Reforming Fantasy Scout (yearly rankings)
milivella
reform... ;) But before officially proposing it, I'd like to have as
much feedback as possible, so please reply with your opinions and
criticisms.
1. Each year, a scout can pick up to 10 players.
(We can make it 12 as the number of the months, or any other number.)
2. There is a separate scout ranking for each year of pick. The score
of a scout is the sum of the points earned by the players he picked in
that year.
(So there will be a ranking for players picked in 2010, a different
one for players picked in 2011, and so on.)
(Going for the sum instead of the average makes things straighter,
because it's easier to compare the performance of two scouts: "I have
4 starters", "But I have 6, 2 of which are goal-machines: I win!";
with the current system, you have to divide by the number of picked
players.)
(One nice feature of yearly rankings is that we'll be able to change
rules - about picking players or scoring - without hurting what has
happened so far.)
(Of course the ranking for the 2007-09, or for 2007-08 and 2009, will
be based on average, because it would be unfair to change the past.)
3. Then there is a meta-ranking based on the rankings of each year.
This meta-ranking is based on: most 1st places; as a tie-breaker,
most 2nd places; as a tie-breaker, most 3rd places; and so on.
(So one 1st place beats four 2nd places.)
(This way, even if constancy is rewarded, you can enter the game late
and overcome the early scouts.)
To you.
--
Cheers
milivella
Benny
> From : mili...@gmail.com
> 1. Each year, a scout can pick up to 10 players.
>
> (We can make it 12 as the number of the months, or any other number.)
I'm fine either way but still want the option to drop players, at least
3 a year.
--
http://soccer-europe.com
Rss feed : http://soccer-europe.com/RSS/News.xml
milivella
> > Subject : Reforming Fantasy Scout (yearly rankings)
> > From : milive...@gmail.com
>
> > 1. Each year, a scout can pick up to 10 players.
> >
> > (We can make it 12 as the number of the months, or any other number.)
>
> I'm fine either way but still want the option to drop players, at least
> 3 a year.
I see why you want to drop players now that scout score is an
*average* (in fact, a 0 point player takes down your average), but in
this proposal scout score is a *sum*, so you have no advantage
dropping players (in fact, having or not having a 0 pointer is the
same).
Or do you mean something like "in the first months of 2010 I pick 10
players, then in September I drop 3 of them and pick 3 new players
instead"?
(Thanks for the feedback!)
--
Cheers
milivella
milivella
In the latter case, it would be the same to allow scouts pick 13
players (instead of pick 10, then drop 3 and pick another 3).
--
Cheers
milivella
milivella
> 2. There is a separate scout ranking for each year of pick. The score
> of a scout is the sum of the points earned by the players he picked in
> that year.
> (Going for the sum instead of the average makes things straighter,
> because it's easier to compare the performance of two scouts: "I have
> 4 starters", "But I have 6, 2 of which are goal-machines: I win!";
> with the current system, you have to divide by the number of picked
> players.)
What I don't like in my proposal is that the scoring system is less
complex on one hand (because all the scouts would have the same number
of players), but more complex on the other hand (because there would
be many rankings, one for year).
An alternative could be something like this (leaving the yearly
rankings). Each year, a scout can:
- pick up to 10 players
- select 1 out of the players picked the previous year to permanently
represent him
So things would be simpler: for each year, I would have one player, to
compete against one Jesus' player, one Alessandro's player, etc.
--
Cheers
milivella
Futbolmetrix
news:2e1c921a-5a78-4c1b...@y9g2000yqg.googlegroups.com...
> It's that time of the year again: I'm itching for a Fantasy Scout
> reform... ;) But before officially proposing it, I'd like to have as
> much feedback as possible, so please reply with your opinions and
> criticisms.
I think the lesson that we learned from last year's lengthy discussion is
that it's almost impossible to get everybody on board with any suggested
change. And rightly so. The competition started under one set of rules,
contestants chose their strategy under those rules, and it's very difficult
to change the rules in such a way that you don't penalize anybody for his
past choices.
So we can try to think of various alternative or intermediate "prizes" (just
as in real football, you can rank teams by the number of goals scored,
number of goals conceded, who won the most number of corners, etc.; but in
the end the only thing that matters is the total number of points).
Having said this, maybe it's time to launch a completely new game,
independent of Fantasy Scout: Fantasy Scout Champions' League
Rules:
Pick any number of players, from any nationality, under the age of 21, who
have never appeared in a CL match. (Age = current year - year of birth).
Points = Appearances + goals in the Champions' League proper - 10 points for
every player picked.
D
milivella
Thanks for the feedback. :)
> I think the lesson that we learned from last year's lengthy discussion is
> that it's almost impossible to get everybody on board with any suggested
> change.
You're right. I like trying nearly impossible tasks, though. ;) And
anyway I'd like to know what the feelings about the games are (see my
next post).
I know that it's not everybody's feeling, but I enjoyed last year's
discussion, and I think that its little output (you now have to pick
at least 2 players each year) has fixed the major "bugs" in the game.
> And rightly so. The competition started under one set of rules,
> contestants chose their strategy under those rules, and it's very difficult
> to change the rules in such a way that you don't penalize anybody for his
> past choices.
I agree, but I see some changes that should be OK from this point of
view:
- Switching to a yearly ranking (my proposal in the first post of this
thread) doesn't harm anyone (does it?).
- Switching to a sum-with-penalty system (your proposal last year), if
the penalty is proportional to the number of scouts in the moment of
the pick doesn't harm either.
- I'm still not sure whether using the H-index as score would really
penalize - in the long run - who picked few players (Michael has
argued so, but maybe he should be more patient and write a little
more).
> So we can try to think of various alternative or intermediate "prizes"
Nice idea. I'll think about it.
> (just
> as in real football, you can rank teams by the number of goals scored,
> number of goals conceded, who won the most number of corners, etc.; but in
> the end the only thing that matters is the total number of points).
Do you mean that the intermediate prizes would be - in the end -
meaningless?
> Having said this, maybe it's time to launch a completely new game,
> independent of Fantasy Scout: Fantasy Scout Champions' League
>
> Rules:
> Pick any number of players, from any nationality, under the age of 21, who
> have never appeared in a CL match. (Age = current year - year of birth).
> Points = Appearances + goals in the Champions' League proper - 10 points for
> every player picked.
Nice. I've had a similar idea some time ago, but your rules (above all
the age limit) works better than mine (as expected!). (My advisor, who
doesn't play FS, and - if I'm not wrong - Alberto came to similar
ideas as well.)
If you're interested in my humble opinion, here are the only points
*against* the game that I can imagine (i.e. all the rest is perfect!):
- Do all the world class players play in Europe (and so in sides that
can reach UCL)? I'm not sure about it.
- Many of the things that you maybe would like to be fixed in FS are
still there in FSCL under the rules you have posted (the nice things
as well: it's simple, you aren't requested a constant effort).
- I don't like to have the same player on my side in FS and against me
in FSCL (for all his career!), but this is probably just a
ideosincracy of mine. Anyway, we could give the option (only at the
beginning of the game) to carry with you in FSCL players already
picked in FS.
--
Cheers
milivella
milivella
> I'd like to know what the feelings about the games are (see my
> next post).
Here it is. Three questions to you all:
1. Do you care about *immediately* (e.g. in the same moment as you
know who played and scored) knowing the effect of international
matches on your Fantasy Scout score?
2. Are you able to immediately know the aforementioned effect? (e.g.
to say who between Benny and Alberto has had a better June so far?)
3. What is the thing in Fantasy Scout that annoy you the most?
Thanks.
--
Cheers
milivella
milivella
Futbolmetrix:
> > The competition started under one set of rules,
> > contestants chose their strategy under those rules, and it's very difficult
> > to change the rules in such a way that you don't penalize anybody for his
> > past choices.
>
> I agree, but I see some changes that should be OK from this point of
> view:
I still keep this assertion, but my subsequent proofs were pretty
silly. :(
> - Switching to a yearly ranking (my proposal in the first post of this
> thread) doesn't harm anyone (does it?).
Yes, it could harm. Even in the simplest example (2 scouts, 2 years, 1
player picked by each scout per year) things could change with the
proposed rules:
Year 1: A 10 points, B 1 point
Year 2: A 1 point, B 2 points
Under the actual rules: A 5.5 points, B 1.5 points. A wins.
Under the proposed rules: A wins year 1, B wins year 2. It's a tie.
So this proposal doesn't pass my own "razor"
http://groups.google.com/group/rec.sport.soccer/msg/c85c605062fe3723
and I retract it. Sorry. :(
> - Switching to a sum-with-penalty system (your proposal last year), if
> the penalty is proportional to the number of scouts in the moment of
> the pick doesn't harm either.
?!
Anyway, as we said last year, it's possible to switch from average to
sum, if you weight players picked under the average system (the more
the players, the less their individual weight). But it's not elegant,
and it was not accepted by everyone.
> - I'm still not sure whether using the H-index as score would really
> penalize - in the long run - who picked few players (Michael has
> argued so, but maybe he should be more patient and write a little
> more).
It doesn't pass the "razor" either.
(Of course not passing the razor doesn't automatically mean that a
proposal is unfair: indeed, I think that in the aforementioned cases
probably no scout should feel penalized, so these changes could be
proposed and approved.)
--
Cheers
milivella
milivella
> Having said this, maybe it's time to launch a completely new game,
> independent of Fantasy Scout: Fantasy Scout Champions' League
>
> Rules:
> Pick any number of players, from any nationality, under the age of 21, who
> have never appeared in a CL match. (Age = current year - year of birth).
> Points = Appearances + goals in the Champions' League proper - 10 points for
> every player picked.
I know that I'm spoiling your fun ;) but here is an incomplete ranking
of players (no penalty applied, no Champions Cup, updated to 2008-09
semifinals), thanks to
http://www.uefa.com/printoutfiles/competitions/ucl/2009/e/e_13_md.pdf
(page 3 = 5 of 45)
http://euro.futbal.org/EC1.scorers.php
187 123+64 Raúl González (Real Madrid CF)
154 104+50 Thierry Henry (AS Monaco FC, Arsenal FC, FC Barcelona)
141 94+47 Andriy Shevchenko (FC Dynamo Kyiv, AC Milan, Chelsea FC)
139 114+25 Ryan Giggs (Manchester United FC)
136 120+16 Roberto Carlos (Real Madrid CF, Fenerbahçe SK)
127 103+24 Luís Figo (FC Barcelona, Real Madrid CF, FC Internazionale
Milano)
128 107+21 Paul Scholes (Manchester United FC)
119 103+16 David Beckham (Manchester United FC, Real Madrid CF)
111 109+ 2 Paolo Maldini (AC Milan)
111 101+10 Clarence Seedorf (AFC Ajax, Real Madrid CF, AC Milan)
109 94+15 Guti (Real Madrid CF)
105 103+ 2 Gary Neville (Manchester United FC)
103 103+ 0 Oliver Kahn (FC Bayern München)
97 96+ 1 Claude Makelele (FC Nantes Atlantique, Real Madrid CF,
Chelsea FC)
56+ ?+56 Ruud van Nistelrooy (PSV Eindhoven, Manchester United FC,
Real Madrid CF)
42+ ?+42 Filippo Inzaghi (Juventus, AC Milan)
41+ ?+41 Alessandro Del Piero (Juventus)
Notable absences: Ronaldo (he should have 54 points) and Zidane (93?).
--
Cheers
milivella
milivella
> Notable absences: Ronaldo (he should have 54 points)
58?
http://www.footballdatabase.eu/football.joueurs.luis-nazario.ronaldo.424.en.html
Anyway, you got the idea.
--
Cheers
milivella
milivella
> I know that I'm spoiling your fun ;) but here is an incomplete ranking
> of players (no penalty applied, no Champions Cup, updated to 2008-09
> semifinals)
Let's make it in the right way. The 100 players with most appearances,
sorted by Fantasy Scout score:
http://spreadsheets.google.com/pub?key=rEzshUHG69hkYEbXWK8URTw&output=html
Excel version:
http://spreadsheets.google.com/pub?key=rEzshUHG69hkYEbXWK8URTw&output=xls
It should be correct (at least starting 1993-94). The source is:
http://www.footballdatabase.eu/clubparticipe.php?compet=(C1)%20Ligue%20des%20Champions&lieu=Europe&cltype=9&sens=down&journee=-50&page=1
If you are lazy, here are the basic data (name, score):
Gonzalez Raul 187
Thierry Henry 154
Ryan Giggs 144
A. Shevchenko 140
Roberto Carlos 135
Paul Scholes 129
Luis Figo 128
A. Del Piero 128
R. Van Nistelrooy 128
F. Morientes 122
David Beckham 119
Filippo Inzaghi 113
Paolo Maldini 112
C. Seedorf 111
José Maria Guti 109
Gary Neville 106
Oliver Kahn 103
Ivan Helguera 102
Rivaldo 100
P. Kluivert 100
Claude Makelele 97
H. Salihamidzic 97
O. G. Solskjaer 97
Robert Pires 96
Michael Ballack 96
Zinedine Zidane 94
Pavel Nedved 94
Giovane Elber 94
Iker Casillas 93
Hernandez Xavi 92
Sylvain Wiltord 91
Hernan Crespo 90
Roy Makaay 90
Frank de Boer 89
Roy Keane 89
Carles Puyol 87
Philip Cocu 87
C. Panucci 86
Fernando Hierro 85
de Souza Deco 84
Frank Lampard 83
E. Van der Sar 82
Michel Salgado 82
Roar Strand 82
Mehmet Scholl 82
A. Tacchinardi 81
Steven Gerrard 81
A. Nesta 78
M. Silvestre 78
Patrick Vieira 77
Edgar Davids 77
F. Ljungberg 77
P. Juninho 77
Andrea Pirlo 76
Jamie Carragher 75
Z. Ibrahimovic 75
M. Reiziger 74
D. Deschamps 74
Gennaro Gattuso 74
Samuel Kuffour 73
Ashley Cole 73
Kakha Kaladze 72
John Arne Riise 71
P. Djordjevic 71
Dejan Stankovic 70
Emerson 70
Lilian Thuram 69
Javier Zanetti 69
Nicky Butt 69
Sidney Govou 69
G. Pessotto 68
Dida 67
Rio Ferdinand 67
J. Zé Roberto 67
John Terry 67
J. S. Veron 67
Santiago Solari 67
A. Costacurta 66
Ricardo Carvalho 66
Denis Irwin 66
S. Cañizares 65
Jaap Stam 65
Nwankwo Kanu 65
Willy Sagnol 64
John O'Shea 64
Florent Malouda 64
H. K. Touré 63
O. Shovkovsky 62
B. Lizarazu 62
M. Ambrosini 62
Grégory Coupet 61
Erik Hoftun 61
Ciro Ferrara 61
Fabio Cannavaro 61
Phil Neville 61
M. Sanchis Hontiyuelo 59
Petr Cech 59
Jens Lehmann 59
G. Zambrotta 59
Fabien Barthez 57
--
Cheers
milivella
Futbolmetrix
milivella:
> Let's make it in the right way. The 100 players with most appearances,
> sorted by Fantasy Scout score:
a) Brilliant!
b) My main concern about Fantasy Scout CL was whether it would be possible
to retrieve accurate data on CL appearances and goals. Let's just hope that
footballdatabase.eu has a long and healthy life.
c) Now to some analysis
> If you are lazy, here are the basic data (name, score):
In brackets, their National Team FS (NTFS) score (just for the top 10)
> 1. Gonzalez Raul 187 [146]
> 2. Thierry Henry 154 [159]
> 3. Ryan Giggs 144 [76 *]
> 4. A. Shevchenko 140 [128 *]
> 5. Roberto Carlos 135 [136]
> 6. Paul Scholes 129 [80]
> 7. Luis Figo 128 [159 *]
> 8. A. Del Piero 128 [118]
> 9. R. Van Nistelrooy 128 [97]
> 10. F. Morientes 122 [74]
3 of the top 10 players would never have been eligible for Fantasy Scout.
This is the full distribution by nationality:
France 15
Italy 14
Spain 11
England 11
Netherlands 10
Brazil 7
Germany 4
Argentina 4
Norway 4
Portugal 3
Ireland 3
Ukraine 2
Sweden 2
Serbia/Yugoslavia 2
Czech Republic 2
Wales 1
Nigeria 1
Ghana 1
Cote d'Ivoire 1
Bosnia-Herzegovina 1
76 out of the top 100 players belong to one of the FS nations. Should Norway
be the next FS entrant?
Is it my impression, or is the CLFS list even more tilted towards forwards
than the NTFS list?
I couldn't resist, so I calculated the regular FS score of the players on
the list. The combined ranking (NTFS + CLFS) is the following (top 10 only,
I'll send you the Excel file):
Gonzalez Raul 333
Thierry Henry 313
Luis Figo 287
Roberto Carlos 271
A. Shevchenko 268
David Beckham 248
A. Del Piero 246
Paolo Maldini 245
Zinedine Zidane 233
Michael Ballack 229
I think that's a pretty decent summary of the best players in the game born
before 1980, even though one could quibble with the ranking. The one glaring
omission is Ronaldo. His combined FS score is 217 (NTFS = 159, CLFS = 58),
which would put him in 14th place, squeezed between Kluivert and Frank de
Boer.
D
Futbolmetrix
news:h1di4i$ja1$1...@news.eternal-september.org...
>
> Is it my impression, or is the CLFS list even more tilted towards forwards
> than the NTFS list?
Here is the distribution of players by role, and average FS points by role:
Number of
players Avg. FS Avg.CLFS Avg.NTFS
Goalkeepers: 10 149.00 70.80 78.20
Defenders: 36 149.78 74.89 74.89 (!)
Midfielders: 39 163.72 86.74 76.97
Forwards: 15 209.60 111.6 98.00
I don't think the differences are statistically significant, but there was
something to my hunch.
D
milivella
> "milivella" <milive...@gmail.com> wrote in message
>
> news:2d70e0a2-42ea-47e9...@f16g2000vbf.googlegroups.com...
> milivella:
>
> > Let's make it in the right way. The 100 players with most appearances,
> > sorted by Fantasy Scout score:
>
> a) Brilliant!
It's just a matter of copy-paste... I know that I have still a long
way to go through, my Dark Master! [1]
> b) My main concern about Fantasy Scout CL was whether it would be possible
> to retrieve accurate data on CL appearances and goals.
I know. International caps and goals OTOH are one of the first stats
that are quoted in a career summary. But Champions League match
reports (or season-wise appearances data for the single players) for
the last and - above all - the next years will be all around, so this
is not a problem.
It seems that the CLFS idea is going forward. Good. :) My suggestion
to you and to anybody interested in the project (I am the first one!)
is to treasure NTFS' experience [3]: there have been a lot of ideas
about where NTFS fall short and how to fix these problems. I guess
that you remember the main points, but if you want I can recall them.
> Let's just hope that
> footballdatabase.eu has a long and healthy life.
:) Really! While I have bookmarked 8 or so database sites, this one is
becoming my favorite one (e.g. it's the only one with 100% accurate
stats about recent caps).
> c) Now to some analysis
OK Master, my mind is ready: let's learn.
> 3 of the top 10 players would never have been eligible for Fantasy Scout.
Of course one of the advantages of CLFS over NTFS is that you don't
have to artificially limit who can be picked (Brazilian yes, Welsh
no). (this is also the reason behind my unrealistic idea of a FS based
only on World Cup finals)
> 76 out of the top 100 players belong to one of the FS nations.
Better than I expected. I guess that in the last years (less
represented in the ranking) UCL has become more global, though.
> Should Norway
> be the next FS entrant?
:)
> I couldn't resist, so I calculated the regular FS score of the players on
> the list. The combined ranking (NTFS + CLFS) is the following (top 10 only,
> I'll send you the Excel file):
I've received it, thanks. It's a lot of work, and a lot of food for
the brain.
> Gonzalez Raul 333
> Thierry Henry 313
> Luis Figo 287
> Roberto Carlos 271
> A. Shevchenko 268
> David Beckham 248
> A. Del Piero 246
> Paolo Maldini 245
> Zinedine Zidane 233
> Michael Ballack 229
>
> I think that's a pretty decent summary of the best players in the game born
> before 1980
Yep. In fact, they are all (bar Del Piero) well ranked in Enzo's
Golden Ball decade meta-ranking [2]:
7 138.6 Rau'l Gonza'lez
1 294.2 Thierry Henry
9 116.6 Lui's Figo
14 107.8 Roberto Carlos
2 218.5 Andriy Shevchenko
23 84.1 David Beckham
8 117.0 Paolo Maldini
3 215.8 Zinedine Zidane
28 66.1 Michael Ballack
The best placed players in the Golden Ball ranking are still young and
will surely get a lot of FS points in the next years:
4 194.0 Ronaldinho [OK, maybe not "a lot" of points...]
5 158.4 Cristiano Ronaldo
6 140.9 Kaka'
10 114.9 Lionel Messi
> The one glaring
> omission is Ronaldo. His combined FS score is 217 (NTFS = 159, CLFS = 58),
> which would put him in 14th place
But he's worse placed (24th) in the Golden Ball ranking: being injured
for many seasons affected his position in both the rankings.
[1] Darth Metrix, Dark Lord of Stats... I've just had the idea for a
movie plot. The basic idea is that all our universe is just a
simulation run in a spreadsheet in Daniele's own computer *in this
same universe* (Wachowski bros, you can carry my bags around). The
title is of course The FutbolMatrix.
[2] I'm using Elko's ranking:
http://groups.google.com/group/rec.sport.soccer/msg/42326d783b63052b
[3] Yes, I've already sued Microsoft for using the National Team
Fantasy Scout acronym for a filesystem.
--
Cheers
milivella
milivella
> Here is the distribution of players by role, and average FS points by role:
Great.
> Number of
> players Avg. FS Avg.CLFS Avg.NTFS
> Goalkeepers: 10 149.00 70.80 78.20
> Defenders: 36 149.78 74.89 74.89 (!)
> Midfielders: 39 163.72 86.74 76.97
> Forwards: 15 209.60 111.6 98.00
Isn't it better to count the average score of the _same_ number of
players for each role? This is the average score of the top 10 GKs,
the top 10 DFs, etc.:
GKs 149.0 = 70.8 + 78.2
DFs 205.4 = 87.5 + 117.9 (!!)
MFs 224.4 = 110.2 + 114.2
FWs 239.5 = 125.2 + 114.2
> I don't think the differences are statistically significant, but there was
> something to my hunch.
Always follow your hunches,
Always scratch when it itches.
You feel time-wasting maybe,
but that's the research, baby.
--
Cheers
milivella
milivella
> maybe it's time to launch a completely new game,
> independent of Fantasy Scout: Fantasy Scout Champions' League
>
> Rules:
> Pick any number of players, from any nationality, under the age of 21, who
> have never appeared in a CL match. (Age = current year - year of birth).
> Points = Appearances + goals in the Champions' League proper - 10 points for
> every player picked.
Let me list the pros and cons of the game when compared to National
Team Fantasy Scout. *
Pro:
1. Every football player on the planet is eligible (I've already
written about this pro).
2. A scout isn't forced to pick players, but it's rewarded if he does
it (wisely, otherwise he is punished by the penalty!).
3. Computations are easier to do: it's straighter to do a sum (even
with penalties) than to compute an average (see the first post in this
thread).
4. Less rules: in NTFS, there are specifications about eligible
players (see pro #1) and the number of players to be picked (see pro
#2).
Cons:
1. The game doesn't scale well, because the number of football players
expected to have a positive score is fixed, no matter how many scouts
play the game. E.g. if there are 2 scouts the best thing for them is
to pick a lot of players, but if there are 200 scouts the best
strategy is probably to pick just one player.
2. Would the game be based on transfer market? (e.g. "Manchester
United bought this prospect, so he'll score some point: I pick him") I
don't know.
* Some disclaimers:
- Here, for the sake of discussion, I'm assertive, but my opinions are
way more blurred.
- The list is by no mean complete.
- I'm assuming that the quoted rules are all the basic rules of the
game (e.g. there is no duty to call at least 1 player per season).
- As I've written, there are some features that the CLFS and NTFS
share. I'm not writing about these features here.
MH
milivella wrote:
> It's that time of the year again: I'm itching for a Fantasy Scout
> reform... ;) But before officially proposing it, I'd like to have as
> much feedback as possible, so please reply with your opinions and
> criticisms.
>
> 1. Each year, a scout can pick up to 10 players.
>
> (We can make it 12 as the number of the months, or any other number.)
10 is plenty.
Like Benny, I would not mind dropping a certain number per year. Even one.
>
> 2. There is a separate scout ranking for each year of pick. The score
> of a scout is the sum of the points earned by the players he picked in
> that year.
>
> (So there will be a ranking for players picked in 2010, a different
> one for players picked in 2011, and so on.)
>
> (Going for the sum instead of the average makes things straighter,
> because it's easier to compare the performance of two scouts: "I have
> 4 starters", "But I have 6, 2 of which are goal-machines: I win!";
> with the current system, you have to divide by the number of picked
> players.)
>
> (One nice feature of yearly rankings is that we'll be able to change
> rules - about picking players or scoring - without hurting what has
> happened so far.)
>
No problem with this as a yearly ranking but ongoing should be average
as that is the current system, and at least one player objected to any
changes to to that.
Enzo
>
> - Show quoted text -- Hide quoted text -
>
> - Show quoted text -
I like the dropping idea too. When I started playing this
game, I did so under certain assumptions that I now realize
were incorrect. I have been watching quietly for a while
and have a better idea of how to play this game now. I
should be able to drop a player to trim my losses.
milivella
Thanks fr the feedback. Indeed I have already retired the proposal,
because it shouldn't be strictly fair:
http://groups.google.com/group/rec.sport.soccer/msg/57a9f5243b4d1643
In this moment, I think that, having started with a system based on
average (my idea!), we should keep it. This means that a match-by-
match update is maybe useless (checking the scores once every 6 or 12
months should work); I'll keep updating the score after every "round"
of international matches, though. (I'm always open to change my mind
about the scoring system, anyway.)
I've also asked you, Michael, to write something - if you want - to
explain us why switching to the H-index wouldn't harm the scouts who
picked few players. It could be the right way to go.
> Like Benny, I would not mind dropping a certain number per year. Even one.
It's OK for me. I only add that, to be fair, the number of players to
be dropped should be not fixed (e.g. 1 for each scout), but
proportional to the number of players picked by that scout (e.g. 10%:
if a scout has picked 20 players, the can drop 2; if 13, he can drop
1). In fact, dropping 1 player could be a big improvement if you have
picked 2 players, but it would do pretty nothing to your score if you
have picked 100 players. (You see, this time I want to protect who has
picked more players, just like in other cases I felt that the
proposals would have harmed who had picked few players.)
(I also don't know how, if everyone can drop players, anyone can have
an advantage out of it, and so why you like this rule. But this point
doesn't matter: if someone will officially propose to add a rule to
drop x% of the players that a scout has picked (in the previous
year?), I'll vote for such a proposal.)
--
Cheers
milivella
milivella
> I like the dropping idea too. When I started playing this
> game, I did so under certain assumptions that I now realize
> were incorrect. I have been watching quietly for a while
> and have a better idea of how to play this game now. I
> should be able to drop a player to trim my losses.
As I've just written, if we all will be able to drop a player, nobody
will trim his losses. Am I wrong?
Anyway: if we shape the dropping rule in a more precise form, someone
will be able to officially propose it. My two suggestions (see my
previous post) are:
- number of players to be dropped should be proportional to the number
of picked players
- you shouldn't be able to drop players whose career has ended,
because this way it's too easy to spot who was bad
--
Cheers
milivella
Enzo
> Enzo:
>
> > I like the dropping idea too. When I started playing this
> > game, I did so under certain assumptions that I now realize
> > were incorrect. I have been watching quietly for a while
> > and have a better idea of how to play this game now. I
> > should be able to drop a player to trim my losses.
>
> As I've just written, if we all will be able to drop a player, nobody
> will trim his losses. Am I wrong?
>
> Anyway: if we shape the dropping rule in a more precise form, someone
> will be able to officially propose it. My two suggestions (see my
> previous post) are:
> - number of players to be dropped should be proportional to the number
> of picked players
Dont like that. If I select 10 players of whom 9 are doing badly,
I can drop more than someone who selects 4 players of whom 1
is doing badly? I think it is best not to allow to drop players.
milivella
in Fantasy Scout. I'll briefly write it here (but I'll be a bit more
verbose, or link to more verbose posts, in the notes).
Fantasy Scout is not a game of immediate rewards, but instead the nice
thing is discussing with the other contestants, after years of play,
about who has been better in spotting young talents [1]. But the
scoring system makes such discussions hard to have [2]. So the only
thing that we can do is to check the numbers: who has the highest
score? [3]
Now, this is not something the you can't live with. It just makes the
game too abstract to make it _funny_ in any sense of this word. OTOH,
it's the only game that I know about about spotting young talents [4],
so it's better than nothing [5].
Notes:
[1] See http://groups.google.com/group/rec.sport.soccer/browse_thread/thread/419571c5cb6cef39
[2] See the first post in this thread.
[3] To make my argument a little more concrete, here is a fictional
dialog between the two fictional scouts A and D (any resemblance to
real persons, living or dead...):
August 2009
A: I'm a better scout than you, because yesterday you scored 1 goal
through Amauri, but I scored 2 through "Bimbo" Walcott!
D: Why this hurry? These goals are just drops in the ocean. In ten
years we'll see who has picked better players in 2008-09.
A: OK. So... see you in 2019. Bye.
D: Bye.
August 2019
A: Hey, ten years have passed. I'm a better scout than you, because
you haven't picked any 120-point player, while I've picked Gago!
D: OK, but you have picked 150 players, while I've picked just 30. So
you can't just compare your best player with my best player; the
average of your best 5 players vs. my best player would be a better
start. And anyway there are 176 more players to consider...
A: ...So I only can compare our scores. It's just this, a number vs.
another number, isn't it?
D: Yes it is.... And I win!
A: Fantasy Scout sucks. [in 10 years, A will be heavily influenced by
D's slang]
[4] Enzo will probably disagree... but I think he's slightly wrong in
the long run. ;)
[5] I now see Fantasy Scout a bit like the Elo rating system, because
the following description is true for them both: it's the best system
if you want to tell who is better, because direct comparisons are not
possible (e.g. because of the different number of players/matches);
but it's not intuitive, so, when they discuss, people will always use
different arguments (in the Elo case, WC performance).
--
Cheers
milivella
milivella
> Fantasy Scout is not a game of immediate rewards, but instead the nice
> thing is discussing with the other contestants, after years of play,
> about who has been better in spotting young talents [1]. But the
> scoring system makes such discussions hard to have [2]. So the only
> thing that we can do is to check the numbers: who has the highest
> score? [3]
> [3] To make my argument a little more concrete, here is a fictional
> dialog
Even more concrete: we had a nice discussion about picked players in
March:
http://groups.google.com/group/rec.sport.soccer/browse_thread/thread/50169286070c22e7/
Well, I sadly don't see how something similar could be done when the
number of picked players will be higher.
--
Cheers
milivella
Enzo
> I've finally (you know, I'm slow) had the insight about what's wrong
> in Fantasy Scout. I'll briefly write it here (but I'll be a bit more
> verbose, or link to more verbose posts, in the notes).
>
> Fantasy Scout is not a game of immediate rewards, but instead the nice
> thing is discussing with the other contestants, after years of play,
> about who has been better in spotting young talents [1]. But the
> scoring system makes such discussions hard to have [2]. So the only
> thing that we can do is to check the numbers: who has the highest
> score? [3]
>
> Now, this is not something the you can't live with. It just makes the
> game too abstract to make it _funny_ in any sense of this word. OTOH,
> it's the only game that I know about about spotting young talents [4],
> so it's better than nothing [5].
>
> Notes:
>
Well it is not. If it was about spotting young talent,
it would have certain rules like a player cannot be
older than a certain age, cannot have played more than
a certain number of games as a professional etc. Thats
what I originally thought.
No, it is, and I am fine with that, a game where people
try to be the first to snap up a player who is kind
of on the fringes after making a quick assessment of
the strength in the NT at his position. Most players
picked are already (very) well-known.
milivella
> On Jun 22, 2:09 pm, milivella <milive...@gmail.com> wrote:
>
> > it's the only game that I know about about spotting young talents [4],
>
> > [4] Enzo will probably disagree... but I think he's slightly wrong in
> > the long run. ;)
>
> Well it is not. If it was about spotting young talent,
> it would have certain rules like a player cannot be
> older than a certain age, cannot have played more than
> a certain number of games as a professional etc. Thats
> what I originally thought.
Can you be more precise about the limits (about age, games, etc.) that
you would put?
I ask it for two reasons: first of all, because it would be great to
make the game better by the addition of such limits (let's say from
next year) [1]; in second place, because I myself can't think about
any limit that would work well enough in all the cases, and, having
you in high esteem, have great hopes in your ability to find the sweet
spots.
> No, it is, and I am fine with that, a game where people
> try to be the first to snap up a player who is kind
> of on the fringes after making a quick assessment of
> the strength in the NT at his position. Most players
> picked are already (very) well-known.
I think that, with more scouts playing and the duty to pick at least
two players each year, this won't be true anymore. See, Italy (the
nation with most picks so far) has not only all the players with a
reasonably predictable career in NT picked, but practically all the
U-21 team picked as well. I don't think that "very well-known" (we
mean at international level, let's remember it) Italian players are
free now.
Of course, I can be wrong, and anyway making the game more difficult
(while still not based on luck) would be great, so let me re-iterate
my invitation for you to provide a detailed view of the limits that
you would put. I'll play the devil's advocate and try to criticize
them, but I'm pretty sure that in the end we'll make a better game. :)
[1] The only con that I see is that scouts who'll join after the
reform would have a disadvantage... But they would have it anyway
because more scouts = more competition. And I guess that in the long
run the disadvantages will be less and less. And: we could agree to
adopt these limits retroactively, e.g. deleting players who were too
old when picked (I know that it could seem unrealistic to have
everyone agreeing on this, but something similar has happened in the
past).
--
Cheers
milivella
milivella
> > it's the only game that I know about about spotting young talents [4]
>
> > [4] Enzo will probably disagree... but I think he's slightly wrong in
> > the long run. ;)
>
> Well it is not. If it was about spotting young talent,
> it would have certain rules like a player cannot be
> older than a certain age, cannot have played more than
> a certain number of games as a professional etc. Thats
> what I originally thought.
>
> No, it is, and I am fine with that, a game where people
> try to be the first to snap up a player who is kind
> of on the fringes after making a quick assessment of
> the strength in the NT at his position. Most players
> picked are already (very) well-known.
What I've written in my first reply stand, but I fear that there is
another point, because I could switch my assertion from "FS is about
spotting young talents" to "FS is about spotting the next Ballon d'Or-
shortlisted players", but I'm not sure if you agree...
--
Cheers
milivella
milivella
> I've finally (you know, I'm slow) had the insight about what's wrong
> in Fantasy Scout.
But I think that it's possible to solve the main problems of the
games. I have some ideas about the solutions, but *I need your help*.
So I propose to you all to carry forward a two-step project:
1. Let's identify the PROBLEMS of the game
2. Let's try to find the possible SOLUTIONS to those problems.
I'm writing down my first thoughts about both the points, but what
I'll write can't be anything more than a stimulus for you to reply
with *your* ideas, criticisms, proposals.
The game can get better, and the only way to obtain such a result is
putting together everybody's ideas.
--
Cheers
milivella
milivella
Here are all the negative aspects of the game that I'm able to
identify at the present. If something is not clear, please tell me,
and I'll try to clarify. Here onwards, scout = who plays Fantasy Scout
(e.g. Jesus), while player = football player (e.g. Pato).
CHANGES: It's difficult to change the rules, because there is a past
situation to be kept intact for fairness (and because harmed scouts
would not accept the changes!). This is in my opinion the most
important thing to fix, because, if you fix it, then you’ll be always
able to change things later.
CONTINUUM: The score of a player is a continuum, i.e. each cap or goal
is 1 more point, there are not moments that are more important than
others (and anyone who prefer our low-scoring sport over high-scoring
ones like basketball knows that having few scoring moments is more
fun).
WORLD: You can pick only players from 8 nations.
DROPS: You can't drop the players that you’ve picked.
CLUBS: Only the NT career of a player counts in Fantasy Scout, but in
reality his club career is important as well.
SIMPLE MODEL: Winning matches or playing important matches (e.g. the
WC final) are not rewarded.
POSITIONS: There is no difference in the scoring system for the
different positions: GKs and FWs are given points by the same rule.
TIME: There's a lot of time between your action (you picking a player)
and the effect (he scoring points). Also, there aren't intermediate
goals, because there is just one eternal ranking.
NOT INTUITIVE: Comparing the performance of different scouts is not
something intuitive, because it involves averages. Also, different
scouts could have picked players of different ages.
EASY: It's maybe too easy to spot players who will have some caps.
Anyway, betting on youngest players is not rewarded per se.
Note that probably to fix some of these problems you in turn
necessarily create some other problems (because there actually are
good things in the game!). So I'm not saying that everything can/
should be fixed. But it's better to have a complete list of problems,
and only then check what we really want to realize.
--
Cheers
milivella
milivella
I will here cite the problems that I've highlighted in the last post.
---
First of all, as I've said, we need to fix CHANGES. Here is my
proposal:
1. The overall ranking is based on limited rankings, with criteria
being in order: most 1st places (in limited rankings); most 2nd
places; most 3rd places; and so on.
2. The first limited ranking is given by the average score of players
picked in the years 2007-09. [So what has happened so far will never
be changed: picks made so far will be valued by the rules that have
been in effect so far.]
I stop here, because everything else can be discussed now, and anyway
changed in the future. But I also suggest: to have, starting with
2010, yearly rankings, based on players picked in each given year; and
to give a 2x weight to the 2007-09 ranking, because it involves more
time than the next rankings.
A little note: I wrote about _limited_ rankings. I know that this
adjective is ambiguous and not elegant. But it's _intentionally_
ambiguous, because "limited" can mean "yearly", "monthly", "for each
position on the field", "for each nation", etc. If "limited" means a
temporal limit (I’m suggesting an yearly limit), then the second part
of the TIME problem is fixed, because there would be intermediate
goals: at some point - let's say in 2025 (!) - we'll have a definitive
champion for the 2007-09 ranking.
---
Then there are all the other problems. Once we fix CHANGES, we have no
hurry to tackle them, because we know that we’ll always be able to do
it later. This said, I personally feel that some problems are more
urgent than the others. In this moment, I can live with CONTINUUM,
DROPS, SIMPLE MODEL, POSITIONS, TIME (the first part: for the second
one, see above) and EASY. Then there are WORLD and (possibly related)
CLUBS, that could be easily fixed, but it will need (if I’m not wrong)
more computations, and I don’t think that anyone want to do those
computations...
Indeed, I'd like to fix NOT INTUITIVE. I0ve already posted a simple
solution to both in this thread, let me quote it:
1. Each year, a scout can:
- pick up to 10 players
- select 1 out of the players picked the previous year to permanently
represent him
2. The yearly rankings are based on the score of the single players
that represent the scouts.
So, a little example (involving just two scouts) to be clear: in 2010
Alessandro and Alberto both pick 10 players. In 2011, Alessandro
chooses on of those 10 players, Okaka, to represent him, and Alberto
similarly select Poli. Alessandro will top the 2010 ranking if Okaka,
at the end of their career, will have more points than Poli; Alberto
will top the 2010 ranking if Poli will have more points than Okaka.
The NOT INTUITIVE problem would be fixed, because it would be very
intuitive to compare one player of yours vs. one player for each of
the other scouts (they will also probably be almost of the same age),
considering that in any moment you probably won’t have more than 15
active players.
--
Cheers
milivella
milivella
you don't accept the second one):
1. Many rankings, from which an eternal meta-ranking is computed. The
first ranking, spanning the three years 2007-09, is based on the old
rules (= average of all the picked players).
2. The next rankings are yearly and are based on the following rule:
each year, a scout select one player to represent him (so the
player=scout with most points at the end of his career tops the
ranking for that year).
...and now I wait for your ideas.
--
Cheers
milivella
Enzo
> Enzo:
>
> > On Jun 22, 2:09 pm, milivella <milive...@gmail.com> wrote:
>
> > > it's the only game that I know about about spotting young talents [4],
>
> > > [4] Enzo will probably disagree... but I think he's slightly wrong in
> > > the long run. ;)
>
> > Well it is not. If it was about spotting young talent,
> > it would have certain rules like a player cannot be
> > older than a certain age, cannot have played more than
> > a certain number of games as a professional etc. Thats
> > what I originally thought.
>
> Can you be more precise about the limits (about age, games, etc.) that
> you would put?
>
> I ask it for two reasons: first of all, because it would be great to
> make the game better by the addition of such limits (let's say from
> next year) [1]; in second place, because I myself can't think about
> any limit that would work well enough in all the cases, and, having
> you in high esteem, have great hopes in your ability to find the sweet
> spots.
>
For starters, you can rule out all players who have U-21 caps.
This will make the job much harder. The problem of course
is that the game will become more boring, in the sense
that no one will see immediate returns on investment.
I am actually ok with the rules now, I only say the above
to point out that this is not so much a "scout" competition
as a very interesting game.
Btw, is the website still fantasyscout.altervista.org?
The data has not been updated in a while. I should
have a higher score.
> > No, it is, and I am fine with that, a game where people
> > try to be the first to snap up a player who is kind
> > of on the fringes after making a quick assessment of
> > the strength in the NT at his position. Most players
> > picked are already (very) well-known.
>
> I think that, with more scouts playing and the duty to pick at least
> two players each year, this won't be true anymore. See, Italy (the
> nation with most picks so far) has not only all the players with a
> reasonably predictable career in NT picked, but practically all the
> U-21 team picked as well. I don't think that "very well-known" (we
> mean at international level, let's remember it) Italian players are
> free now.
>
You can compensate for that by throwing the doors open
to at least a few more NTs. These have to be NTs which
a reasonably large number of scouts/future scouts follow, else it
will give someone a unfair advantage. Lets say Portugal,
United States, Mexico, Australia etc
milivella
> For starters, you can rule out all players who have U-21 caps.
I know that you have the right ideas! (For this reason, I ask you to
comment on the more general problems-and-solutions topic.)
I like your proposal. The only two negative aspects that I can think
of:
- Brazil and Argentina doesn't have an U-21 team. They have an U-20
team, but if U-20 is under 20 then U-21 is under 23, so they are not
comparable at all.
- Where do you can find 100% complete data about U-21 caps? I don't
know.
> This will make the job much harder. The problem of course
> is that the game will become more boring, in the sense
> that no one will see immediate returns on investment.
This is not a issue IMHO, because even now in most cases you don't see
any immediate return.
> Btw, is the website still fantasyscout.altervista.org?
Yes (I swear that, if I'll ever change the address, I'll add a re-
direct from the old address to the new one!)
> The data has not been updated in a while. I should
> have a higher score.
You're right. I've written before the end of May (it was the e-mail
with the Fantasy Scout Football Club photo!) that I'll update
everything after the Confederations Cup final. In general, my policy
is to update the scores after each international match "window" (do
you prefer more updates?).
> You can compensate for that by throwing the doors open
> to at least a few more NTs. These have to be NTs which
> a reasonably large number of scouts/future scouts follow, else it
> will give someone a unfair advantage. Lets say Portugal,
> United States, Mexico, Australia etc
This is a recurrent proposal. The points usually risen are:
- Is it OK to reward a scout who picks Paul Wade (94 points, 1986-96)
more than a scout who picks John Barnes, Careca or Bergomi (90, 89, 87
points in the same era)?
- How do you choose the national teams to include? It's not easy (see
http://groups.google.com/group/rec.sport.soccer/browse_thread/thread/b8c1e637a19b511d#
second post and below).
--
Cheers
milivella
Enzo
> Enzo:
>
> > For starters, you can rule out all players who have U-21 caps.
>
> I know that you have the right ideas! (For this reason, I ask you to
> comment on the more general problems-and-solutions topic.)
>
> I like your proposal. The only two negative aspects that I can think
> of:
> - Brazil and Argentina doesn't have an U-21 team. They have an U-20
> team, but if U-20 is under 20 then U-21 is under 23, so they are not
> comparable at all.
> - Where do you can find 100% complete data about U-21 caps? I don't
> know.
Not sure. I can find out if needed.
>
> > This will make the job much harder. The problem of course
> > is that the game will become more boring, in the sense
> > that no one will see immediate returns on investment.
>
> This is not a issue IMHO, because even now in most cases you don't see
> any immediate return.
>
> > Btw, is the website still fantasyscout.altervista.org?
>
> Yes (I swear that, if I'll ever change the address, I'll add a re-
> direct from the old address to the new one!)
>
> > The data has not been updated in a while. I should
> > have a higher score.
>
> You're right. I've written before the end of May (it was the e-mail
> with the Fantasy Scout Football Club photo!) that I'll update
> everything after the Confederations Cup final. In general, my policy
> is to update the scores after each international match "window" (do
> you prefer more updates?).
No, no, thats fine.
>
> > You can compensate for that by throwing the doors open
> > to at least a few more NTs. These have to be NTs which
> > a reasonably large number of scouts/future scouts follow, else it
> > will give someone a unfair advantage. Lets say Portugal,
> > United States, Mexico, Australia etc
>
> This is a recurrent proposal. The points usually risen are:
> - Is it OK to reward a scout who picks Paul Wade (94 points, 1986-96)
> more than a scout who picks John Barnes, Careca or Bergomi (90, 89, 87
> points in the same era)?
True.
Another thought, for a different kind of scouting game.
Instead of only awarding points for international appearances,
also award points for player signings by big clubs.
First, make a set of big clubs ( EPL big 4, LL big 2, SA big 3
should do for the moment ). The restriction is that the player
should not have played outside his home country ever, and not
in any of the big 3 leagues. Also, he should not have any
international caps ( max 2 ). To avoid situations where a 30
year old player is signed by say Liverpool as cover, make
a age restriction ( max 21 ). Then people with knowledge
of regional football can use that to gain some points,
instead of being restricted to the big nations.
I havent really thought about the above in detail,
maybe I can start a separate competition if it is not
used in fantasy scout, but it seems like a scouting
competition.
> - How do you choose the national teams to include? It's not easy (seehttp://groups.google.com/group/rec.sport.soccer/browse_thread/thread/...
Futbolmetrix
> > 1. Let's identify the PROBLEMS of the game
>
First, the gamemaster has too many ideas, writes too many posts and
follows up on them on his own, and it's difficult to keep track of
him :-)
> CHANGES: It's difficult to change the rules, because there is a past
> situation to be kept intact for fairness (and because harmed scouts
> would not accept the changes!). This is in my opinion the most
> important thing to fix, because, if you fix it, then you’ll be always
> able to change things later.
Theorem: there does not exist a game such that, if it began under a
set of rules, and players chose an optimal strategy under that set of
rules, then the rules can be changed and none of the players will be
left worse off. Prove the theorem or show a counterexample.
> CONTINUUM: The score of a player is a continuum, i.e. each cap or goal
> is 1 more point, there are not moments that are more important than
> others (and anyone who prefer our low-scoring sport over high-scoring
> ones like basketball knows that having few scoring moments is more
> fun).
I'm not particularly bothered by this.
> WORLD: You can pick only players from 8 nations.
Nor by this.
> DROPS: You can't drop the players that you’ve picked.
That is part of the original package of the game. If a scout knew that
he could have dropped players, he would have adopted a different
strategy.
> CLUBS: Only the NT career of a player counts in Fantasy Scout, but in
> reality his club career is important as well.
This can be easily fixed, even within the context of the current game:
CLFS!
> SIMPLE MODEL: Winning matches or playing important matches (e.g. the
> WC final) are not rewarded.
>
That's one of the advantages of the game.
> POSITIONS: There is no difference in the scoring system for the
> different positions: GKs and FWs are given points by the same rule.
Not a problem:
a) there's a tradeoff between simplicity of the rule and the
unfairness of different positions being treated in the same way.
b) It's also more realistic: forwards are valued more than defenders
in the real world. I'm sure that the real world scout that was the
first to discover Maradona earned more than the first to discover
Maldini.
> TIME: There's a lot of time between your action (you picking a player)
> and the effect (he scoring points). Also, there aren't intermediate
> goals, because there is just one eternal ranking.
This is a shortcoming.
> NOT INTUITIVE: Comparing the performance of different scouts is not
> something intuitive, because it involves averages. Also, different
> scouts could have picked players of different ages.
If I were to start a new game, I would pick a different scoring
formula. As it is, we're stuck with averages.
> EASY: It's maybe too easy to spot players who will have some caps.
> Anyway, betting on youngest players is not rewarded per se.
In a new game, maybe you would put an age restriction on the players
that you can pick.
D
Futbolmetrix
> Futbolmetrix:
>
>
> Pro:
>
> 1. Every football player on the planet is eligible (I've already
> written about this pro).
Yep.
> 2. A scout isn't forced to pick players, but it's rewarded if he does
> it (wisely, otherwise he is punished by the penalty!).
But we could of course institute a minimum number of players
requirement.
> 3. Computations are easier to do: it's straighter to do a sum (even
> with penalties) than to compute an average (see the first post in this
> thread).
The sum also has the property that the contribution of each player is
independent of the contributions of all the others. Something we had
already discussed.
> 4. Less rules: in NTFS, there are specifications about eligible
> players (see pro #1) and the number of players to be picked (see pro
> #2).
But of course one could add more rules if they make the game better.
> Cons:
>
> 1. The game doesn't scale well, because the number of football players
> expected to have a positive score is fixed, no matter how many scouts
> play the game. E.g. if there are 2 scouts the best thing for them is
> to pick a lot of players, but if there are 200 scouts the best
> strategy is probably to pick just one player.
Hmmm.. not sure I understand, and not sure why it's a problem.
> 2. Would the game be based on transfer market? (e.g. "Manchester
> United bought this prospect, so he'll score some point: I pick him") I
> don't know.
Probably that will end up being many scouts' strategy. A bit like in
NTFS that scouts pick players on the fringe of the national team, and
in positions of relative weakness of their NT.
Anyway, here is an updated set of rules:
1) The basic competition is yearly. One of course can then aggregate
the score over years.
2) Every August, each scout must select N players, under the age of X,
with no more than Y CLFS points up to that point of their career.
3) In January (the January transfer window), one can replace Z players
in his roster.
4) The winner of the yearly competition is the scout with most CLFS
points in that year (August-May)
D
milivella
> > - Where do you can find 100% complete data about U-21 caps? I don't
> > know.
>
> Not sure. I can find out if needed.
It would be useful if you want to make it an official proposal.
> Another thought, for a different kind of scouting game.
> Instead of only awarding points for international appearances,
> also award points for player signings by big clubs.
>
> First, make a set of big clubs ( EPL big 4, LL big 2, SA big 3
> should do for the moment ). The restriction is that the player
> should not have played outside his home country ever, and not
> in any of the big 3 leagues. Also, he should not have any
> international caps ( max 2 ). To avoid situations where a 30
> year old player is signed by say Liverpool as cover, make
> a age restriction ( max 21 ). Then people with knowledge
> of regional football can use that to gain some points,
> instead of being restricted to the big nations.
It seems similar to the Champions League Fantasy Scout by Daniele
(proposed above in this thread, and then again after your post). And I
like them both!
> I havent really thought about the above in detail,
> maybe I can start a separate competition if it is not
> used in fantasy scout, but it seems like a scouting
> competition.
As I've written, I have a slight preference for the integration of
these new ideas in the old Fantasy Scout (I think that it's possible:
let's switch to yearly rankings), because I don't like to have a
player forever on my side in a game and against me in another one.
But anyway I would absolutely play such a new scouting game based on
club football! :)
--
Cheers
milivella
milivella
> On Jun 22, 7:01 pm, milivella <milive...@gmail.com> wrote:
>
> > > 1. Let's identify the PROBLEMS of the game
>
> First, the gamemaster has too many ideas, writes too many posts and
> follows up on them on his own, and it's difficult to keep track of
> him :-)
:) Really, I know. But I see a lot of things that we can do in this
space! [1] Anyway, I swear that I'll be quieter now, and my effective
proposals are limited to two radical but simple changes, that I've
summed up in less than 400 characters (spaces included!) here:
http://groups.google.com/group/rec.sport.soccer/msg/aa4613a8b84b7f28
[1] This is exactly the reason why I'm proposing to have yearly
rankings: I guess that we'll keep having new ideas forever (you and
Enzo had the idea of using club football after almost 3 years since my
first post about Fantasy Scout), and yearly rankings are the way to
include these new ideas in the game. Otherwise, we'll come up with a
lot of different games, or with a lot of unused brilliant ideas!
> Theorem: there does not exist a game such that, if it began under a
> set of rules, and players chose an optimal strategy under that set of
> rules, then the rules can be changed and none of the players will be
> left worse off. Prove the theorem or show a counterexample.
Counterexample (I hope, please tell me if I'm wrong): the present set
of rules is not touched, but it's used to compute a first ranking.
Then we change the rules (picking, scoring, etc.) and all that happens
after this change is used to compute a different ranking.
A possible objection could be that who had the lead under the old
rules restarts on pair with the other scouts under the new rules. But
let's remember that he was forced to keep picking players under the
old rules, so he had to do better picks than his opponents in any way.
> > CONTINUUM: The score of a player is a continuum, i.e. each cap or goal
> > is 1 more point, there are not moments that are more important than
> > others (and anyone who prefer our low-scoring sport over high-scoring
> > ones like basketball knows that having few scoring moments is more
> > fun).
>
> I'm not particularly bothered by this.
Thanks for the detailed feedback. I'm keeping track of our opinions
and feelings about the different problems.
> > DROPS: You can't drop the players that you’ve picked.
>
> That is part of the original package of the game. If a scout knew that
> he could have dropped players, he would have adopted a different
> strategy.
I agree with you.
> > CLUBS: Only the NT career of a player counts in Fantasy Scout, but in
> > reality his club career is important as well.
>
> This can be easily fixed, even within the context of the current game:
> CLFS!
I was obviously thinking about CLFS as a solution when I wrote about
the problem.
> > SIMPLE MODEL: Winning matches or playing important matches (e.g. the
> > WC final) are not rewarded.
>
> That's one of the advantages of the game.
Why? Are there other reason other than the easiness of computations?
Just curious.
> > POSITIONS: There is no difference in the scoring system for the
> > different positions: GKs and FWs are given points by the same rule.
>
> Not a problem:
> a) there's a tradeoff between simplicity of the rule and the
> unfairness of different positions being treated in the same way.
> b) It's also more realistic: forwards are valued more than defenders
> in the real world. I'm sure that the real world scout that was the
> first to discover Maradona earned more than the first to discover
> Maldini.
Maldini actually has more FS points than Maradona (133 vs. 125)! ;)
Anyway, I've said too many times that the current formula in my
opinion just works: using just two easy-to-find data, it gives you
rankings that are quite close to the aggregate subjective ones (best-
player-of-all-time polls, etc.).
> > TIME: There's a lot of time between your action (you picking a player)
> > and the effect (he scoring points). Also, there aren't intermediate
> > goals, because there is just one eternal ranking.
>
> This is a shortcoming.
How would you fix it? With the season-wise CLFS you've propose in your
next post?
> > NOT INTUITIVE: Comparing the performance of different scouts is not
> > something intuitive, because it involves averages. Also, different
> > scouts could have picked players of different ages.
>
> If I were to start a new game, I would pick a different scoring
> formula.
I know. ;)
> As it is, we're stuck with averages.
See my counterexample to your theorem: IMHO we're stuck with averages
*for the players picked so far*, but we can change things *for the
players picked e.g. from 2010*.
> > EASY: It's maybe too easy to spot players who will have some caps.
> > Anyway, betting on youngest players is not rewarded per se.
>
> In a new game, maybe you would put an age restriction on the players
> that you can pick.
Of course the Patos and the Santons of this world escape any age
restriction, but it could be a good idea.
--
Cheers
milivella
milivella
> On Jun 20, 3:12 am, milivella <milive...@gmail.com> wrote:
>
> > 3. Computations are easier to do: it's straighter to do a sum (even
> > with penalties) than to compute an average (see the first post in this
> > thread).
>
> The sum also has the property that the contribution of each player is
> independent of the contributions of all the others. Something we had
> already discussed.
True, and isn't it odd that the economist among us thinks that making
the picking strategy simpler is so important? ;) (under a sum-with-
penalty system, you just have to ask e.g. "will this player play at
least 10 UCL matches? if yes, pick him; if no, don't pick him")
> > 4. Less rules: in NTFS, there are specifications about eligible
> > players (see pro #1) and the number of players to be picked (see pro
> > #2).
>
> But of course one could add more rules if they make the game better.
Of course. I'm just saying that you've obtained a game of the same
quality of NTFS with less rules.
> > 1. The game doesn't scale well, because the number of football players
> > expected to have a positive score is fixed, no matter how many scouts
> > play the game. E.g. if there are 2 scouts the best thing for them is
> > to pick a lot of players, but if there are 200 scouts the best
> > strategy is probably to pick just one player.
>
> Hmmm.. not sure I understand, and not sure why it's a problem.
You've maybe partly replied writing that "we could of course institute
a minimum number of players requirement", but let me try to be less
obscure. Let's suppose that you launch CLFS, and in the first month
only two scouts join the game: they'll pick every prospect who is
worth at least points, let's say they'll pick 20 players each. Then
the world opens his eyes and understand that this game is really nice,
and in few days 198 more scouts join the game: the number of good
enough (i.e. worth 10 points) players will be the same, but now the
scouts will be 100x more, so each of them will be happy if he picks 1
or 2 good players. (Using average, you have not such a situation,
because when the number of scouts increases the number of points that
make a player good enough decreases.)
Maybe this is not a problem, it depends on your taste. I just wanted
to say that the game nature changes when the number of scouts changes,
and that with a high enough number of scouts it could be *really* hard
to find just 1 good enough player.
> Anyway, here is an updated set of rules:
>
> 1) The basic competition is yearly. One of course can then aggregate
> the score over years.
> 2) Every August, each scout must select N players, under the age of X,
> with no more than Y CLFS points up to that point of their career.
> 3) In January (the January transfer window), one can replace Z players
> in his roster.
> 4) The winner of the yearly competition is the scout with most CLFS
> points in that year (August-May)
Interesting. I'll think a bit about it and eventually write something.
--
Cheers
milivella
Enzo
> Enzo:
>
> > > - Where do you can find 100% complete data about U-21 caps? I don't
> > > know.
>
> > Not sure. I can find out if needed.
>
> It would be useful if you want to make it an official proposal.
>
Later.
> > Another thought, for a different kind of scouting game.
> > Instead of only awarding points for international appearances,
> > also award points for player signings by big clubs.
>
> > First, make a set of big clubs ( EPL big 4, LL big 2, SA big 3
> > should do for the moment ). The restriction is that the player
> > should not have played outside his home country ever, and not
> > in any of the big 3 leagues. Also, he should not have any
> > international caps ( max 2 ). To avoid situations where a 30
> > year old player is signed by say Liverpool as cover, make
> > a age restriction ( max 21 ). Then people with knowledge
> > of regional football can use that to gain some points,
> > instead of being restricted to the big nations.
>
> It seems similar to the Champions League Fantasy Scout by Daniele
> (proposed above in this thread, and then again after your post). And I
> like them both!
>
Not exactly the same. This is more restrictive, in the same
sense as the NTFS with its 8 or 9 teams. Whereas that
unless I misread it is for the entire CL competition.
> > I havent really thought about the above in detail,
> > maybe I can start a separate competition if it is not
> > used in fantasy scout, but it seems like a scouting
> > competition.
>
> As I've written, I have a slight preference for the integration of
> these new ideas in the old Fantasy Scout (I think that it's possible:
> let's switch to yearly rankings), because I don't like to have a
> player forever on my side in a game and against me in another one.
>
Same here.
milivella
> On Jun 24, 6:24 am, milivella <milive...@gmail.com> wrote:
>
> > It would be useful if you want to make it an official proposal.
>
> Later.
OK, we have no hurry at all.
> > It seems similar to the Champions League Fantasy Scout by Daniele
> > (proposed above in this thread, and then again after your post). And I
> > like them both!
>
> Not exactly the same. This is more restrictive, in the same
> sense as the NTFS with its 8 or 9 teams. Whereas that
> unless I misread it is for the entire CL competition.
You're right. Daniele's proposal has the pro that you haven't to
choose who are the big teams and who aren't: whatever team reaches the
later stages of UCL is good, and the further they go the better they
are. And, about players: the more they play the better they are.
> > As I've written, I have a slight preference for the integration of
> > these new ideas in the old Fantasy Scout (I think that it's possible:
> > let's switch to yearly rankings), because I don't like to have a
> > player forever on my side in a game and against me in another one.
>
> Same here.
So you agree that we should try to integrate your and Daniele's ideas
in Fantasy Scout?
--
Cheers
milivella
milivella
> On Jun 22, 7:01 pm, milivella <milive...@gmail.com> wrote:
>
> > > 1. Let's identify the PROBLEMS of the game
>
> First, the gamemaster has too many ideas, writes too many posts and
> follows up on them on his own, and it's difficult to keep track of
> him :-)
I know that adding a reply in this context is not wise :) but there's
a little thing that I _have_ to say:
I am *not* the gamemaster.
I am the guy who wrote the first version of the rules. But they've
been made far better than they were by you all. Just read the first
version and compare with the actual one:
http://groups.google.com/group/rec.sport.soccer/browse_thread/thread/93dda93dc1d54914
I am the guy who keeps track of the picks and of the scores. But
anyone can do it; even if, for some time, nobody updates the score,
the game can go on.
So the game is (and, given its "eternal" horizon, must be)
decentralized.
--
Cheers
milivella
milivella
> A bit like in
> NTFS that scouts pick players on the fringe of the national team, and
> in positions of relative weakness of their NT.
(This reply came to my mind watching Riera on the field...) This one
of the reasons why I'm proposing 1. one player per year 2. one first
place in the yearly rankings counts more than a lot of mid-table
placements. In fact, with such rules you'll be pushed to pick
eventually great players, not surely good ones.
--
Cheers
milivella
milivella
> Another thought, for a different kind of scouting game.
> Instead of only awarding points for international appearances,
> also award points for player signings by big clubs.
>
> First, make a set of big clubs ( EPL big 4, LL big 2, SA big 3
> should do for the moment ).
On a thread in BigSoccer you can find who are the best players that
don't play for any of your teams (or Bayern):
http://www.bigsoccer.com/forum/showthread.php?t=1064672
--
Cheers
milivella
Enzo
>
> > > - Where do you can find 100% complete data about U-21 caps? I don't
> > > know.
>
> > Not sure. I can find out if needed.
>
> It would be useful if you want to make it an official proposal.
Sratch U-21. My proposal is U-20. That should cover it.
milivella
> Sratch U-21. My proposal is U-20. That should cover it.
Would a European player that has caps only in the U-19 team be
eligible? (the teams for the U-20 WC are basically the same of the
previous year's U-19 Euro)
--
Cheers
milivella
Enzo
Any player would be eligible as long as he does not have a under-20
cap. He could be 21, which means he will never again have a under-20
cap. The age limit of course is a different issue, I think there
shouldnt be one. If you want to pick an older player, go ahead.
milivella
> On Jun 29, 12:53 pm, milivella <milive...@gmail.com> wrote:
>
> > Enzo:
>
> > > Sratch U-21. My proposal is U-20. That should cover it.
>
> > Would a European player that has caps only in the U-19 team be
> > eligible? (the teams for the U-20 WC are basically the same of the
> > previous year's U-19 Euro)
>
> Any player would be eligible as long as he does not have a under-20
> cap.
Perfect. We only need sources for U-20 caps: national federations'
site should be OK, but maybe it's not always the case.
> He could be 21, which means he will never again have a under-20
> cap. The age limit of course is a different issue, I think there
> shouldnt be one. If you want to pick an older player, go ahead.
Fine.
--
Cheers
milivella
Enzo
> Enzo:
>
> > On Jun 29, 12:53 pm, milivella <milive...@gmail.com> wrote:
>
> > > Enzo:
>
> > > > Sratch U-21. My proposal is U-20. That should cover it.
>
> > > Would a European player that has caps only in the U-19 team be
> > > eligible? (the teams for the U-20 WC are basically the same of the
> > > previous year's U-19 Euro)
>
> > Any player would be eligible as long as he does not have a under-20
> > cap.
>
> Perfect. We only need sources for U-20 caps: national federations'
> site should be OK, but maybe it's not always the case.
>
Needless to say, he should not have full international caps either.
It could be made slightly easier, say max 2 U-20 caps.