capitalized heavily while Sachin
> was injured,
The nexus escapes me.
> but since then Ponting has become rather mortal (not even
> averaging 50 in a batting heaven period for 3 yrs continuously)
> whereas Sachin has reinvented himself. The early noughties both of
> them are good.
Oty Howzzat.
Nirvanam
> On Tue, 5 Jan 2010 11:58:38 -0800 (PST), Nirvanam wrote:
> > But if I may, would
> > you like to share the reasons of why you think Sachin is not the best
> > batsman in Test cricket ever (according to me, no one is better than
> > him in the history of the game considering subjective and objective
> > criteria...
>
> I'd be absolutely rivetted to know what "...subjective and objective
> criteria" exist that can overcome a 45 run difference in batting average.
For one, to do well playing against at least 7 considerably equal or
better sides in the world than just ONE.
Playing on all 3 difficult tyes of surfaces facing some of the best
bowlers the game of cricket has known, rather than just 1 or 2 great
bowlers in just 2 countries of the world
Anyway, the point is that Sachin is not greater than Bradman but
definitely not lesser. In my opinion they are virtuosos along with
Lara. No other batsmen in the history of the Test game can claim that
spot, in my opinion.
alvey
And you are perfectly entitled to your opinion. Just don't get snotty, or
worse, even more verbose, if people point at you and laugh.
alvey
jzfredricks
> you like to share the reasons of why you think Sachin is not the best
> batsman in Test cricket ever
A few simple reasons;
1 and 2) Don Bradman
There are two thing that are universally accepted in the cricket
world;
1) Don Bradman was the greatest batsman ever. NB I've not included the
word "Test" in there.
2) Anyone who doesn't agree with #1 is a first class idiot
jzfredricks
No he's not.
Dangerous things should be taken off kids.
Don speaks the truth
3) Anyone who agrees with the first dosen't have a brain.
Cue: Bradman never played the shorter format.
Don
Don speaks the truth
Pot, Kettle, Black.
Don
> or
> worse, even more verbose, if people point at you and laugh.
>
> alvey- Hide quoted text -
>
> - Show quoted text -
skp
IMO, Nirvanam goes overboard when he says that SRT is the greatest
test bat ever i.e. better than the Don and presumably much better than
Lara and Ponting.
But you are going seriously underwater when you say that he barely
makes the top 10. Who are your top 9, pray tell? Scratch that. Some
of the oldtimers you might put on your list I probably would never
have heard of.
What post 1970 batsmen have been better test bats than SRT? IMO, no
one other than Lara and Ponting can be compared to SRT. If you say
that those 2 guys have been better than SRT...fine...I would disgaree
but would not vehemently argue the point. But I am wondering if you
would place any others above SRT.
Howzzat
> > Look - I never called you a clueless Sachin-fanatic. You are as
> > entitled to your opinion as I am to mine. And anyway, this was not a
> > response to your post per se - I was actually pointing out flaws in
> > that analyst's methodology. I would have responded similarly if
> > someone else had posted that analysis. So no reason to take this
> > personally.
>
> Nah...I didn't take it personally dude.
>
> > Regarding the points you raised:
> > 1. I am aware of the Wisden 100 list (and the subsequent debate it
> > generated), but didn't know that it was compiled by the same person
> > who did this analysis. Regardless, my point about the flaw in his
> > methodology still stands. He has overweighted the "aggregate" runs -
> > category in which Sachin is the clear leader. Do you seriously buy
> > that? You could tweak some parameters easily to make any of the other
> > greats come out on top. The only (mild) criticism I have of you is
> > your glee in treating this analysis as gospel, just because the
> > outcome suits your pov.
>
> Nope I haven't treated this as gospel at all.
> Also, just on the methodology you might want to consider that over-all
> aggregate bit a lil more...if you read the article Anat has actually
> normalized the aggregate values. You are only looking at the table
> displayed...read the whole article and you'll see he has normalized
> even the aggregate.
>
Yes he has normalized, but only partially. If you read the comments,
someone brought up the Richards issue in ODIs (better strike rate and
average). Anant's response was that although SRT score 3x runs as
Richards did, after normalization he "only" gets twice as many points
as Richards. Please tell me that you find this as ridiculous as I do.
> > 2. Your point about career vs temporal analysis is laughable. The
> > "temporal" analysis is the last TEN years - the entire decade -,
> > during which Tendulkar played 89 of his 162 tests (55%). This is not
> > selective, unlike the "Ponting vs Tendulkar" post yesterday showing
> > averages from the last 2 years.
>
> It is temporal. There was a period between 2004 and 2007 when Ponting
> scored heavily and Sachin was seriously injured (every other batsman
> of any note averaged more than Sachin during those 3 yrs). Now, if you
> remove those 3 yrs you'll see Sachin averages more than Ponting in all
> other years of the decade. If not more in each and every year, at
> least not significantly behind Ponting (not even 5 runs behind Is'd
> think).
This I have a serious problem with. You are the one who is making this
"temporal" by selectively filtering out Tendulkar's worst years.
Injuries are part and parcel of every player's career. If Tendulkar's
injuries were so serious, he should have stayed the hell out. There is
no one to blame but him for rushing back and spoiling his figures. And
anyway, if you insist on excluding his worst years, then in the
interest of fairness also remove his best years (1997-2000) and see
where he ends up at. Probably won't be too different from where he is
now. Over a long career, you can't be filtering selectively. If you
do, then you have to do it for everyone else you are comparing him to
as well - how do you know they didn't have their own issues that
prevented them for continuing their upward trajectory?
And this not about Tendulkar vs Ponting. That was a separate thread,
and we already beat it to death :-)
>
> > We can debate about the lists and rankings forever. You obviously
> > consider SRT to be the best test batsman of the lot, so the analysis
> > appears objective to you. Fair enough. But to me, SRT had done too
> > much damage in the last 8-10 years to be considered worthy of that
> > crown. ODIs are a different story though - there isn't too much to
> > argue about his status as #1 or #2 (behind IVAR).
>
> Well, you have your opinion that is fair enough. But if I may, would
> you like to share the reasons of why you think Sachin is not the best
> batsman in Test cricket ever (according to me, no one is better than
> him in the history of the game considering subjective and objective
> criteria...he along with Don and Lara are in a class apart). And who
> would your best batsman be if we consider only the last 40 yrs? How
> does that batsman do better than Sachin....if you'd like to share your
> reasons pls do so...else ignore it. If you consider only the last 10
> yrs, definitely Ponting has better numbers than Sachin...that I
> maintain (to be specific Ponting, and not only him but 20 others would
> have been better than Sachin in the mid noughties)
1. First, if 22 batsmen have averaged more than Tendulkar (see
cricinfo article) over an entire decade (representing 55% of
Tendulkar's tests), there is no way in hell any objective critic can
consider him the best batsman of his time, let alone of all time. I
don't care whether he was injured or whatever, but if over half his
career he isn't among the top 2-3 batsman of his time, then the
argument stops there. If we all close our eyes and pretend that
Tendulkar's career stopped in 2002, I will grant you that there is
merit in considering him as the best of his time - even that would be
highly arguable. See below.
2. My primary reason was not considering Sachin as the best (in Tests)
is the paucity of "landmark" or "defining" inning he has played. When
I think of Sachin the ODI batsman, I have fond memories of his 90 vs
Australian in the 1996 world cup, the Desert Storm, and numerous
others that stand out and have withstood the test of time. However in
test cricket, I struggle - there's the Perth knock, the 136, the 155,
and the 103*, but over a 20 year career I expected a lot more. The
fact that he doesn't appear in the top 100 innings list (even allowing
for Anant's flawed analysis) is too damning. So is the inability to
put together a 500 run series, cross 900 on the ICC ratings, score 2
hundreds in a test even once, lack of a really big knock and finally
the numerous failures in 2nd innings of a test, when the match was
there to be won or lost. The 103* is so celebrated now because it is
so unique and rare - it took Tendulkar 19 years to produce a match
winning knock in the 4th innings. And IIRC, his second innings average
against Australia is in the teens. Maybe I am being too harsh, but for
someone being argued as the best of all time there are too many boxes
left unchecked.
Also there's the style. The reason Tendulkar ruled our hearts in the
early years was his counter-attacking nature, the boy-on-the-burning-
deck persona. Again the 96 world cup vs Australia, desert storm, the
155* vs Warne, the 136 etc come to mind. How many of these do you
recall in the last 8 years? He has become more like Dravid and Kallis
in style, which might be effective but won't leave an imprint on your
soul like some of Tendulkar's earlier knocks did.
3. For some of the reasons mentioned above, FWIW, I consider Lara as
the best of the last 20 years. After that there's not much to choose b/
w Tendulkar, Ponting, Kallis, and Dravid (maybe Sehwag considering how
many games he has turned). Going back another 20 years you'd add
Richards, Gavaskar, Miandad, Greg Chappell and maybe a couple others
to the list - any one of whom can be reasonably argued to be greater
than or equal to Tendulkar.
Howzzat
> > entitled to your opinion as I am to mine. And anyway, this was not a
> > response to your post per se - I was actually pointing out flaws in
> > that analyst's methodology. I would have responded similarly if
> > someone else had posted that analysis. So no reason to take this
> > personally.
>
> Nah...I didn't take it personally dude.
>
> > Regarding the points you raised:
> > 1. I am aware of the Wisden 100 list (and the subsequent debate it
> > generated), but didn't know that it was compiled by the same person
> > who did this analysis. Regardless, my point about the flaw in his
> > methodology still stands. He has overweighted the "aggregate" runs -
> > category in which Sachin is the clear leader. Do you seriously buy
> > that? You could tweak some parameters easily to make any of the other
> > greats come out on top. The only (mild) criticism I have of you is
> > your glee in treating this analysis as gospel, just because the
> > outcome suits your pov.
>
> Nope I haven't treated this as gospel at all.
> Also, just on the methodology you might want to consider that over-all
> aggregate bit a lil more...if you read the article Anat has actually
> normalized the aggregate values. You are only looking at the table
> displayed...read the whole article and you'll see he has normalized
> even the aggregate.
>
Yes he has normalized, but only partially. If you read the comments,
someone brought up the Richards issue in ODIs (better strike rate and
average). Anant's response was that although SRT score 3x runs as
Richards did, after normalization he "only" gets twice as many points
as Richards. Please tell me that you find this as ridiculous as I do.
> > 2. Your point about career vs temporal analysis is laughable. The
> > "temporal" analysis is the last TEN years - the entire decade -,
> > during which Tendulkar played 89 of his 162 tests (55%). This is not
> > selective, unlike the "Ponting vs Tendulkar" post yesterday showing
> > averages from the last 2 years.
>
> It is temporal. There was a period between 2004 and 2007 when Ponting
> scored heavily and Sachin was seriously injured (every other batsman
> of any note averaged more than Sachin during those 3 yrs). Now, if you
> remove those 3 yrs you'll see Sachin averages more than Ponting in all
> other years of the decade. If not more in each and every year, at
> least not significantly behind Ponting (not even 5 runs behind Is'd
> think).
This I have a serious problem with. You are the one who is making this
"temporal" by selectively filtering out Tendulkar's worst years.
Injuries are part and parcel of every player's career. If Tendulkar's
injuries were so serious, he should have stayed the hell out. There is
no one to blame but him for rushing back and spoiling his figures. And
anyway, if you insist on excluding his worst years, then in the
interest of fairness also remove his best years (1997-2000) and see
where he ends up at. Probably won't be too different from where he is
now. Over a long career, you can't be filtering selectively. If you
do, then you have to do it for everyone else you are comparing him to
as well - how do you know they didn't have their own issues that
prevented them for continuing their upward trajectory?
And this not about Tendulkar vs Ponting. That was a separate thread,
and we already beat it to death :-)
>
> > We can debate about the lists and rankings forever. You obviously
> > consider SRT to be the best test batsman of the lot, so the analysis
> > appears objective to you. Fair enough. But to me, SRT had done too
> > much damage in the last 8-10 years to be considered worthy of that
> > crown. ODIs are a different story though - there isn't too much to
> > argue about his status as #1 or #2 (behind IVAR).
>
> Well, you have your opinion that is fair enough. But if I may, would
> you like to share the reasons of why you think Sachin is not the best
> batsman in Test cricket ever (according to me, no one is better than
> him in the history of the game considering subjective and objective
> criteria...he along with Don and Lara are in a class apart). And who
> would your best batsman be if we consider only the last 40 yrs? How
> does that batsman do better than Sachin....if you'd like to share your
> reasons pls do so...else ignore it. If you consider only the last 10
> yrs, definitely Ponting has better numbers than Sachin...that I
> maintain (to be specific Ponting, and not only him but 20 others would
> have been better than Sachin in the mid noughties)
1. First, if 22 batsmen have averaged more than Tendulkar (see
Howzzat
> that those 2 guys have been better than SRT...fine...I would disgaree
> but would not vehemently argue the point. But I am wondering if you
> would place any others above SRT.
Talking SRT's test career in totality (not just the 90s or excluding
his injury-hit phase), post 1970 I would put Lara, Chappell, Richards
and Sobers (he overlapped this period) ahead of him. Arguables would
be Ponting, Kallis, Dravid, Gavaskar (maybe Miandad and Border too)
but let's take two of them. Add about half a dozen pre-70s guys and
you can see why SRT barely makes my top 10. But that's my opinion :-)
There's no shame in being among the top 10 test bats of all time, is
there?
I think the reason some of us are so passionate about SRT and him
being numero uno (leaving aside Bradman obviously) is that he was
clearly headed there, and we can't bring ourselves to accept that
things went pear-shaped post 2002. In our heart and minds we are still
stuck in that fantasy world, refusing to accept the reality because it
hurts so much.
Howzzat
> that those 2 guys have been better than SRT...fine...I would disgaree
> but would not vehemently argue the point. But I am wondering if you
> would place any others above SRT.
Talking SRT's test career in totality (not just the 90s or excluding
Howzzat
> that those 2 guys have been better than SRT...fine...I would disgaree
> but would not vehemently argue the point. But I am wondering if you
> would place any others above SRT.
Talking SRT's test career in totality (not just the 90s or excluding
Howzzat
Apologies for duplicate postings. I have no idea how this happened.
Hopefully this will appear only once :-)
Will S
"Nirvanam" <viz.s...@gmail.com> wrote in message
news:7c788e23-6c2f-49d0...@m3g2000yqf.googlegroups.com...
ffs, his average is below Pontings
nuff said
further more ... protective gear including helmets and covered wickets
more then nuff said
skp
>
> - Show quoted text -
I suppose that I can only lament the fact that I never saw him play
before 2003. Thanks for your POV.
skp
> cricinfo article) over an entire decade (representing 55% of
> Tendulkar's tests), there is no way in hell any objective critic can
> consider him the best batsman of his time, let alone of all time. I
> don't care whether he was injured or whatever, but if over half his
> career he isn't among the top 2-3 batsman of his time, then the
> argument stops there. If we all close our eyes and pretend that
> Tendulkar's career stopped in 2002, I will grant you that there is
> merit in considering him as the best of his time - even that would be
> highly arguable. See below.
>
at the top. What Nirvanam is saying that he has had 3 or 4 bad years
in a 20 year career. You are right in saying "too bad...one can't be
selective". But OTOH, you also can't be too selective in
overemphasizing the second decade average. It's the career
performance that counts which includes the good and bad times and a 20
year career carries more weight than a 10 year career.
> 2. My primary reason was not considering Sachin as the best (in Tests)
> is the paucity of "landmark" or "defining" inning he has played. When
> I think of Sachin the ODI batsman, I have fond memories of his 90 vs
> Australian in the 1996 world cup, the Desert Storm, and numerous
> others that stand out and have withstood the test of time. However in
> test cricket, I struggle - there's the Perth knock, the 136, the 155,
> and the 103*, but over a 20 year career I expected a lot more. The
> fact that he doesn't appear in the top 100 innings list (even allowing
> for Anant's flawed analysis) is too damning.
I don't know much about he Perth knock but the other 3 arguably at
least ought to rank in a top 100 if fairly drawn up today.
So is the inability to
> put together a 500 run series,
A truly meaningles and arbitrary landmark. Besides, he came damn
close with 490 in OZ 2 years ago. Plus, he has had a stretch of *3
innings* (never mind a series) where he scored 495 odd runs without
being dismissed (Sydney '04 and Multan '04).
cross 900 on the ICC ratings, score 2
> hundreds in a test even once, lack of a really big knock
Points taken.
and finally
> the numerous failures in 2nd innings of a test, when the match was
> there to be won or lost. The 103* is so celebrated now because it is
> so unique and rare - it took Tendulkar 19 years to produce a match
> winning knock in the 4th innings.
I could be wrong but there aren't many players in the modern era who
have sored more second inning centuries than SRT. Heck, even the
recent second innings 100 against Sri Lanka ought to be considered a
match saver but is never mentioned as an achivement to speak of.
Plus, if his second innings average against OZ is so poor, his first
innings average must be quite a bit higher than his unfiltered overall
average. First innings scores (which set up wins) are not necessarily
any less valuable than second innings scores.
And IIRC, his second innings average
> against Australia is in the teens. Maybe I am being too harsh, but for
> someone being argued as the best of all time there are too many boxes
> left unchecked.
Point taken. But I suspect that almost anyone else that you night
nominate as the best in the modern era would also have several - but
perhaps different - boxes unchecked if their performances were
scrutinized to the nth degree as is being done with SRT's numbers.
> Also there's the style. The reason Tendulkar ruled our hearts in the
> early years was his counter-attacking nature, the boy-on-the-burning-
> deck persona. Again the 96 world cup vs Australia, desert storm, the
> 155* vs Warne, the 136 etc come to mind. How many of these do you
> recall in the last 8 years? He has become more like Dravid and Kallis
> in style, which might be effective but won't leave an imprint on your
> soul like some of Tendulkar's earlier knocks did.
>
> 3. For some of the reasons mentioned above, FWIW, I consider Lara as
> the best of the last 20 years. After that there's not much to choose b/
> w Tendulkar, Ponting, Kallis, and Dravid (maybe Sehwag considering how
> many games he has turned). Going back another 20 years you'd add
> Richards, Gavaskar, Miandad, Greg Chappell and maybe a couple others
> to the list - any one of whom can be reasonably argued to be greater
> than or equal to Tendulkar.- Hide quoted text -
>
> - Show quoted text -- Hide quoted text -
dechucka
"skp" <vika...@gmail.com> wrote in message
news:b1e27464-16d8-4ea1...@q4g2000yqm.googlegroups.com...
>> 1. First, if 22 batsmen have averaged more than Tendulkar (see
>> cricinfo article) over an entire decade (representing 55% of
>> Tendulkar's tests), there is no way in hell any objective critic can
>> consider him the best batsman of his time, let alone of all time. I
>> don't care whether he was injured or whatever, but if over half his
>> career he isn't among the top 2-3 batsman of his time, then the
>> argument stops there. If we all close our eyes and pretend that
>> Tendulkar's career stopped in 2002, I will grant you that there is
>> merit in considering him as the best of his time - even that would be
>> highly arguable. See below.
>>
> Longevity plays a big factor in many (Bharat for example) who rank him
> at the top. What Nirvanam is saying that he has had 3 or 4 bad years
> in a 20 year career. You are right in saying "too bad...one can't be
> selective". But OTOH, you also can't be too selective in
> overemphasizing the second decade average. It's the career
> performance that counts which includes the good and bad times and a 20
> year career carries more weight than a 10 year career.
OK so he is the Worlds Greatest Batsman except for those times he played
shit.
snip
skp
> On Jan 5, 9:01 am, Nirvanam <viz.sha...@gmail.com> wrote:
>
> >http://blogs.cricinfo.com/itfigures/archives/2010/01/_sachin_tendulka...
>
> > He leads the number 2 by around 10% points!
>
> > 'nuff said!
>
> Glad that it makes you feel good. Me too actually since I am fellow
> Sachin admirer.
>
> However, it doesn't take a genius to figure out right away that the
> author started with the hypothesis that Tendulkar was the best, and
> then come up with the appropriate criteria to make that happen. Just
> one example is enough - Viv Richards has a better average and strike
> rate than Tendulkar in ODIs, yet is given 10% less points. The
> author's bias is clear when he overweights the aggregate runs -
> Tendulkar was always going to win this one.
>
> I am not suggesting that Tendulkar shouldn't be near the top. He is
> probably #1 or #2 in ODIs, but among the names considered, would
> barely make my top 10 in tests. Loads of batsmen have played more
> significant innings in tests than Tendulkar. And in a separate
> analysis on Cricinfo yesterday, Tendulkar was shown at #23 (yes -
> twenty third) in batting averages of the last 10 years (against top 8
> sides) - 'nuff said!
I have only now briefly perused the author's work. I will study it
more carefully later.
Who would be your top 5 in order over the past 40 years on a combined
test and ODI basis?
skp
> "skp" <vikani...@gmail.com> wrote in message
>
> - Show quoted text -
eh? Read the post again and, in particular the last line of the
paragraph you have quoted.
Howzzat
> > 1. First, if 22 batsmen have averaged more than Tendulkar (see
> > cricinfo article) over an entire decade (representing 55% of
> > Tendulkar's tests), there is no way in hell any objective critic can
> > consider him the best batsman of his time, let alone of all time. I
> > don't care whether he was injured or whatever, but if over half his
> > career he isn't among the top 2-3 batsman of his time, then the
> > argument stops there. If we all close our eyes and pretend that
> > Tendulkar's career stopped in 2002, I will grant you that there is
> > merit in considering him as the best of his time - even that would be
> > highly arguable. See below.
>
> Longevity plays a big factor in many (Bharat for example) who rank him
> at the top. What Nirvanam is saying that he has had 3 or 4 bad years
> in a 20 year career. You are right in saying "too bad...one can't be
> selective". But OTOH, you also can't be too selective in
> overemphasizing the second decade average. It's the career
> performance that counts which includes the good and bad times and a 20
> year career carries more weight than a 10 year career.
>
I am not overemphasizing (or underemphasizing) any part at all. In
fact, I am insisting that the career as a whole be considered. If you
considered only the second half of his career, SRT wouldn't even make
a third World XI. All my assessments of other players vs SRT were
based on everyone's careers taken in totality.
> > 2. My primary reason was not considering Sachin as the best (in Tests)
> > is the paucity of "landmark" or "defining" inning he has played. When
> > I think of Sachin the ODI batsman, I have fond memories of his 90 vs
> > Australian in the 1996 world cup, the Desert Storm, and numerous
> > others that stand out and have withstood the test of time. However in
> > test cricket, I struggle - there's the Perth knock, the 136, the 155,
> > and the 103*, but over a 20 year career I expected a lot more. The
> > fact that he doesn't appear in the top 100 innings list (even allowing
> > for Anant's flawed analysis) is too damning.
>
> I don't know much about he Perth knock but the other 3 arguably at
> least ought to rank in a top 100 if fairly drawn up today.
>
> So is the inability to
>
> > put together a 500 run series,
>
> A truly meaningles and arbitrary landmark. Besides, he came damn
> close with 490 in OZ 2 years ago. Plus, he has had a stretch of *3
> innings* (never mind a series) where he scored 495 odd runs without
> being dismissed (Sydney '04 and Multan '04).
The landmark may be arbitrary, but so is a century, an average of 50
etc. My point is that all ATG batsmen have at least 1-2 monster series
in their careers - Tendulkar has yet to manage one.
>
> cross 900 on the ICC ratings, score 2
>
> > hundreds in a test even once, lack of a really big knock
>
> Points taken.
>
> and finally
>
> > the numerous failures in 2nd innings of a test, when the match was
> > there to be won or lost. The 103* is so celebrated now because it is
> > so unique and rare - it took Tendulkar 19 years to produce a match
> > winning knock in the 4th innings.
>
> I could be wrong but there aren't many players in the modern era who
> have sored more second inning centuries than SRT. Heck, even the
> recent second innings 100 against Sri Lanka ought to be considered a
> match saver but is never mentioned as an achivement to speak of.
> Plus, if his second innings average against OZ is so poor, his first
> innings average must be quite a bit higher than his unfiltered overall
> average. First innings scores (which set up wins) are not necessarily
> any less valuable than second innings scores.
>
> And IIRC, his second innings average
>
> > against Australia is in the teens. Maybe I am being too harsh, but for
> > someone being argued as the best of all time there are too many boxes
> > left unchecked.
>
> Point taken. But I suspect that almost anyone else that you night
> nominate as the best in the modern era would also have several - but
> perhaps different - boxes unchecked if their performances were
> scrutinized to the nth degree as is being done with SRT's numbers.
>
True, but none of these other contenders' fans claim them as the
undisputed #1s of all time. I've already said that there's isn't much
to choose b/w the 5-6 top players of the last 20 years. But if someone
wants to claim their candidate as the clear #1, there needs to be a
watertight argument that can stand up to any and all scrutiny. And as
you can see, there isn't one for SRT. I do consider Lara a better test
bat, partially because he checks pretty much all the boxes above
(admittedly my subjective criteria). Sure he has his flaws too, but
then I am not comparing him to Bradman, just Tendulkar :-)
Point is that nobody would argue much if you said that SRT/Ponting/
Lara/* is the best modern day bat. Arguments can be made and
statistics manipulated to present a case either way, but if you then
claimed that x is not only the best of the current era but of all
times you are inviting trouble.
<snip>
Howzzat
You know, that would be very tough to answer. For starters, it would
depend on how much weight you assigned to tests vs odis. The analysis
being debated here assigns equal weight to each, which I don't agree
with. But I don't know what the right ratio should be - it would be
highly subjective.
At the end of the day it would be between SRT, Lara and Richards
depending on your weighting of tests vs odis, closely followed by
Ponting and Kallis. I assume we are not considering Sobers.
dechucka
=======================================
I'm agreeing. This cherry picking of statistics is stupid.
jzfredricks
> > world;
>
> > 1) Don Bradman was the greatest batsman ever. NB I've not included the
> > word "Test" in there.
> > 2) Anyone who doesn't agree with #1 is a first class idiot
>
> 3) Anyone who agrees with the first dosen't have a brain.
> Cue: Bradman never played the shorter format.
Right. So you're an idiot. I guess knowing that should be #3 on the
list.
Nirvanam
> test bat ever i.e. better than the Don and presumably much better than
> Lara and Ponting.
Hang on there! I never said Sachin is greater than Don or Lara. I
always maintained that Sachin, Don, and Lara are a class apart and
none of them is greater or lesser than the other except statistically.
But yes, all of them are pretty much better off than Ponting
> But you are going seriously underwater when you say that he barely
> makes the top 10. Who are your top 9, pray tell? Scratch that. Some
> of the oldtimers you might put on your list I probably would never
> have heard of.
Ponting does make it to the top 10, not just barely but definitely, at
least in my opinion.
> What post 1970 batsmen have been better test bats than SRT? IMO, no
> one other than Lara and Ponting can be compared to SRT. If you say
> that those 2 guys have been better than SRT...fine...I would disgaree
> but would not vehemently argue the point. But I am wondering if you
> would place any others above SRT.
I am saying, at this point in time, I do not see any batsman who can
be placed above Sachin...leave alone post 1970. Having said that, I
also maintain that there are 2 batsmen who are ALONGSIDE Sachin: Don
and Lara. Look if we keep only numbers in the picture, its a no-
comparison...Don wins outright. I am coming from an overall
perspective....and frankly I haven't known of any batsmen who are
better than these 3. Do you?
Nirvanam
To each his own :-)
Nirvanam
> "Nirvanam" <viz.sha...@gmail.com> wrote in message
HAR HAR HAR HAR HAR HAR!!!!!!!
LOL...that's crazy seriously.....the big man here said the same thing
2 months ago when Sachin was at 54.2x and Ponting was 55.8x. Today it
is 54.72 to Sachin and 54.79 to Ponting....lol! Hang on till the
season gets over and you'll see your lone argument of Ponting's
greatness over Sachin, i.e. he having a higher average, will also bite
the dust.
Wanna be reminded on the very exact day it happens? I'll be more than
willing to.
Nirvanam
Very well said!
Nirvanam
> cricinfo article) over an entire decade (representing 55% of
> Tendulkar's tests), there is no way in hell any objective critic can
> consider him the best batsman of his time, let alone of all time. I
> don't care whether he was injured or whatever, but if over half his
> career he isn't among the top 2-3 batsman of his time, then the
> argument stops there. If we all close our eyes and pretend that
> Tendulkar's career stopped in 2002, I will grant you that there is
> merit in considering him as the best of his time - even that would be
> highly arguable. See below.
skp has argued perfectly for this one...am not gonna repeat the same
point here.
I think your not ranking Sachin high has got to do with your
expectations out of Sachin...nothing to do with what Sachin has
delivered really. Here's the thing: person A plays brilliantly in time
T1 and so you expect him to become god in time T2. But he shows his
human nature and performs as effectively, dare I say more effectively,
than the first version in time T2. Around the time T2 you see another
person B who you did not expect to be any ATG kinda fellow suddenly
rising up the ladder and putting together a purple patch. It coincides
with person A's non-attaining God status. Person B remains mortal, was
always mortal, was NEVER considered immortal, NEVER even expected to
get to such a high stage and doesn't even get there. Mind playing
games with Sachin's and Ponting's true value, you reckon?
> 3. For some of the reasons mentioned above, FWIW, I consider Lara as
> the best of the last 20 years. After that there's not much to choose b/
> w Tendulkar, Ponting, Kallis, and Dravid (maybe Sehwag considering how
> many games he has turned). Going back another 20 years you'd add
> Richards, Gavaskar, Miandad, Greg Chappell and maybe a couple others
> to the list - any one of whom can be reasonably argued to be greater
> than or equal to Tendulkar.
Opinion entitlement is free in a free society :-)
Andrew Dunford
"jzfredricks" <jzfre...@gmail.com> wrote in message
news:7486a39b-196f-4e17...@r5g2000yqb.googlegroups.com...
<snip>
> 2) Anyone who doesn't agree with #1 is a first class idiot
Which would still be preferable to a limited overs idiot.
Andrew
John Hall
alvey <al...@play.com> writes:
>On Tue, 5 Jan 2010 11:58:38 -0800 (PST), Nirvanam wrote:
>
>
>> But if I may, would
>> you like to share the reasons of why you think Sachin is not the best
>> batsman in Test cricket ever (according to me, no one is better than
>> him in the history of the game considering subjective and objective
>> criteria...
>
>I'd be absolutely rivetted to know what "...subjective and objective
>criteria" exist that can overcome a 45 run difference in batting average.
The list only covers those who played since 1969, when ODIs began.
--
John Hall
"Acting is merely the art of keeping a large group of people
from coughing."
Sir Ralph Richardson (1902-83)
Andrew Dunford
"Don speaks the truth" <don2...@rediffmail.com> wrote in message
news:64ad1ef4-bc69-4774...@a15g2000yqm.googlegroups.com...
If Nirvanam is kettle and alvey the pot, you are an entire kitchen.
Andrew
Howzzat
> always mortal, was NEVER considered immortal, NEVER even expected to
> get to such a high stage and doesn't even get there. Mind playing
> games with Sachin's and Ponting's true value, you reckon?
1. It's not about Sachin vs Ponting
2. If you seriously believe that Sachin was equally (or more)
effective in T2, then we are so far apart that this debate is
pointless. He was 23rd in averages among his peers in T2, for cryin'
out loud! This is not my expectation of Sachin talking, but the
reality of his output vs. his contemporaries.
3. If you choose to play in your way (T1/T2), then I can offer other
candidates whose T1 was superior to Sachin's, and T2 just as
effective. Remember Viv Richards - his T1 was absolutely mind blowing.
Ask anyone who followed his career then - he averaged about 60 after
50 odd tests, but it was more than just the average. It was surreal
watching him bat and just annihilate the opposition - Tendulkar never
reached those heights in tests, not even close. Viv's average fell
away in the later years, but you can argue that he was just as
effective as Tendulkar in T2. And has more than a reasonable claim to
superiority over Tendulkar in ODIs.
You will not find too many people who buy into your argument of
splitting a career into two phases. How many people really consider
Botham as the best 1-2 allrounders ever? A lot if you went down the T1/
T2 route, but the reality is that he was pretty average in the second
half of his career so is considered "just" one of the best 5-6
allrounders of the modern era, not the best ever or the equal of
Sobers. In the above sentence, replace Botham with Tendulkar, Sobers
with Bradman, and allrounder with batsman, and you'll get my drift :-)
Nirvanam
I am not the one who chose to break their careers into two halves. In
fact I always looked it at it as a whole. It was you or someone else
who said that this decade Ponting was better (which I agree to). The
original point of this thread was that in a numerical analysis done by
a statistician with cricinfo Sachin came out as the best batsman the
world has seen in both Tests and ODIs together. And while coming out
on top, he is ahead by 10% points from the second. This analysis is
based on career not specific timeboxes or specific pitches or specific
oppositions.
Just coz Ponting had a purple patch in the mid noughties which
coincided with Sachin's dip in form, does not make Ponting a greater
bat. Across their two overlapping careers which is 14 yrs now, Ponting
has done better than Sachin only in 4-5 yrs which coincided during his
purple patch. Once Sachin came back from injury all clear he has
overtaken Ponting (which coincides with Ponting's drop to mortality).
I don't see how you can ever say Sachin is not a greater batsman than
Ponting? Seriously, on what basis? Lay out your criteria and I'll
argue with you. I'll prove it to you that Sachin is greater than
Ponting.
You bring forth any damn batsman thru out the history of the game
(except Don on numbers alone and Lara...these 3 fellas are virtuosos),
and I'll prove it to you that Sachin is better than anyone you can
throw up. Bring forth your criteria...I challenge you.
Howzzat
Dude - I don't understand your obsession with Ponting. Regardless of
what is being discussed, you make everything about Tendulkar vs.
Ponting. And where did I say that Ponting is better than Tendulkar? Go
back and real all posts and let me know where I claimed that?
I have already listed my criteria in a previous post - most of which
Sachin failed, and you conveniently chose to ignore. I am sure you can
make Sachin appear better than anyone based on specious arguments and
manipulated statistics - I can do the same for at least 20 other
batsmen in cricketing history. The only thing going unarguably in
Tendulkar's favor is his longevity (and therefore the highest
aggregate). At the end of the day, that is the sole basis of your (and
Anant's) argument - everything else is highly subjective and specious.
I have no doubt that if Tendulkar played for another 10 years and
averaged 10-15 points less than his previous record and/or his
contemporaries, he will become an even greater batsman in your eyes -
certainly would via Anant's methodology. But my intent here is not to
change your mind about what you strongly believe. And you can't change
mine (and almost everyone else who posts here) no matter which "proof"
you come up with.
higgs
> > Look - I never called you a clueless Sachin-fanatic. You are as
> > entitled to your opinion as I am to mine. And anyway, this was not a
> > response to your post per se - I was actually pointing out flaws in
> > that analyst's methodology. I would have responded similarly if
> > someone else had posted that analysis. So no reason to take this
> > personally.
>
> Nah...I didn't take it personally dude.
>
> > Regarding the points you raised:
> > 1. I am aware of the Wisden 100 list (and the subsequent debate it
> > generated), but didn't know that it was compiled by the same person
> > who did this analysis. Regardless, my point about the flaw in his
> > methodology still stands. He has overweighted the "aggregate" runs -
> > category in which Sachin is the clear leader. Do you seriously buy
> > that? You could tweak some parameters easily to make any of the other
> > greats come out on top. The only (mild) criticism I have of you is
> > your glee in treating this analysis as gospel, just because the
> > outcome suits your pov.
>
> Nope I haven't treated this as gospel at all.
> Also, just on the methodology you might want to consider that over-all
> aggregate bit a lil more...if you read the article Anat has actually
> normalized the aggregate values. You are only looking at the table
> displayed...read the whole article and you'll see he has normalized
> even the aggregate.
>
> > 2. Your point about career vs temporal analysis is laughable. The
> > "temporal" analysis is the last TEN years - the entire decade -,
> > during which Tendulkar played 89 of his 162 tests (55%). This is not
> > selective, unlike the "Ponting vs Tendulkar" post yesterday showing
> > averages from the last 2 years.
>
> It is temporal. There was a period between 2004 and 2007 when Ponting
> scored heavily and Sachin was seriously injured (every other batsman
> of any note averaged more than Sachin during those 3 yrs). Now, if you
> remove those 3 yrs you'll see Sachin averages more than Ponting in all
> other years of the decade. If not more in each and every year, at
> least not significantly behind Ponting (not even 5 runs behind Is'd
> think).
>
Given your admiration for longevity and aggregate runs, if you take
out those three years, Tendulkar is no longer the leading aggregate
run getter nor has his career been as lengthy as some others.
Of course, if you take out those three years, it ups his average, and
with a bit of 'normalising', and due regard for both subjective and
objective data, TAV would have an average in excess of 99.94 for both
Test and odo........
Higgs
dechucka
"higgs" <kenh...@hotmail.com> wrote in message
news:b4a524d2-6e34-43ad...@e37g2000yqn.googlegroups.com...
Higgs
______________________________________________
Your right Huggies stats can proove anything. For instance by normalising
Pontings Test career by taking out any innings when he shouldn't have been
given out we are reduced to 27 innings and an infinite average
:-)
Trying to compare players in the same era is next to impossible as there are
so many variables. Comparing players between eras is impossible IMO. I have
read arguments where Bradam was not the greatest batsman and some valid
points were made ( I don't agree with this argument )
Nirvanam
> what is being discussed, you make everything about Tendulkar vs.
> Ponting. And where did I say that Ponting is better than Tendulkar? Go
> back and real all posts and let me know where I claimed that?
Yaar, I did not bring Ponting into the discussion either. My original
pst was just a pointer to an analysis done on cricinfo. Someone down
the thread brought Ponting into the discussion.
> I have already listed my criteria in a previous post - most of which
> Sachin failed, and you conveniently chose to ignore. I am sure you can
> make Sachin appear better than anyone based on specious arguments and
> manipulated statistics - I can do the same for at least 20 other
> batsmen in cricketing history. The only thing going unarguably in
> Tendulkar's favor is his longevity (and therefore the highest
> aggregate). At the end of the day, that is the sole basis of your (and
> Anant's) argument - everything else is highly subjective and specious.
> I have no doubt that if Tendulkar played for another 10 years and
> averaged 10-15 points less than his previous record and/or his
> contemporaries, he will become an even greater batsman in your eyes -
> certainly would via Anant's methodology. But my intent here is not to
> change your mind about what you strongly believe. And you can't change
> mine (and almost everyone else who posts here) no matter which "proof"
> you come up with.
Frankly, I don't remember those criteria...why not just lay it down or
direct me to those criteria?
Longevity is not the sole basis for my argument of Sachin being the
greatest batsman along with Lara and Don. That is why I am readily
challenging anyone who can throw up an candidate other than Lara and
Don on comprehensive criteria.
I mean, seriously, just try it yourself and you'll find out why I have
so much conviction behind this argument. Alright, just present your
comprehensive criteria and present 1 or n number of batsmen that you
think maybe better than Sachin. And let's see if they truly are
better? Fair?
I am not trying to change your mind. I am only trying to understand
why u think Sachin is not the best. Once you give me the reasons, if I
can logically challenge them I'll present evidence to the
contrary...that's all.
Nirvanam
Higgs, First you show me one post where I have said longevity and
aggregates are everything? Why can't you do the easier thing
instead...throw up your comprehensive criteria, throw up your
candidates and let's see if any of your candidates...ANY of your
candidates other than Don and Lara can beat Sachin.
Such an open, all-encompassing (Don and Lara out) challenge is rarely
made by anyone. Why don't you prove me wrong if you think I am wrong?
Just try it, mate. Why do you not have the conviction in your opinions
to challenge me? Take it up mate...challenge me.
If you'd like, I'll challenge you to show me one batsman (Don, Lara
excluded) better than Sachin on a comprehensive set of criteria to
encompass the whole criteria (no need to remove those 3 or 4 injury
years at all).
jzfredricks
> why u think Sachin is not the best.
Because The Don was easily better.
Nirvanam
If you are going to bring numbers to argue Don's case....well there is
no case at all. I myself say there is no one better than Don on that
and that is as obvious as obvious can be. But if you put context
around those fabulous stats then I can argue why Sachin is not behind
Don...why they are equal...rather why there is no point in attempting
to say one is better than the other.
JZ, please try to understand my point of view here. I am nowhere
claiming that Sachin is greater than Don. But I am saying Sachin is
not lesser than Don and so is Lara not lesser than Don. These 3 are
virtuosos....the most perfect batsmen our game has produced. No other
batsmen in the history of the game can compare with them. All other
batsmen come close but lose out on one or the other virtuoso-defining
parameter. Arguing for Sachin and Lara's supreme batsmanship is not a
denial of Don's supreme batsmanship. Clubbing Sachin and Lara with Don
does not bring Don down nor does it make Sachin and Lara go higher.
Each one of those 3 deserves to be in the virtuoso bracket.
Howzzat
And that's fair enough. However, the problem arises because not
everyone measures greatness by this nebulous "virtuoso" criteria,
which is as subjective as subjective can be. You have been asking for
criteria, so here are some off the top of my head. While writing these
down, I am not in any way trying to bring SRT down (how can I when I
am his #1 fan - ok maybe #2 after you). For someone to be considered
among the top 3 batsmen of all time, they must check most, if not all
of the following boxes. In no particular order:
1. Longevity, aggregate, number of centuries
2. Some combination of average/strike rate comfortably better than
their contemporaries
3. At least 1-2 monster innings in their career (say 300+)
4. At least 1-2 monster series in their career (say 500+ runs)
5. At least 1-2 instances of 2 hundreds in a test match (over a long
enough career of, say, 100+ tests)
6. Break through the "holy grail" - 900 points barrier on ICC rating
points at least once in their career. Arbitrary I know, but very hard
to achieve since it requires sustained peak performance over a stretch
of time. Bonus points if they can stay there for any significant
length of time
7. A handful (3-4) of very very special innings that are timeless and
would make any "top 100" list by default
8. Excellence in all conditions against all comers (perhaps allowing
for 1 outlier due to whatever circumstances/excuses)
8. Through their playing career, being considered as the top 3-4 bats
in the world every year, for majority of their careers
I am not going to offer names or evaluate players just yet, because I
want to know what you think of this. I probably missed a couple, but I
think it's important to get on a similar wavelength while evaluating
players, otherwise this will go on forever. You will notice that I
have left out the more subjective aspects like elegance, off field
conduct etc. because they only muddy the waters. We should stick to
the "measurable" aspects so this can actually get somewhere.
Nirvanam
batsmen more. Instead I suggest we have a cut-off of minimum number of
centuries (say 10, 15?) and then use Tests per century or Tests per
Innings as the more appropriate metric.
Also Longevity would disadvantage the guys who did not play high
number of Tests in years gone by. Instead we use peer comparison for
longevity. So let's say a guy who debuted in the 1970s played for 10
yrs. And the average number of Tests played by a country during those
10 yrs was say 85. Then if the guy is involved in all 85 he gets a
perfect score...and proportionately reduce index.
> 2. Some combination of average/strike rate comfortably better than
> their contemporaries
SR will be a disadvantage for many of history's greatest batsmen
because they were not tracked as religiously as they do now-a-days.
So, you could remove the SR thing
> 3. At least 1-2 monster innings in their career (say 300+)
How many triple centurions are there in the history of the game? To
date there are only 3 batsmen who have scored 2 triples. Maybe 20-25
individuals who have scored triples. It disadvantages guys like
Richards, Sunny, Ponting, Sachin, 3 Ws, Pollock, Headley, etc. Or if
you'd like to keep it at 300 itself, then you'll need to reduce the
weightage provided to this metric.
Here's a suggestion: how about the average number of runs above 100 in
all the guys centuries...not outs add to the denominator unlike
batting average....or ignore the not out innings.
> 4. At least 1-2 monster series in their career (say 500+ runs)
Monster series definition will need to change to Monster Average
Series. It depends on opportunities don't you think? It would terribly
disadvantage players who did not play many 3+ Test-series. Many modern
batsmen will get disadvantaged.
> 5. At least 1-2 instances of 2 hundreds in a test match (over a long
> enough career of, say, 100+ tests)
> 6. Break through the "holy grail" - 900 points barrier on ICC rating
> points at least once in their career. Arbitrary I know, but very hard
> to achieve since it requires sustained peak performance over a stretch
> of time. Bonus points if they can stay there for any significant
> length of time
Should this be binary or should it be a distance from or beyond 900
points? Also the time for which the batsman was ahead of by 800
(arbitrary) points would give a reasonable insight into the sustenance
of great batsmanship. Also, think over this rating because many of the
modern and pre 60s batsmen will get an advantage on this because they
played in an age when the average bowler was not as good as let's say
the 70s or 80s. So although the bowler (for ex - Kulasekhara) rating
is 800 and the bowler is the best at that point in time, his rating
would've been probably 600-700 had there been better bowlers around
him. Think this one through.
> 7. A handful (3-4) of very very special innings that are timeless and
> would make any "top 100" list by default
Whose 100's list? I am not sure there are any such lists except one
which Wisden had done way back in 2000. This would disadvantage the
guys who scored centuries after the year 2000.
> 8. Excellence in all conditions against all comers (perhaps allowing
> for 1 outlier due to whatever circumstances/excuses)
What is outlier here? Does outlier mean an innings, a Test, or an
entire 'type of pitch'? Here's what I suggest: we pick the 3 most
difficult surface types i.e. Speed Kings, Swing Havens, and Dust
Bowls. It is reasonable to expect that a truly great virtuoso will be
consistent on all 3 difficult tracks? Also to avoid averages of 2.06
and 3.13 and 2.67 showing up as a greatly consistent, we may want to
say that at least in 2 of these 3 tracks the batsman should've had an
average of 40 (whatever)?
Recently I started a post on identifying the different grounds in the
world which will fall under each category. We can revisit that thread
and finalize the list.
> 8. Through their playing career, being considered as the top 3-4 bats
> in the world every year, for majority of their careers
Considered by whom? The ICC rating point should address this one
although not directly. I am raising this because many of the past
batsmen may not have been considered greats in a particular year but
we just remember them as great...so we tend to say that they were
great as a whole.
> I am not going to offer names or evaluate players just yet, because I
> want to know what you think of this. I probably missed a couple, but I
> think it's important to get on a similar wavelength while evaluating
> players, otherwise this will go on forever. You will notice that I
> have left out the more subjective aspects like elegance, off field
> conduct etc. because they only muddy the waters. We should stick to
> the "measurable" aspects so this can actually get somewhere.
Nice effort dude. I think all the criteria seem good. Just added my
thoughts from the practicalities and advantage/disadvantage caused to
batsmen of different ages.
Hey by the way there's one more very important criteria...what is the
minimum number of Tests / Innings for a batsman to have played. Also
what is the minimum number of countries / grounds on which the guy
played.
We'll give it a serious shot dude and see what we can discover. Also,
I suggest let's not use Don for peer calculations and keep him out of
this...he will disadvantage his peers in peer-comparison by a heavy
amount and anyway there is never an argument of whether Don is the
best.
jzfredricks
Trust me, I do. My opinion just differs to yours. It also differs so
much that I think anyone who objectively tries to rate *any* batsman
higher than Bradman is a nutter.
As some/most of you know, I'm an Englishman living in Oz. Strangely, I
still support England. Always have, always will. There's something
about Oz cricket I just don't like (no offence my Aussie comrades)(it
probably boils down to "the way it's played"). I'm just letting you
know all that so that my position on The Don is not one based on blind
love. You'll never catch me calling him "immortal" (unlike others
w.r.t.SRT).
Like all things, I look at the facts, have a think about it, and make
a decision. This is an easy one.
Read some more threads here, and you'll also see me pimping SRT. He's
a wonderful batsman, the "best" I've ever seen. I say "between Lara
and SRT: Lara to bat for my entertainment, SRT to bat for my life".
If we could magically pick an all-time cricket team, you'd be mad not
to pick Bradman first. I think most students of the game would.
Howzzat
total nerds, let's give it a shot...
> I'd also think that this will advantage Sachin, and a few other modern
> batsmen more. Instead I suggest we have a cut-off of minimum number of
> centuries (say 10, 15?) and then use Tests per century or Tests per
> Innings as the more appropriate metric.
> Also Longevity would disadvantage the guys who did not play high
> number of Tests in years gone by. Instead we use peer comparison for
> longevity. So let's say a guy who debuted in the 1970s played for 10
> yrs. And the average number of Tests played by a country during those
> 10 yrs was say 85. Then if the guy is involved in all 85 he gets a
> perfect score...and proportionately reduce index.
>
Fine. To keep it simple, let's say we only consider players with >6000
Test runs - that address longevity and aggregate. Or you can raise the
threshold to reduce the pool further. Then the metric for those who
qualify would be innings per century scored.
To be fair to older players, we should exclude performances against
Bangladesh. Zimbabwe I think are OK since they were relatively decent
for most of the time they were playing test cricket.
> > 2. Some combination of average/strike rate comfortably better than
> > their contemporaries
>
> SR will be a disadvantage for many of history's greatest batsmen
> because they were not tracked as religiously as they do now-a-days.
> So, you could remove the SR thing
OK. So average only, but not absolute. An average of 50 in the 1970s
is not the same as in 2000s. It has to be relative to the average of
the best performing batsman of their era.
>
> > 3. At least 1-2 monster innings in their career (say 300+)
>
> How many triple centurions are there in the history of the game? To
> date there are only 3 batsmen who have scored 2 triples. Maybe 20-25
> individuals who have scored triples. It disadvantages guys like
> Richards, Sunny, Ponting, Sachin, 3 Ws, Pollock, Headley, etc. Or if
> you'd like to keep it at 300 itself, then you'll need to reduce the
> weightage provided to this metric.
>
> Here's a suggestion: how about the average number of runs above 100 in
> all the guys centuries...not outs add to the denominator unlike
> batting average....or ignore the not out innings.
Hmm - this gets complicated. What you are suggesting does not reward
someone like Lara enough for monstrous scores like 375 and 400 - one
of the main reasons he is held in such high regard. There's got to be
something that considers the relative magnitude of, say, the 3 highest
test scores of each batsman.
>
> > 4. At least 1-2 monster series in their career (say 500+ runs)
>
> Monster series definition will need to change to Monster Average
> Series. It depends on opportunities don't you think? It would terribly
> disadvantage players who did not play many 3+ Test-series. Many modern
> batsmen will get disadvantaged.
So how about the number of series in which a batsman averaged say 125
runs per test. Minimum qualification is a 3-test series. Would also
need to normalize since those who have played more would be
advantaged. So something like the % of 3+ test series in which the
batsman averaged 125 (?) runs per test
>
> > 5. At least 1-2 instances of 2 hundreds in a test match (over a long
> > enough career of, say, 100+ tests)
> > 6. Break through the "holy grail" - 900 points barrier on ICC rating
> > points at least once in their career. Arbitrary I know, but very hard
> > to achieve since it requires sustained peak performance over a stretch
> > of time. Bonus points if they can stay there for any significant
> > length of time
>
> Should this be binary or should it be a distance from or beyond 900
> points? Also the time for which the batsman was ahead of by 800
> (arbitrary) points would give a reasonable insight into the sustenance
> of great batsmanship. Also, think over this rating because many of the
> modern and pre 60s batsmen will get an advantage on this because they
> played in an age when the average bowler was not as good as let's say
> the 70s or 80s. So although the bowler (for ex - Kulasekhara) rating
> is 800 and the bowler is the best at that point in time, his rating
> would've been probably 600-700 had there been better bowlers around
> him. Think this one through.
That's why the comparison across eras is so difficult. But I think we
should leverage the ICC ratings points somehow, since they have
already done a lot of the analysis we are trying to do. Percentage of
a player's career at a rating of 800+ would be fine I think.
>
> > 7. A handful (3-4) of very very special innings that are timeless and
> > would make any "top 100" list by default
>
> Whose 100's list? I am not sure there are any such lists except one
> which Wisden had done way back in 2000. This would disadvantage the
> guys who scored centuries after the year 2000.
I agree this one is subjective. But an important measure nonetheless -
I don't know if anyone else wants to jump in with suggestions.
>
> > 8. Excellence in all conditions against all comers (perhaps allowing
> > for 1 outlier due to whatever circumstances/excuses)
>
> What is outlier here? Does outlier mean an innings, a Test, or an
> entire 'type of pitch'? Here's what I suggest: we pick the 3 most
> difficult surface types i.e. Speed Kings, Swing Havens, and Dust
> Bowls. It is reasonable to expect that a truly great virtuoso will be
> consistent on all 3 difficult tracks? Also to avoid averages of 2.06
> and 3.13 and 2.67 showing up as a greatly consistent, we may want to
> say that at least in 2 of these 3 tracks the batsman should've had an
> average of 40 (whatever)?
> Recently I started a post on identifying the different grounds in the
> world which will fall under each category. We can revisit that thread
> and finalize the list.
I think what you are suggesting would be too complicated. Why not just
say that the batsman must average 45+ in each country where they
played a minimum of say 5 tests.
>
> > 8. Through their playing career, being considered as the top 3-4 bats
> > in the world every year, for majority of their careers
>
> Considered by whom? The ICC rating point should address this one
> although not directly. I am raising this because many of the past
> batsmen may not have been considered greats in a particular year but
> we just remember them as great...so we tend to say that they were
> great as a whole.
I agree. This would be subjective too, and the ICC ratings points (%
of playing career spent over a certain threshold) addresses this
indirectly.
>
> > I am not going to offer names or evaluate players just yet, because I
> > want to know what you think of this. I probably missed a couple, but I
> > think it's important to get on a similar wavelength while evaluating
> > players, otherwise this will go on forever. You will notice that I
> > have left out the more subjective aspects like elegance, off field
> > conduct etc. because they only muddy the waters. We should stick to
> > the "measurable" aspects so this can actually get somewhere.
>
> Nice effort dude. I think all the criteria seem good. Just added my
> thoughts from the practicalities and advantage/disadvantage caused to
> batsmen of different ages.
>
> Hey by the way there's one more very important criteria...what is the
> minimum number of Tests / Innings for a batsman to have played. Also
> what is the minimum number of countries / grounds on which the guy
> played.
I suggested min 6000 Test runs above - that should be a good enough
threshold. Don't see how you can address the countries/grounds issue
though - for example you'd have to exclude oldies like Bradman if you
did.
>
> We'll give it a serious shot dude and see what we can discover. Also,
> I suggest let's not use Don for peer calculations and keep him out of
> this...he will disadvantage his peers in peer-comparison by a heavy
> amount and anyway there is never an argument of whether Don is the
> best.
Yes sounds good.
So this is what we have so far as the criteria. Next we'll need to
figure out the methodology to allocate points and weights to each
criteria.
Qualification: 6000 test runs
Analysis excludes Bangladesh
1. Average number of innings per century
2. Batting average relative to the best performing batsman of their
era
3. Relative magnitude of batsman's 3 highest test scores
4. % of 3+ test series in which the batsman averaged 125 (?) runs per
test
5. Number of times a batsman scored 2 hundred per test - normalized by
number of tests played
6. Percentage of a player's career at ICC rating of 800+
7. Some TBD metric of the relative number of ATG innings played
8. Number of countries in which the player averaged <45 (min 5 tests)
That's all I can think of for now. A nice one to add would be the
frequency of playing match-winning innings, but that would seriously
disadvantage players from weaker teams. That is why we need something
like #7 at least. I'd invite others to jump in here if they find this
amusing :-)
Finally, I'd suggest that we do a dry run with 10 or so players. At
the end of the day, we are just trying to prove (or disprove) your
hypothesis that no player in the history of test cricket is better
than Tendulkar, apart from Bradman and Lara who are his equals -
correct?
Here's my suggestion for moving the analysis forward. For the pool of
players under consideration, let's assign them points out of 10 for
each of the criteria. Then we'll assign a weight to each criteria and
do the math. Alternatively, if we working with only ten players, we
can just rank them from 1-10 for each of the criteria above, then do a
weighted average.
Ben Gussey
> Can't believe we are actually doing this. At the risk of looking like
> total nerds, let's give it a shot...
>
>
> > > 3. At least 1-2 monster innings in their career (say 300+)
>
> > How many triple centurions are there in the history of the game? To
> > date there are only 3 batsmen who have scored 2 triples. Maybe 20-25
> > individuals who have scored triples. It disadvantages guys like
> > Richards, Sunny, Ponting, Sachin, 3 Ws, Pollock, Headley, etc. Or if
> > you'd like to keep it at 300 itself, then you'll need to reduce the
> > weightage provided to this metric.
>
> > Here's a suggestion: how about the average number of runs above 100 in
> > all the guys centuries...not outs add to the denominator unlike
> > batting average....or ignore the not out innings.
>
> Hmm - this gets complicated. What you are suggesting does not reward
> someone like Lara enough for monstrous scores like 375 and 400 - one
> of the main reasons he is held in such high regard. There's got to be
> something that considers the relative magnitude of, say, the 3 highest
> test scores of each batsman.
I didn't read the rest of this thread so I am not sure what you guys
are getting at but average above 100 would obviously reward someone
who achieved monstrous scores, as Lara's average scores above 100
would be added to considerably by a 375 & 400?
Howzzat
Not necessarily if someone's got a lot of hundreds - then the effect
gets diluted. For the sake of discussion (with bothering to look up
real numbers) let's say Tendulkar or Waugh's average 100+ score is 140
- I am sure Lara's (or any other contemporary player's) will at most
be in 150/160 range. Higher, but the effect has been diluted due the
large sample size of centuries. Not everyone goes out every other day
and smashes 375 or 400 - it must carry significant weight while
evaluating someone. Put another way, would you consider Lara the
player he is if he had not made those two scores, but had been
dismissed at 150 odd instead, and to compensate, had those extra
300-400 runs spread equally over his remaining 36 hundreds?
Which is why I suggested to look at the top 3 scores each batsman has
managed in their careers. Lara would score here, while someone like
Kallis would (and should) be disadvantaged - IIRC he doesn't even have
a double hundred? If I was assigning points out of 10, Lara would get
a 10 while Kallis probably a 3 or 4. I am sure things would be much
closer if you just calculated the average score above 100 - I would
not be overly surprised if Kallis came within touching distance of
Lara.
Howzzat
OK - so I just ran the numbers on Statsguru. Prepare to be amazed.
We all know Lara has scored 400*, 375, 277 etc. Kallis' highest score,
OTOH, is 189*.
Here are their filtered records when you only include innings > 100.
Lara
33 5889 400* 184.03 34
Kallis
32 4396 189* 199.81 33
Kallis's average score over 100 is not only higher than Lara's, but
also more than his own highest score :-)
Clearly this is due to the number of not-out hundreds he's got, but
there is no satisfactory way around it. You can't ignore the not-outs
altogether either, because Kallis might claim he would have converted
each one of those into 400+ had he been allowed to continue batting.
Nirvanam
> > JZ, please try to understand my point of view here.
>
> Trust me, I do. My opinion just differs to yours. It also differs so
> much that I think anyone who objectively tries to rate *any* batsman
> higher than Bradman is a nutter.
But that's exactly where you are misunderstanding or misinterpreting
my opinion. I have never said there is anyone better than Bradman.
There is a difference in saying x, y and z are in the same set of
virtuosos i.e. U( batting Virtuosos) = {x, y, z} and saying x > y, no?
jzfredricks
> > much that I think anyone who objectively tries to rate *any* batsman
> > higher than Bradman is a nutter.
>
> But that's exactly where you are misunderstanding or misinterpreting
> my opinion.
Please allow me to change what I said from "better than" to "better
than or equal to".
Andrew B.
> > JZ, please try to understand my point of view here.
>
> Trust me, I do. My opinion just differs to yours. It also differs so
> much that I think anyone who objectively tries to rate *any* batsman
> higher than Bradman is a nutter.
You could make a case for Grace (even further ahead of his
contemporaries, took batting to a new level). Not a very good case,
though.
Re Tendulkar and Lara - I wouldn't dismiss out of hand the idea that
their the best batsmen since Bradman (though I wouldn't necessarily
agree, either). But I can't see any reason for saying they're "a class
above" the rest.
jzfredricks
> contemporaries, took batting to a new level). Not a very good case,
> though.
Grace would be my #2 pick, if someone didn't take him first!
Nirvanam
> Fine. To keep it simple, let's say we only consider players with >6000
> Test runs - that address longevity and aggregate. Or you can raise the
> threshold to reduce the pool further. Then the metric for those who
> qualify would be innings per century scored.
Sure, this sounds good. Maybe the formula for inns / century can be
modified to count only the 'out'-inns / centuries? Whatever, we can
include both of them and see how the data behaves.
> To be fair to older players, we should exclude performances against
> Bangladesh. Zimbabwe I think are OK since they were relatively decent
> for most of the time they were playing test cricket.
This is actually not fair. I'll tell why: Each team except Eng/Aus was
a minnow at some point in time. Therefore we need to define minnows
temporally. Here's what I suggest: a team has gained status when it
wins its first series (home or abroad) or has ensured consistent away
draws/wins. From then onwards the team can be moved to the higher
grade. I did a similar thing some time ago. Below is the suggested
years for each team and you review them and see if we need to change
the dates, ok?
Aus - 1877
Ban - not yet
Eng - 1877
Ind - 1952
NZ - 1949
Pak - 1952
RSA - 1930
SL - 1985
WI - 1934
Zim - 1994 to 2003
Alright, now do you want to exclude the data of minnows or just reduce
their weight?
Also we cannot penalize a batsman who has scored against the minnows.
If he is good enough he will have scored against others as well. We
could also consider centuries in diff countries?
> OK. So average only, but not absolute. An average of 50 in the 1970s
> is not the same as in 2000s. It has to be relative to the average of
> the best performing batsman of their era.
No. not best performing batsman but the general average. It will be a
huge huge huge disadvantage for guys who played during Don's time.
Secondly we cannot arbitrarily choose a 10 yr period. Instead we have
a running peer-comparison-average. For example, Dravid will not have
just 2 averages i.e. 90s and 00s but will have averages for each of
the years 1996 to 2009. Maybe we can change it to season instead of
calendar year.
> Hmm - this gets complicated. What you are suggesting does not reward
> someone like Lara enough for monstrous scores like 375 and 400 - one
> of the main reasons he is held in such high regard. There's got to be
> something that considers the relative magnitude of, say, the 3 highest
> test scores of each batsman.
It will reward them by not counting the not outs. Ex let's say 3
scores of 100, 100*, 100. The average will be 300/2 instead of 300/3
Another thought is how many of the centuries were big ones (not
monsters but >150) i.e. percentage of big ones (again not outs to be
included in >150)
> So how about the number of series in which a batsman averaged say 125
> runs per test. Minimum qualification is a 3-test series. Would also
> need to normalize since those who have played more would be
> advantaged. So something like the % of 3+ test series in which the
> batsman averaged 125 (?) runs per test
We'll use the regular average itself, no? Also we need to combine this
with standard deviation because a guy who has let's say 9 of 10 series
where he averaged 50 will lose out to a guy who averaged 75 in 2
series and averaged maybe 38 in the other 8. So we need to reward both
peak performance and consistency.
> That's why the comparison across eras is so difficult. But I think we
> should leverage the ICC ratings points somehow, since they have
> already done a lot of the analysis we are trying to do. Percentage of
> a player's career at a rating of 800+ would be fine I think.
Cool, let's use this one.
> > Whose 100's list? I am not sure there are any such lists except one
> > which Wisden had done way back in 2000. This would disadvantage the
> > guys who scored centuries after the year 2000.
>
> I agree this one is subjective. But an important measure nonetheless -
> I don't know if anyone else wants to jump in with suggestions.
I suggest we leave this one out for now. Maybe include notes at the
end of analysis that "by the way mr.x also had n of his centuries in
the wisden top 100 100's of the 20th century)
> I think what you are suggesting would be too complicated. Why not just
> say that the batsman must average 45+ in each country where they
> played a minimum of say 5 tests.
That's fine but there are some very well established ordeal surfaces
in the world of cricket for decades now. Let's leverage them. Also if
we go by country then lots of the older players will either be greatly
advantaged or greatly disadvantaged because they played in few
countries. For ex, Headley played in 2 countries only whereas Waugh
played in 10. So if Headley has an average of 85 in one and 20 in the
other then he will be disadvantaged, otoh if he had 75 and 50 whereas
Waugh has a consistent 50-55 in 9 countries and a 25 in the 10th then
it is unfair to Waugh.
Second thing is that we can say 8 innings instead of 5 tests?
> So this is what we have so far as the criteria. Next we'll need to
> figure out the methodology to allocate points and weights to each
> criteria.
For methodology we use a technique called Analytical Hierarchy Process
or AHP. It is a combination of pair-wise comparison and weighted
evaluation on criteria. Here are some links on AHP:
http://en.wikipedia.org/wiki/Analytic_Hierarchy_Process
http://www.cs.toronto.edu/~sme/CSC340F/slides/tutorial-prioritization.pdf
http://thequalityportal.com/q_ahp.htm
For pairwise comparison ppl use different ways. For ex, if A and B are
being compared on a criteria C. Then, in some cases subjectively ppl
say A-B is High, Medium, Low, or 9, 3, 1 or 5, 4, 3, 2, 1. Those are
subjective associations. Luckily we are dealing with numbers directly
so what we can do is:
Let's say we are comparing average of A and B. A is 54 and B is 57. So
in the cell A v/s B the value will be A/B = 0.95 and in cell B v/s A
the value will be B/A = 1 / (A/B). So at any point u will see that the
average of both cells will be 1. What we are doing essentially is to
proportionately increase or decrease rating on every pair.
If you already know AHP then ignore the above else when you view those
links you'll be able to relate to this quite easily.
I'll create an excel template for this.
> That's all I can think of for now. A nice one to add would be the
> frequency of playing match-winning innings, but that would seriously
> disadvantage players from weaker teams. That is why we need something
> like #7 at least. I'd invite others to jump in here if they find this
> amusing :-)
Its not fair to associate wins on a batsman's performance...its a team
game. What we can do though is to consider only the wins and see how
significant the contribution of the batsman was in them. For ex - if a
batsman scores a 100 or 150+ in both innings and is also the top
scorer in his team for the game, or not less than 25 runs of the top
scorer then his contribution is "significant". If a batsman scores 2
fifties, or 75+ in both innings combined (provided there are others
who have scored more) then his contribution will be regarded as
"supportive". What say? We could also look at the consistency factor
of performance in wins, draws, losses.
Finally, we can also reward based on top class bowlers faced by the
batsman. How about career-to-date averages of bowlers but that'll be
very complicated. Rating points of bowlers in specific innings (ex
centuries)?
> Finally, I'd suggest that we do a dry run with 10 or so players. At
> the end of the day, we are just trying to prove (or disprove) your
> hypothesis that no player in the history of test cricket is better
> than Tendulkar, apart from Bradman and Lara who are his equals -
> correct?
Yep
> Here's my suggestion for moving the analysis forward. For the pool of
> players under consideration, let's assign them points out of 10 for
> each of the criteria. Then we'll assign a weight to each criteria and
> do the math. Alternatively, if we working with only ten players, we
> can just rank them from 1-10 for each of the criteria above, then do a
> weighted average.
10 players will be good I think. What I'll suggest is let's look at
all the criteria and intuitively figure out which 10 batsmen will
address all these criteria very well. Then we both look at the list
and pick the common ones (Sachin, Lara, Ponting get a default
entry...they are the motivation behind the analysis anyway lol).
Folks, help us with your points
Howzzat
will try to do a more back of the envelope calculation in excel during
the weekend and post here. That should give us some idea, but your
analysis will likely be more thorough and accurate.
Some comments below:
I thought we were excluding Don's stats for peer comparisons. Anyway,
on the above two points, it's fine to do it your way, but it's a lot
more work. What I was trying to attempt was not very scientific, but
if you have the bandwidth feel free. Like I said I will try to do a
rough calculation first.
>
> > Hmm - this gets complicated. What you are suggesting does not reward
> > someone like Lara enough for monstrous scores like 375 and 400 - one
> > of the main reasons he is held in such high regard. There's got to be
> > something that considers the relative magnitude of, say, the 3 highest
> > test scores of each batsman.
>
> It will reward them by not counting the not outs. Ex let's say 3
> scores of 100, 100*, 100. The average will be 300/2 instead of 300/3
> Another thought is how many of the centuries were big ones (not
> monsters but >150) i.e. percentage of big ones (again not outs to be
> included in >150)
This is seriously flawed. Did you see my post on this type of math
with Kallis and Lara examples? It just doesn't work.
And anyway, why try to devise a derivative metric when you can just
look at their top 3 or so scores and make a call on their ability of
play huge knocks?
>
> > So how about the number of series in which a batsman averaged say 125
> > runs per test. Minimum qualification is a 3-test series. Would also
> > need to normalize since those who have played more would be
> > advantaged. So something like the % of 3+ test series in which the
> > batsman averaged 125 (?) runs per test
>
> We'll use the regular average itself, no? Also we need to combine this
> with standard deviation because a guy who has let's say 9 of 10 series
> where he averaged 50 will lose out to a guy who averaged 75 in 2
> series and averaged maybe 38 in the other 8. So we need to reward both
> peak performance and consistency.
Average we are already covering as a separate metric, and that in a
way accounts for consistency. My point here is the same as the
monstrous innings one. There is something very special (and rare)
about the ability to play a 375 or 400 knock, or to score 650 runs in
a 3-test series in Murali's backyard. By averaging this out over a
long career, you are failing to reward someone like Lara enough for
these incredible achievements. Lara's overall average is not as high
as some of his peers, so he will take a hit (and rightly so) on the
average metric for his relative lack of consistency. But at the same
time he deserves extra credit for those monstrous peaks. If you don't
allow for that, players like Kallis will come out ahead of geniuses
like Lara every time.
>
> > I think what you are suggesting would be too complicated. Why not just
> > say that the batsman must average 45+ in each country where they
> > played a minimum of say 5 tests.
>
> That's fine but there are some very well established ordeal surfaces
> in the world of cricket for decades now. Let's leverage them. Also if
> we go by country then lots of the older players will either be greatly
> advantaged or greatly disadvantaged because they played in few
> countries. For ex, Headley played in 2 countries only whereas Waugh
> played in 10. So if Headley has an average of 85 in one and 20 in the
> other then he will be disadvantaged, otoh if he had 75 and 50 whereas
> Waugh has a consistent 50-55 in 9 countries and a 25 in the 10th then
> it is unfair to Waugh.
>
> Second thing is that we can say 8 innings instead of 5 tests?
Sure 8 innings is good. Your Headley/Waugh example is valid, but I
don't know the perfect solution. Your surfaces suggestion will have
the same problem, and would penalize people for failing in only 1-2
innings on that ground. Which is why country as a whole makes more
sense, since you can get a reasonable sample size (8 innings plus).
Generally speaking, if someone had 8+ innings in a country and
averaged less than 45, they deserve to be docked a point, whether it's
Headley or Waugh, although you raise a valid point about the problem
inherent in comparing people across eras.
> For methodology we use a technique called Analytical Hierarchy Process
> or AHP. It is a combination of pair-wise comparison and weighted
> evaluation on criteria. Here are some links on AHP:http://en.wikipedia.org/wiki/Analytic_Hierarchy_Processhttp://www.cs.toronto.edu/~sme/CSC340F/slides/tutorial-prioritization...http://thequalityportal.com/q_ahp.htm
>
Cool. Give this a shot, while I do my quick and dirty analysis.
>
> Its not fair to associate wins on a batsman's performance...its a team
> game. What we can do though is to consider only the wins and see how
> significant the contribution of the batsman was in them. For ex - if a
> batsman scores a 100 or 150+ in both innings and is also the top
> scorer in his team for the game, or not less than 25 runs of the top
> scorer then his contribution is "significant". If a batsman scores 2
> fifties, or 75+ in both innings combined (provided there are others
> who have scored more) then his contribution will be regarded as
> "supportive". What say? We could also look at the consistency factor
> of performance in wins, draws, losses.
Think we'll need to hire a full time analyst to do all this :-)
I realize quantifying matchwinning ability in a team game is hard,
that's why I mentioned it but left it out of the criteria.
>
> Finally, we can also reward based on top class bowlers faced by the
> batsman. How about career-to-date averages of bowlers but that'll be
> very complicated. Rating points of bowlers in specific innings (ex
> centuries)?
Doesn't the ICC rating points take care of that already? I thought
they took into account the quality of bowling faced.
>
> 10 players will be good I think. What I'll suggest is let's look at
> all the criteria and intuitively figure out which 10 batsmen will
> address all these criteria very well. Then we both look at the list
> and pick the common ones (Sachin, Lara, Ponting get a default
> entry...they are the motivation behind the analysis anyway lol).
>
For the first pass, you could consider 10 players from the last 40
years only (plus Bradman since he is one of the subjects in the
hypothesis). We can always extend it later to include the older
players if there is need/interest.
Here's my suggested list:
0. Bradman
1. Lara
2. Tendulkar
3. Ponting
4. Kallis
5. Dravid
6. V. Richards
7. G. Chappell
8. Sobers
9. Gavaskar
10. One of Jayawardene/Sangakkara/Yousuf
Nirvanam
> with Kallis and Lara examples? It just doesn't work.
>
> And anyway, why try to devise a derivative metric when you can just
> look at their top 3 or so scores and make a call on their ability of
> play huge knocks?
I saw the example but in that the not outs were not included hence the
higher than highest score. I don't think we can conclude that a
batsman does not have the ability to score a triple if he hasn't done
it especially for serious batting talents. For ex - Miandad was once
in the 270s and Imran declared...what's to say he wouldn't have got
there? Don was left stranded on 299. Lax and Dravid both of them were
out at 281 and 270 so maybe you could argue (however not strongly)
that they couldn't have. What abt Ponting - 257 (not out, no?).
Sachin's top scores are both not outs 248 and 241. We can only use a
derived metric for this dude else it will be very unfair to the guys
who have at least 1 NO in their top 3. Also consider Lara itself...he
made 400 n.o but he could have even made 500 if allowed to continue.
Let's give this one some more thought.
> Average we are already covering as a separate metric, and that in a
> way accounts for consistency. My point here is the same as the
> monstrous innings one. There is something very special (and rare)
> about the ability to play a 375 or 400 knock, or to score 650 runs in
> a 3-test series in Murali's backyard. By averaging this out over a
> long career, you are failing to reward someone like Lara enough for
> these incredible achievements. Lara's overall average is not as high
> as some of his peers, so he will take a hit (and rightly so) on the
> average metric for his relative lack of consistency. But at the same
> time he deserves extra credit for those monstrous peaks. If you don't
> allow for that, players like Kallis will come out ahead of geniuses
> like Lara every time.
The average will take care of this. For ex we say a cut-off of 60 as a
great series. Let's say a guy has 1 series where he averaged 125 and 9
others where he averaged around 65-70. The 125 will definitely pull up
the entire average. Now suppose another fellow had 4 series averaging
around 90 and 6 around 65-70. He also will be recognized for being
consistent enough to maintain a deadly average. Another case where a
guy has 3 series at 90, 1 series at 125, and 6 at 65-70. Now this guy
has to be the best among all 3 cases...which will come out if we take
the average of top series averages.
> Sure 8 innings is good. Your Headley/Waugh example is valid, but I
> don't know the perfect solution. Your surfaces suggestion will have
> the same problem, and would penalize people for failing in only 1-2
> innings on that ground. Which is why country as a whole makes more
> sense, since you can get a reasonable sample size (8 innings plus).
> Generally speaking, if someone had 8+ innings in a country and
> averaged less than 45, they deserve to be docked a point, whether it's
> Headley or Waugh, although you raise a valid point about the problem
> inherent in comparing people across eras.
I'll give you the surfaces...there are at least 5 in each of the 3
categories. You have a look at the surfaces and then decide if we can
go with surface approach or not...will post it separately.
> Cool. Give this a shot, while I do my quick and dirty analysis.
> Think we'll need to hire a full time analyst to do all this :-)
> I realize quantifying matchwinning ability in a team game is hard,
> that's why I mentioned it but left it out of the criteria.
True...so let's leave this one out. I had already done it for Sachin
and Dravid so we have 2 data points ready...lol!
> Doesn't the ICC rating points take care of that already? I thought
> they took into account the quality of bowling faced.
Yep it does but the ICC rating has many parameters in which the
quality of bowler gets lost. For ex - ICC has let's say 20 parameters.
Out of those 4 parameters are for bowlers. Each of these 4 parameters
will then be summed up by all bowlers who played the game. So when a
McGrath and a Paul Brendon Julian bowl then the effective bowling
parameter will not be much, not even good enough, to match a situation
of W-W or Ambrose-Walsh, or McWarne, or Kumble-Bhajji, or WIPT.
> For the first pass, you could consider 10 players from the last 40
> years only (plus Bradman since he is one of the subjects in the
> hypothesis). We can always extend it later to include the older
> players if there is need/interest.
>
> Here's my suggested list:
> 0. Bradman
> 1. Lara
> 2. Tendulkar
> 3. Ponting
> 4. Kallis
> 5. Dravid
> 6. V. Richards
> 7. G. Chappell
> 8. Sobers
> 9. Gavaskar
> 10. One of Jayawardene/Sangakkara/Yousuf
How about Inzi at 10 instead of J/S/Y? Of course we could include AB/
SW but 2 Aussies are anyway represented in the top 10
Fuck, this is going to be a big time exercize hanh!
jzfredricks
> SW but 2 Aussies are anyway represented in the top 10
And there are 3 Indians.
I would suggest dropping Gavaskar, include one of J/S/Y, and then also
include a peer of Bradman.. someone like Barrington or Hammond etc.
Nirvanam
Yep we can remove Gavva and bring in Inzi. Whom do we leave out to
bring in Barrington?
Howzzat
Looks like we'll have agree to disagree on the above two points. I am
very keen to reward extremely high peaks, because that is something
very few players achieve, as opposed to say a 50+ average. Just
looking at the overall average or scores above 150 etc. will always be
unfair to Lara. I'll ask you this - you consider Lara to be equal of
Sachin and Bradman. What about Lara appeals to your really? To me a
significant portion of his greatness is attributable to the
fascination around his 375, 400 type scores and the humongous series
like the one against Murali. When I look back and think of Lara that's
the first thing that comes to mind. Why have Sachin and Kallis been
unable to achieve that despite playing 150 plus tests? By averaging it
out over a (long) career, you are allowing players like Sachin and
Kallis to catch up to him in areas where he has proven himself to be
clearly superior.
And I don't buy the highest score argument of 270* etc. Over a long
career of 100+ tests, I am sure people batting in the top order have
had opportunities to make the really big ones. One clear reason that
Lara (and Sehwag for example) has actually done it is his attacking
style, that he did not change even towards the end of his career.
Hence the likelihood of his converting a 250* into a 300+, while most
others use up too much time to get to 250* and have to stop due to a
declaration or whatever. We can't be getting into ifs and buts -
either you have the numbers on the board or you don't.
Howzzat
consideration. Hope the formatting goes through okay.
Notes:
1. I did not exclude Bangladesh yet, as it is too much work. Perhaps
I'll give it a shot later
2. I could not figure out how to get to ICC ratings point progression
for each player. It is too hard to determine % of a player's career at
ICC ratings > 800 unless we can access that progressive database. So
for now I am just looking at peak ratings achieved.
3. # of series played is 3+ test series, "big series" is the number of
3+ test series in which the player averaged 125+ runs per test
4. Twin hundreds is the number of times a player scored twin hundreds
in a test
5. "Countries avg < 45" is the number of countries in which (NOT
"against" which) the player played 5 tests/8 innings and averaged < 45
6. This is just raw data, NOT any ratings/rankings
Batsman Innings Runs 100s Average HS1, HS2, HS3 # series Big
series Twin Hundreds "Peak ICC rating points" "Countries avg < 45"
Bradman 80 6996 29 99.94 334,304,299* 11 6 1
961 0
Tendulkar 265 12970 43 54.72 248*, 241*, 217 36 4 0
898 2
Lara 232 11953 34 52.88 400*, 375, 277
26 9 1 911 2
Ponting 238 11561 38 54.79 257,242,207 33 5 3
942 2
Kallis 227 10633 33 54.80 189*,186,177 26 2 1
935 2
Dravid 237 11256 28 53.60 270,233,222 29 3 2
892 2
Richards 182 8540 24 50.23 291,232,208 26 3 0
938 1
Sobers 160 8032 26 57.78 365*,226,198 21 3 1
938 1
Chappell 151 7110 24 53.86 247*,235,204 21 2 2
883 1
J'wardene 182 9120 27 53.96 374,275,242 23 2 0
883 3
Barrington 131 6806 20 58.67 256,172,163 17 1
0 914 1
Hammond 140 7249 22 58.45 336*,251,240 20 2 1
897 0
Gavaskar 214 10122 34 51.12 236*,221,220 29 3 3
916 2
Howzzat
results. I spent a lot of time tinkering with criteria and weights,
and the following represents the outcome of my (so far) final
iteration. Some key points:
1. These are the relative ranks of only the 13 players analyzed. Many
other worthies have not been considered in this pass, and some of them
could easily be considered better players than the ones analyzed here.
2. Unfortunately only the measurable, quantitative criteria could be
considered. So this leaves out that X-factor associated with someone
like Richards, and he suffers on that account. Perhaps some other
time.
3. Methodology: In the end I considered 8 criteria (more on this
below), and assigned points out of 10 for each batsmen except Bradman.
For each criterion, I started by allocating 10 points to the best
performing batsman in the list, and calculated other players' points
relative to that score (while normalizing where appropriate). I then
decided to scale up Bradman (as opposed to assigning him 10 points and
scaling everyone else down) so for most criteria Bradman scored more
than the maximum of 10 points
4. In addition to the criteria I listed earlier, I added longevity to
the mix, to reward someone like Tendulkar for continued excellence
over a long (measured by innings played) career. This was an
afterthought as almost everything else is normalized, so I wanted to
reward players for maintaining that level of performance over more
matches.
Here are the criteria and the weights I used (totaling to 100%). Of
course this is subject to debate but after much tinkering I settled on
this.
C1. Frequency of scoring hundreds - 25%
C2. Career batting average relative to peers - 25%
C3. Magnitude of top 3 innings - 10%
C4. Monster series (125+ per test runs in a 3+ test series) - 10%
C5. Twin hundreds in a test - 3%
C6. Highest ICC rating points achieved - 10%
C7. 45+ average in all countries played (8+ innings min) - 10%
C8. Longevity - 7%
Leaving aside Bradman who is off the charts in most categories, I was
surprised by how often Lara has put together monster series relative
to other players on the list. He has 9 such series out of 26 played,
which is more than twice as frequent as any other player not named
Bradman. Also, Ponting and Gavaskar are the only players in this list
to have scored 2 hundreds in a test 3 times, while Tendulkar,
Richards, Jayawardene and Barrington did not manage the feat even
once.
Finally here are the rankings. They are shown in the following format,
with unweighted points out of 10 for each of the eight criteria in
parenthesis. Total points are out of 100, and as I mentioned earlier,
Bradman scores more than 100 as I scaled him up rather than scaling
others down.
Rank "Total Points" "Player Name"(C1, C2, C3, C4, C5, C6, C7, C8)
1. 148 Bradman (22.3, 16.8, 8.9, 15.8, 8.9, 10.4, 10.0, 3.0)
2. 91 Lara (9.0, 9.3, 10.0, 10.0, 3.1, 9.3, 9.0, 8.8)
3. 88 Ponting (9.8, 9.6, 6.7, 4.4, 9.0, 10.0, 9.0, 9.0)
4. 86 Sobers (10.0, 9.9, 7.5, 4.1, 4.5, 9.9, 9.0, 6.0)
5. 84 Tendulkar (10.0, 9.6, 6.7, 3.2, 0.0, 9.1, 9.0, 10.0)
6. 84 Gavaskar (9.8, 9.2, 6.4, 3.0, 10.0, 9.4, 8.5, 8.1)
7. 83 Hammond (9.7, 9.6, 7.9, 2.9, 5.1, 9.0, 10.0, 5.3)
8. 83 Chappell (9.8, 9.7, 6.5, 2.8, 9.4, 8.7, 9.0, 5.7)
9. 80 Kallis (8.9, 9.6, 5.2, 2.2, 3.1, 9.9, 9.0, 8.6)
10. 78 Jayawardene (9.1, 9.4, 8.5, 2.5, 0.0, 8.7, 7.0, 6.9)
11. 78 Barrington (9.4, 10.0, 5.6, 1.7, 0.0, 9.4, 9.0, 4.9)
12. 78 Dravid (7.3, 9.4, 6.9, 3.0, 6.0, 8.9, 9.0, 8.9)
13. 77 Richards (8.1, 9.1, 6.9, 3.3, 0.0, 9.9, 9.0, 6.9)
There it is folks. I am a little surprised by the low ratings of
Dravid and (especially) Richards, but that's what you get if you go
strictly by numbers.
Comments welcome, and I eagerly await Nirvanam's more detailed
analysis.
jzfredricks
> All right folks, for those of you still following along, here are my
> results. I spent a lot of time tinkering with criteria and weights,
> and the following represents the outcome of my (so far) final
> iteration. Some key points:
I like what you're doing. The real debate is over the weighting for
criteria.
Could you please give all criteria an equal rating and post the
results? Just so we can see?
Howzzat
Sure, but realize that this will overly punish those who did not score
twin hundreds in a test. That is the only criteria in which the point
allocations ranged from 0.0 to 10.0 - that is one of the reasons I
assigned it a weight of only 3%. But with equal weights for everything
we get:
1. 120 Bradman
2. 86 Lara
3. 84 Ponting
4. 81 Gavaskar
5. 77 Chappell
6. 76 Sobers
7. 74 Hammond
8. 74 Dravid
9. 72 Tendulkar
10. 71 Kallis
11. 67 Richards
12. 65 Jayawardene
13. 63 Barrington
By the way, if you want to play around with the weights, you can do
that because I gave you all the unweighted points in my previous post.
For this equal weighting exercise I just added up unweighted points
for all eight criteria (max = 80 points), then scaled to calculate
scores out of 100.
Nirvanam
Hey this is well done, Howzatt. I'll add the difficult conditions
thing to it when doing the details. I'd still suggest you use the
rating points above 800 criteria instead of peak rating.
Of course I'll twist the criteria a little when I'll do my
analysis...like the monster innings and all.
But its a damn neat one. Basically what it is saying is that all these
fellas except Don of course are in a kinda clutter. Depending on the
angle of vision we'll see a few ahead of the others inside the
clutter...just 14 points between 2 and 13. Nice one though.
skp
Man, I really don't know how you guys have the time/inclination/brain
power to do all this stuff. Good effort.
My chief comment (and some of this may be addressed in earlier posts)
is that C3,C4, and C5 seem to put too much of a premium on Hot
streaks, and snapshots in time.
To take an example, if one were to compare just Lara, SRT, and
Ponting, most people would say even intuitively that Lara at his best
was the best. His graph has more highs and lows as compared to the
other 2. In your scoring system, his highs have been specifically
rewarded but his lows haven't been specifically penalized. If he had
9 monster series where he averaged at least 62.5 per innings, then he
necessarily must have been quite pedestrian in the other 17 series for
his overall average to be 52'ish.
IMO, there ought to be a consistency barometer specifically thrown
into this somewhere. Something like percentage of series where the
batsman in question averaged less than 80 runs per test. 80 is just a
number off the top of my head. These would be negative points that
are then factored into the equation.
Also, not sure that I agree with just taking into account each
player's top 3 innings. Most if not all of the modern day greats have
all likely played over 200 innings each. Top 10% of innings might be
a better barometer if one chooses to go down this path at all.
The twin hundreds is another strange one. In addition to his 3 twin
hundreds, Poniting in fact also has a 101 and a 99 in a test. If he
scored one extra run in the second innings of that test, am I correct
in assuming that your points scoring system would then reduce the
points scored by all other players? (well, not Richards, SRT, Jaya, or
Barrington as they already have a zero).
Plus other snapshots or hot streaks go unrewarded. For example,
Ponting may be one of the few on this list to get double hundreds in
back to back tests. No reward for that (well, I suppose that there is
an indirect reward in the monster series category but there is no
specific reward for the back to back doubles). As another example,
SRT got 495 runs in 3 consecutive innings without being dismissed.
AFAIK, only SRT and Lara on this list have scored more than 400 runs
in consecutive innings without being dismissed. Again, no points for
that. IIRC, Dravid has scored centuries in 3 consecutive tests but
not all of it in the same series. No reward for that.)
BTW, intuitively, it strikes me that perhaps Dravid comes out so low
because of a much higher incidence of getting out in the nervous
nineties? i.e. this hurts him on the first criterion which is weighted
at 25%.
One last point:
I still haven't studied Anant's analysis very carefully. But what is
it about his analysis that you most disagree with?
Howzzat
I wanted to do the 800 thing, but couldn't figure out how to get the
data. The charts on ICC's web site are static and not much help. If
you have access to the data let me know and I will incorporate.
>
> But its a damn neat one. Basically what it is saying is that all these
> fellas except Don of course are in a kinda clutter. Depending on the
> angle of vision we'll see a few ahead of the others inside the
> clutter...just 14 points between 2 and 13. Nice one though.
Yes, and I think that is how most people see it. The only universal
truth in test match batting ratings is that Don is on a level of his
own. Then there are about 10-15 others any of whom can be considered
first among equals, depending on your point of view. Having said that,
I truly believe that Tendulkar was the clear second on Don until about
2001/2002 - before things went awry and he fell back to the rest of
the pack.
jzfredricks
> fellas except Don of course are in a kinda clutter.
Closer to agreeing that Bradman is better than SRT?
Howzzat
>
> To take an example, if one were to compare just Lara, SRT, and
> Ponting, most people would say even intuitively that Lara at his best
> was the best. His graph has more highs and lows as compared to the
> other 2. In your scoring system, his highs have been specifically
> rewarded but his lows haven't been specifically penalized. If he had
> 9 monster series where he averaged at least 62.5 per innings, then he
> necessarily must have been quite pedestrian in the other 17 series for
> his overall average to be 52'ish.
>
> IMO, there ought to be a consistency barometer specifically thrown
> into this somewhere. Something like percentage of series where the
> batsman in question averaged less than 80 runs per test. 80 is just a
> number off the top of my head. These would be negative points that
> are then factored into the equation.
I have actually been thinking along the same lines. I am going to add
another criterion that penalizes players on having poor 3+ test
series. I think 50 runs per test is a good cutoff. Most batsmen
typically average around 80 runs per test, so anything less that 50
should be considered poor.
>
> Also, not sure that I agree with just taking into account each
> player's top 3 innings. Most if not all of the modern day greats have
> all likely played over 200 innings each. Top 10% of innings might be
> a better barometer if one chooses to go down this path at all.
If you do the top 10%, you are getting into the 150ish type scores.
Ponting and Tendulkar have 250+ innings and 4-5 double hundreds. Their
next 20 innings will likely be scores between 140-180 ish, while for
other players who score hundreds less frequently (e.g. Dravid) it will
be in the low 100s. We already have a centuries scored measure, so I
think what you are suggesting is well accounted for there.
The top 3 innings rewards something that very few people managed to
achieve (as opposed to say a double hundred). And to make sure that
someone didn't have just one huge knock as a fluke, I am averaging
their top 3 scores. And this criterion only has a 10% weight vs. the
centuries one that has 25%.
>
> The twin hundreds is another strange one. In addition to his 3 twin
> hundreds, Poniting in fact also has a 101 and a 99 in a test. If he
> scored one extra run in the second innings of that test, am I correct
> in assuming that your points scoring system would then reduce the
> points scored by all other players? (well, not Richards, SRT, Jaya, or
> Barrington as they already have a zero).
Yes, but that is the nature of our game. In the centuries scored
measure too, there are many 95+ scores that were not counted. I see
what you are saying, but whatever cutoff you apply there will be
scores just below that.
Twin hundreds might appear a strange one (and I don't blame you), but
it is something I have always been fascinated with. Over a long enough
career (100+ tests), I would expect ATG players to have a few
opportunities to accomplish it. And anyway, I did normalize the points
here by number of matches played, and weighted it very low (3%), so it
does not impact the results materially. For instance, the top 5 stay
the same, although the points difference is a little less.
>
> Plus other snapshots or hot streaks go unrewarded. For example,
> Ponting may be one of the few on this list to get double hundreds in
> back to back tests. No reward for that (well, I suppose that there is
> an indirect reward in the monster series category but there is no
> specific reward for the back to back doubles). As another example,
> SRT got 495 runs in 3 consecutive innings without being dismissed.
> AFAIK, only SRT and Lara on this list have scored more than 400 runs
> in consecutive innings without being dismissed. Again, no points for
> that. IIRC, Dravid has scored centuries in 3 consecutive tests but
> not all of it in the same series. No reward for that.)
Well yeah but how do you account for everything? I am going by the
measures that are universally accepted as great batting feats. The not
outs etc. you mention will show up in the overall average measure,
which I weighted as high as 25%. By the way, after those 495
undefeated runs (3 innings), Tendulkar had the following string of 5
scores: 2, 8, 1, 8, 2, 5. Even before the 241* and 60* in Sydney that
started that not out streak, Tendulkar had hardly scored a run all
series. I am not penalizing him for those either. Overall average
takes care of both of these.
Ideally in my analysis I would also include strike rates and ATG
innings (that would help someone like Viv Richards), but unfortunately
the strike rates are not available for all players and ATG lists are
too subjective.
>
> BTW, intuitively, it strikes me that perhaps Dravid comes out so low
> because of a much higher incidence of getting out in the nervous
> nineties? i.e. this hurts him on the first criterion which is weighted
> at 25%.
That's part of it - he has the lowest centuries per innings ratio.
Also some other downers such as relative lack of big series, ICC peak
rating etc. I am the first one to claim that numbers don't tell the
whole story - Dravid and Richards are perfect examples. Unfortunately,
for this debate we are constrained to using only these measurable
criteria. I wouldn't have too many issues with anyone claiming Dravid
or Richards as the best test batsmen behind Bradman, as long as they
put massive daylight between them and Bradman. You believe what you
believe.
>
> One last point:
>
> I still haven't studied Anant's analysis very carefully. But what is
> it about his analysis that you most disagree with?
His massive overweighting of the aggregate. In the example I gave
earlier, Viv Richards has better average and strike rate than
Tendulkar in ODIs, yet is ranked a fair way behind the latter. If you
read Anant's responses to some of the comments made on his article, he
acknowledges that. Based on his methodology, if Tendulkar continued to
add to his aggregate over the next 10 years while performing
substantially below his peers in averages, hundreds, whatever else, he
would actually increase his lead over everyone else, as long as he
continued to have the highest aggregate. I don't know how anyone in
their right mind could agree with that.
skp
> > Plus other snapshots or hot streaks go unrewarded. For example,
> > Ponting may be one of the few on this list to get double hundreds in
> > back to back tests. No reward for that (well, I suppose that there is
> > an indirect reward in the monster series category but there is no
> > specific reward for the back to back doubles). As another example,
> > SRT got 495 runs in 3 consecutive innings without being dismissed.
> > AFAIK, only SRT and Lara on this list have scored more than 400 runs
> > in consecutive innings without being dismissed. Again, no points for
> > that. IIRC, Dravid has scored centuries in 3 consecutive tests but
> > not all of it in the same series. No reward for that.)
>
> Well yeah but how do you account for everything? I am going by the
> measures that are universally accepted as great batting feats. The not
> outs etc. you mention will show up in the overall average measure,
> which I weighted as high as 25%. By the way, after those 495
> undefeated runs (3 innings), Tendulkar had the following string of 5
> scores: 2, 8, 1, 8, 2, 5. Even before the 241* and 60* in Sydney that
> started that not out streak, Tendulkar had hardly scored a run all
> series. I am not penalizing him for those either. Overall average
> takes care of both of these.
>
great stretch exmaple, he is not rewarded for the 3 innings and not
penalized for the string of failures preceding and following those 3
innings. But under the monster series category, a player could have 2
or 3 poor/indifferent series, have a monster series, and then have 2
or 3 poor/indifferent seies and yet be rewarded.
That said I appreciate that you may be ameliorating/rectifying this
with the added "consistency criterion" that we have already addressed.
> > One last point:
>
> > I still haven't studied Anant's analysis very carefully. But what is
> > it about his analysis that you most disagree with?
>
> His massive overweighting of the aggregate. In the example I gave
> earlier, Viv Richards has better average and strike rate than
> Tendulkar in ODIs, yet is ranked a fair way behind the latter. If you
> read Anant's responses to some of the comments made on his article, he
> acknowledges that. Based on his methodology, if Tendulkar continued to
> add to his aggregate over the next 10 years while performing
> substantially below his peers in averages, hundreds, whatever else, he
> would actually increase his lead over everyone else, as long as he
> continued to have the highest aggregate. I don't know how anyone in
> their right mind could agree with that.
Well yeah, I would agree with you there (heh...I am still in my right
mind....)
Hopefully, I will get a chance to read his analysis carefully sometime
soon.
BTW, my thanks to you and Nirvanam for doing this stuff.
ciao
skp
Nirvanam
Look, when it comes to numbers no one can argue against Don. But its
the subjective aspects and specificity of numbers in difficult
conditions that are the basis for my opinion that Sachin, Don, and
Lara are in the same league...virtuosos.
jzfredricks
> On Jan 12, 1:59 am, jzfredricks <jzfredri...@gmail.com> wrote:
>
> > > But its a damn neat one. Basically what it is saying is that all these
> > > fellas except Don of course are in a kinda clutter.
>
> > Closer to agreeing that Bradman is better than SRT?
>
> Look, when it comes to numbers no one can argue against Don.
I'll take that as a yes.
higgs
The results that we currently have appear to indicate Don, followed by
a bunch, the leader of that bunch being Lara. TAV is back in the
middle of the pack.
I suppose you could argue that all those in the pack are in the same
league as Bradman, but I reckon you'd have problems convincing people
that there are upwards of 12 batsmen in the same league as Bradman.
You might have an easier time convincing people that TAV is in the
same league as about 11 other batsmen.
Failing that, you could see which criteria TAV comes out best in,
assign a higher rating to those criteria and then make an arbitrary
cut-off point directly below this point and decare any player above
this line to be the equal of Bradman (virtuoso). Good luck.
Higgs
Nirvanam
You don't need to take it as a yes. It is a yes. It never was a "no".
I don't see where the confusion is?
Nirvanam
Sure Higgs these results based on the criteria chosen show different
things. Pls understand Howzatt himself has done a high level analysis.
Let's wait till the thorough analysis to be done. Howzatt and I will
exchange notes on that as we progress. I am waiting for the last Test
between Pak and Aus to end to have the latest data.
higgs
>
> - Show quoted text -
Like I said, the results we have show TAV back in the middle of the
pack behind Bradman.
If you tweak the weightings and assign more to those areas where TAV
is strongest, you should at least be able to put him at the front of
the pack, though I doubt you'll get him equal with Bradman (virtuoso)
unless you do an extremely high level analysis and assign something
like 90% of the marks for aggregate runs.
I've already wished you luck, what more do you want me to do?
Higgs
Nirvanam
Higgs, what are you not understanding of my posts? Why are you
insisting on misinterpreting and misrepresenting my opinion. Have I
not already told this a thousand times that Don cannot be compared as
far as numbers are concerned?
jzfredricks
You have on numerous times refused to say "Bradman was better than
SRT" (or to that effect). Instead, you do your best to lump them
together in a single quality category and say "they are both
brilliant".
Nirvanam
Yes he is not better than SRT when viewed from all perspectives...that
is my opinion. If viewed only from numbers perspective then he has no
comparison, but given a holistic view my opinion is that Sachin, Don,
and Lara are a class of their own which no other batsmen in the
history of the game have risen to.
higgs
>
> - Show quoted text -
I'd prefer to look at subjective aspects and specificity of numbers in
difficult
conditions to form the basis of my opinion.
It then becomes clear that Don leads the way, followed by Lara.
They are in a class of their own (virtuosos) which no other batsman in
the history of the game has risen to.
Higgs
higgs
>
> - Show quoted text -
What's to misinterpret?
You say that Don, Lara & SRT are equals and you said that a high level
analysis of the figures would prove this.
The first round of figures haven't proved this.
I'm simply waiting for the next set of figures
Higgs
R. Spanditt
> yawn
>
> discussed at length not too long ago
some of us don't spend our lives on here.
Howzzat
Nirvanam - as I have mentioned numerous times I respect your opinion.
I think the problem arises only when you insist that you can prove the
above. I really don't understand how you can do so conclusively,
unless you massively overweight longevity and aggregate (which is the
basis of Anant's analysis as well). On pretty much every normalized
criteria, Don is so much off the charts it beggars belief.
If you want to bring stuff like "virtuoso", "never got a traffic
ticket while driving to practice" etc. into it, you'd have to find a
way to convert that into measurable data. Unless you can prove your
assertion, people will continue to knock you back.
R. Spanditt
> I am not suggesting that Tendulkar shouldn't be near the top. He is
> probably #1 or #2 in ODIs, but among the names considered, would
> barely make my top 10 in tests. Loads of batsmen have played more
> significant innings in tests than Tendulkar.
Found your comments very interesting because i love lists..
I'd love to see your top ten if you have the time to compile it? I'm
only 27 so I don't feel qualified to be able to refer to greats such
as gavaskar, chappell, sobers, bradman and the like who would
obviously feature on many all-time lists.
R. Spanditt
> Please differentiate between a Sachin fanatic claiming things that
> Sachin has not done and a person who is sharing his glee in an
> objective analysis of all batsmen who played tests and odi's. If you
> are going to view each one of my posts on Sachin as some kinda a
> 'clueless' sachin-fanatic's rant then you are neither giving the point
> I am making due consideration nor are you allowing a fair open-minded
> assessment of my viewpoints- Hide quoted text -
I like the way you have put forward your point, but in all fairness
you referred to Sachin as 'he', assuming people would know who you're
talking about. Certainly a little bit Sachin-centric, was your post,
impossible to argue otherwise. Whether or not the weightings favoured
Sachin i'm not sure but I would have assumed he'd be #1 myself anyway.
Howzzat
As I've mention earlier, after Bradman it just boils down to personal
preference. I have already presented the list based on measurable
criteria that *I* consider as important (and even that list was not
complete - we just restricted ourselves to 13 players), but I would
not pick my top 10 based on those numbers alone. One of my key
criteria would be the aura the batsman projects at the crease - almost
the way he makes the bowlers bowl at him. Sachin had that in loads
until the injuries arrived in 2003, and on this measure he has been
very very ordinary since then (IMHO of course). I also appreciate
batsmen who have the ability to dig in and play a 60* type knock over
several hours to save a game - not saying Collingwood would make my
top 10 but I do appreciate the Dravid and Kallis type players. With
Sachin I never really had that confidence, even in the 90s.
Not sure if you were looking for all-time, but in the last 40ish
years, considering their overall careers I would certainly pick
Sobers, Lara, Richards, Gavaskar and Sehwag over Tendulkar in my test
team. Next in line would be the group consisting of Tendulkar, Kallis,
Dravid and Ponting. Just to be sure, this is for Tests only. If
combining with ODIs, Tendulkar is probably my second pick after
Richards.
eusebius
Also don't forget Grant Bradburn..
I agree just from a numbers point of view he doesn't match the others
but taking the wholistic view, considering all variables,
possibilities and probabilities then SRT, Lara, Don, W.G.,Trumper,
Albert Trott, Andy Ganteaume, Lord Frederick Beauclerc, A.E.J.
Collins, Charles Bannerman, Vinod Kambli, Stuart Law and He are a
class of their own which no other batsman in the history of the game
have risen to.-
eusebius
>
> - Show quoted text -
And sorry I negelcted to mention dewpoint. That is another very vital
statistic.
skp
>
> - Show quoted text -
I will have to get to Anant's analysis at some other point when I have
time. But I did have a look ar Viv's ODI stats and SRT's ODI stats.
IMO, SRT's numbers are more impressive. I choose those words
deliberately. It doesn't mean that SRT was a better ODI player.
That's a different argument and requires a look at some different
criteria. But on the numbers alone, that aggregate does makes SRT's
numbers more impressive.
VIV
167 innings
24 Not outs
6721 Runs
189* HS
47 average
90'ish sr
11 centuries
45 50s
SRT
429 innnings
40 Not out
17,394 runs
186 HS
44.71 Average
86'ish SR
43 centuries
54 50s
Both have terrific stats but obviously, terrific stats over 429
innings carries more weight than terrific stats over 167 innings. One
can argue about how much weight but IMO, A 10K difference in aggregate
runs more than makes up for the lower average (note the ratio of "not
outs" as well) and lower SR.
Playing devil's advocate, I would point out this though: VIV had a 16
year ODI career. They simply didn't play as many ODIs back then as
they did in SRT's era. The fact that Richards amassed his stats over
16 years renders the aggregate argument in favor of SRT less
persuasive.
To put it another way, if Richards had amassed his stats in say only 8
years, I could more convincingly argue that he may have had a tough
time keeping those stats at that high level over a 20 year career like
SRT has done. Stuff like age/slower reflexes, complaceny, boredom
coule be expected to kick in and negatively affect performance. But
the fact is that VIV did what he did it over 16 years. So, it's
easier to argue that he may well have maintained stats at that high
level even if he had played 400 innings in those same 16 years as
opposed to the 167 that he actually paid.
Howzzat
>
> To put it another way, if Richards had amassed his stats in say only 8
> years, I could more convincingly argue that he may have had a tough
> time keeping those stats at that high level over a 20 year career like
> SRT has done. Stuff like age/slower reflexes, complaceny, boredom
> coule be expected to kick in and negatively affect performance. But
> the fact is that VIV did what he did it over 16 years. So, it's
> easier to argue that he may well have maintained stats at that high
> level even if he had played 400 innings in those same 16 years as
> opposed to the 167 that he actually paid.
Great - you answered your own question!
To put Viv's strike rate in perspective, in the 70s/80s a 240 ish
score was considered a winning total. I haven't looked it up, but I am
sure most top order players at that time has a SR in the low 60s,
which makes Viv's numbers all the more remarkable.
You really can't compare people across generations because the game
keep on changing/evolving. That's why the only meaningful comparisons
are the ones against contemporaries. That is why Don is considered
untouchable - none of his contemporaries came within touching
distance. Viv was clearly the best ODI player of his era, and SRT is
probably the best of his.