ACC Offense/Defense Ratings (All Games Through 12/02/01)
George W. Harris
series of posts assessing the offensive &
defensive performance of ACC teams in all
games. Obviously, meaningful comparisons
can be difficult, since some teamshave played
challenging non-conference schedules
and other teams have not.
First is the rating of team offensive
performance, from best (highest) to worst
(lowest); then the rating of team defensive
performance, from best (lowest) to worst
(highest); and finally, a tally of each team's
tendencies, from highest offense to highest
defense. The basic form of these ratings is
[points/(def rebounds + turnovers)]
with the obvious team's number for each
category.
Offense
Rank)Team Rating
Current Previous
1)Virginia 2.397 1)2.675
2)Duke 2.298 3)2.367
3)Clemson 2.263 4)2.333
4)Wake 2.171 6)2.084
5)NCState 2.076 2)2.497
6)Maryland 2.074 5)2.161
7)FSU 1.885 7)2.009
8)Ga Tech 1.779 8)1.752
9)UNC 1.572 9)1.367
Average 2.058 (2.142)
Defense
Rank)Team Rating
Current Previous
1)Virginia 1.277 2)1.278
2)FSU 1.308 1)1.217
3)NCState 1.463 3)1.293
4)Duke 1.464 4)1.403
5)Maryland 1.504 5)1.521
6)Clemson 1.799 9)1.884
7)Ga Tech 1.825 7)1.835
8)Wake 1.937 8)1.847
9)UNC 2.035 6)1.824
Average 1.619 (1.557)
Tendencies
Offense Rating
^ Current Previous Change
|
Wake 2.051 1.962 +1
Clemson 2.016 2.095 -1
Duke 1.839 1.822 +1
Ga Tech 1.802 1.794 +3
UNC 1.782 1.573 +4
Maryland 1.771 1.818 -1
Virginia 1.760 1.864 -4
NCState 1.749 1.807 -2
FSU 1.575 1.575 -1
|
v
Defense
Average 1.827 (1.828)
--
Check out all my basketball analysis at http://communityacupuncture.com/Basketball/
George W. Harris For actual email address, replace each 'u' with an 'i'.
Edward M. Kennedy
> [points/(def rebounds + turnovers)]
In terms of offense:
You're trying to get a gauge of points per possession?
Who is to say an offensive rebound is a bad thing, or
even a new possession? The other team never got the
ball. Plus, missing a shot while being double teamed
down low frequently results in a put-back.
Offensive rebounds do relect less efficiency in a way,
but they are still a good thing and a quicker way to
score than running back to play defense, get the ball,
and come back to try again.
In terms of defense:
You're a Duke fan, right? Offensive rebounds is the
price Duke pays for man-to-man help defense.
To mitigate the damage of rebounding to a rating, how
about:
[(points + rebounds)/(rebounds + turnovers)]
It fails in the case where you hardly score and don't
turn the ball over either, but I find that hard to
believe.
--Tedward
George W. Harris
<T...@tkennedy.mc.duke.edu> wrote:
:"George W. Harris" wrote:
:
:> [points/(def rebounds + turnovers)]
:
:In terms of offense:
:
:You're trying to get a gauge of points per possession?
:Who is to say an offensive rebound is a bad thing, or
:even a new possession?
If you take a shot, miss it, get the offensive
rebound, and make a shot, that is functionally equivalent
to making the first shot. Basically, missing a shot and
getting the offensive rebound is a push.
:The other team never got the
:ball. Plus, missing a shot while being double teamed
:down low frequently results in a put-back.
If it does, you get credit for the points. Just as if
you'd made the first shot.
:
:Offensive rebounds do relect less efficiency in a way,
:but they are still a good thing and a quicker way to
:score than running back to play defense, get the ball,
:and come back to try again.
An offensive rebound is a good thing that is
dependent upon a bad thing; missing a shot. In effect,
you're getting credit for the offensive rebound because
the other team *doesn't* get the defensive rebound. If
you don't score after getting the offensive rebound, it
didn't do you any good. So why not eliminate the middle
man and just count points?
:
:In terms of defense:
:
:You're a Duke fan, right? Offensive rebounds is the
:price Duke pays for man-to-man help defense.
:
:To mitigate the damage of rebounding to a rating, how
:about:
:
:[(points + rebounds)/(rebounds + turnovers)]
What's the justification for that? The reasoning
behind the rating is very simple. On one hand, you have
possessions that end well (points). On the other hand,
you have possessions that end badly (turnovers or
opposition defensive rebounds). Simply compare the
two.
:It fails in the case where you hardly score and don't
:turn the ball over either, but I find that hard to
:believe.
It is in no way an improvement.
:
:--Tedward
Edward M. Kennedy
>
> On Mon, 03 Dec 2001 18:21:43 -0500, "Edward M. Kennedy"
> <T...@tkennedy.mc.duke.edu> wrote:
>
> :"George W. Harris" wrote:
> :
> :> [points/(def rebounds + turnovers)]
> :
> :In terms of offense:
> :
> :You're trying to get a gauge of points per possession?
> :Who is to say an offensive rebound is a bad thing, or
> :even a new possession?
I'm not even sure what you're after yet. If you want
some pure efficiency in terms of possessions, I grok
what you are doing, but don't necessarily agree it is
the most useful thing to do. On the other hand, you use
phrases like "performance" and "ratings", and I don't
think offensive efficiency reflects offensive performance.
Princeton might have high efficiency and low offensive
output, while MSU might be the opposite. Teams that
run and gun can get a lower efficiency to pay off.
> If you take a shot, miss it, get the offensive
> rebound, and make a shot, that is functionally equivalent
> to making the first shot. Basically, missing a shot and
> getting the offensive rebound is a push.
It's a push in terms of efficiency. It's not a push in
terms of all the good things that happen while you have
possession (and the other team doesn't). Since you cannot
easily factor in things like 1 of 2 foul shots, looking
at other readily available stats may help. BTW, would
it be hard to factor in treys? I would think subtracting
a half point from the denominator for each three would
make sense. Adding a half point for each missed free
throw makes sense if you could seperate out the bonus free
throws, plus the front end of a one-and-one remains elusive
(I suspect most teams pencil in "0-2" in their private
databases). Still, I think there is enough information
in the free throws that I'd include them rather than
ignore them.
> :The other team never got the
> :ball. Plus, missing a shot while being double teamed
> :down low frequently results in a put-back.
>
> If it does, you get credit for the points. Just as if
> you'd made the first shot.
Right, but virtually all the teams are shooting between a
range of about 10 percentage points (I'd guess 38-48), and
50% for a team is very, very, very good. Offensive rebounds
are important; given identical efficiency, I'd bet on the
team that rebounds better because they are better able
to overcome a bad night or good defenese, and will rack up
more fouls on the other team.
A glance at shooting percentages and turnovers gives pretty
much the same information as efficiency. It doesn't have
the benefit of taking into account the number of possessions
(though you could), but it doesn't ignore good three point
shooting or bad free throw shooting.
Hell, don't use box scores. Use the team stats on Yahoo
or some such.
--Tedward
George W. Harris
<T...@tkennedy.mc.duke.edu> wrote:
:"George W. Harris" wrote:
:>
:> On Mon, 03 Dec 2001 18:21:43 -0500, "Edward M. Kennedy"
:> <T...@tkennedy.mc.duke.edu> wrote:
:>
:> :"George W. Harris" wrote:
:> :
:> :> [points/(def rebounds + turnovers)]
:> :
:> :In terms of offense:
:> :
:> :You're trying to get a gauge of points per possession?
:> :Who is to say an offensive rebound is a bad thing, or
:> :even a new possession?
:
:I'm not even sure what you're after yet.
These ratings actually grew out of a discussion
about how to judge *defensive* performance. Just
looking at any one of shooting %age, scoring average,
defensive rebounding and turnovers does not suffice
because alone the meaning of any one of these can be
influenced by other factors (such as pace or the other
aspects of defense). I wanted to rate defenses with stats
that can be gleaned from any box score, and to balance
what a defense seeks to prevent (points) against what a
defense seeks to cause (an opponent possession that
ends without scoring). The next step is obvious.
: If you want
:some pure efficiency in terms of possessions, I grok
:what you are doing, but don't necessarily agree it is
:the most useful thing to do. On the other hand, you use
:phrases like "performance" and "ratings", and I don't
:think offensive efficiency reflects offensive performance.
:
:Princeton might have high efficiency and low offensive
:output, while MSU might be the opposite. Teams that
:run and gun can get a lower efficiency to pay off.
No, they can't. Every time Team A's possession
ends, Team B's possession begins. So ignoring half-
ending possessions, for every possession you have, your
opponent is going to have one as well. So while running
and gunning will get you more possessions, it will give
your opponent more possessions as well.
What teams that run and gun are trying to do is
to get a shot before the opponent can get their defense
set up. This makes it more likely that you'll score on
each possession, and thus increase your offensive
efficiency.
:
:> If you take a shot, miss it, get the offensive
:> rebound, and make a shot, that is functionally equivalent
:> to making the first shot. Basically, missing a shot and
:> getting the offensive rebound is a push.
:
:It's a push in terms of efficiency. It's not a push in
:terms of all the good things that happen while you have
:possession (and the other team doesn't).
If you have possession, it generally doesn't
matter how long your possession lasts if the number of
points you score is the same.
:Since you cannot
:easily factor in things like 1 of 2 foul shots, looking
:at other readily available stats may help. BTW, would
:it be hard to factor in treys? I would think subtracting
:a half point from the denominator for each three would
:make sense. Adding a half point for each missed free
:throw makes sense if you could seperate out the bonus free
:throws, plus the front end of a one-and-one remains elusive
:(I suspect most teams pencil in "0-2" in their private
:databases). Still, I think there is enough information
:in the free throws that I'd include them rather than
:ignore them.
The top is points scored; that includes treys
and made free-throws. The bottom is *failed*
possessions. That's relatively easy to figure (although
you get some problem when a team makes the first FT
and misses the second). You can try to figure *total*
possessions by including total FGs and some multiple of
FTs, but then you introduce other levels of inaccuracy,
so you gain very little in terms of accuracy for a sacrifice
of simplicity.
:> :The other team never got the
:> :ball. Plus, missing a shot while being double teamed
:> :down low frequently results in a put-back.
:>
:> If it does, you get credit for the points. Just as if
:> you'd made the first shot.
:
:Right, but virtually all the teams are shooting between a
:range of about 10 percentage points (I'd guess 38-48), and
:50% for a team is very, very, very good. Offensive rebounds
:are important; given identical efficiency, I'd bet on the
:team that rebounds better because they are better able
:to overcome a bad night or good defenese, and will rack up
:more fouls on the other team.
If two teams have identical offensive efficiency,
and team A is better at rebounding, then perforce they
must be worse shooters or turn the ball over more. It
doesn't matter how many times you rebound a missed
shot if the end result is the same. If you rack up more
fouls on the opposition, you'll score more on FTs which
will help your efficiency.
:A glance at shooting percentages and turnovers gives pretty
:much the same information as efficiency. It doesn't have
:the benefit of taking into account the number of possessions
:(though you could), but it doesn't ignore good three point
:shooting or bad free throw shooting.
Neither does efficiency. Besides, a team can
have a lower shooting percentage and make up for it
with offensive rebounds (reducing the opposition's
defensive rebounds). Just looking at shooting
percentage & TOs ignores this.
Edward M. Kennedy
> : BTW, would
> :it be hard to factor in treys? [snip]
>
> The top is points scored; that includes treys
> and made free-throws. The bottom is *failed*
> possessions.
At the end of regulation, the score is tied, and neither
team got an offensive rebound or a free throw:
Team A: 20-50 treys, no deuces, 10 TO
off. rating = 60/(30+10) = 1.5
Team B: 30-60 deuces, no treys, 0 TO
off. rating = 60/(30+0) = 2.0
The score is tied, yet one team has a significantly better
rating despite the same number of possessions. Unless I
am missing something, this system does not handle the
different effect of 2/3 point shots well.
You say, team A had 10 more failed possessions. I say,
so what, they made made the same number of points in
the same number of possessions. You say, well, I don't
know how many *real* possessions they had. I agree,
but factoring in the increased value of the three seems
self evident.
--Tedward
Edward M. Kennedy
> :Princeton might have high efficiency and low offensive
> :output, while MSU might be the opposite. Teams that
> :run and gun can get a lower efficiency to pay off.
>
> No, they can't. Every time Team A's possession
> ends, Team B's possession begins. So ignoring half-
> ending possessions, for every possession you have, your
> opponent is going to have one as well. So while running
> and gunning will get you more possessions, it will give
> your opponent more possessions as well.
>
> What teams that run and gun are trying to do is
> to get a shot before the opponent can get their defense
> set up. This makes it more likely that you'll score on
> each possession, and thus increase your offensive
> efficiency.
Run and gun teams tend to shoot more threes, plus
press the other team and try to sucker the opponent
into their style. The trey distorts things, and
the latter two tend to mitigate the "equal possessions"
factor. Not to mention conditioning...
> :> If you take a shot, miss it, get the offensive
> :> rebound, and make a shot, that is functionally equivalent
> :> to making the first shot. Basically, missing a shot and
> :> getting the offensive rebound is a push.
> :
> :It's a push in terms of efficiency. It's not a push in
> :terms of all the good things that happen while you have
> :possession (and the other team doesn't).
>
> If you have possession, it generally doesn't
> matter how long your possession lasts if the number of
> points you score is the same.
It does matter -- D is tougher to play, and results
in more fouls. Both result in softer D later on, and
the latter disqualifies people or brings on lesser
quality players for a time.
Sorry to inject some basketball reality into the purely
statistical aspect...
--Tedward
George W. Harris
<T...@tkennedy.mc.duke.edu> wrote:
:"George W. Harris" wrote:
:
:> :It's a push in terms of efficiency. It's not a push in
:> :terms of all the good things that happen while you have
:> :possession (and the other team doesn't).
:>
:> If you have possession, it generally doesn't
:> matter how long your possession lasts if the number of
:> points you score is the same.
:
:It does matter -- D is tougher to play, and results
:in more fouls. Both result in softer D later on, and
:the latter disqualifies people or brings on lesser
:quality players for a time.
If this is true, it will result in a higher
offensive efficiency later in the game, and so
will be reflected in the numbers.
Otto Bahn
>
> On Fri, 07 Dec 2001 13:38:51 -0500, "Edward M. Kennedy"
> <T...@tkennedy.mc.duke.edu> wrote:
>
> :"George W. Harris" wrote:
> :
> :> :It's a push in terms of efficiency. It's not a push in
> :> :terms of all the good things that happen while you have
> :> :possession (and the other team doesn't).
> :>
> :> If you have possession, it generally doesn't
> :> matter how long your possession lasts if the number of
> :> points you score is the same.
> :
> :It does matter -- D is tougher to play, and results
> :in more fouls. Both result in softer D later on, and
> :the latter disqualifies people or brings on lesser
> :quality players for a time.
>
> If this is true, it will result in a higher
> offensive efficiency later in the game, and so
> will be reflected in the numbers.
True. I'd just reward it all the time.
--Tedward
Edward M. Kennedy
>
> On Fri, 07 Dec 2001 13:38:51 -0500, "Edward M. Kennedy"
> <T...@tkennedy.mc.duke.edu> wrote:
>
> :"George W. Harris" wrote:
> :
> :> :It's a push in terms of efficiency. It's not a push in
> :> :terms of all the good things that happen while you have
> :> :possession (and the other team doesn't).
> :>
> :> If you have possession, it generally doesn't
> :> matter how long your possession lasts if the number of
> :> points you score is the same.
> :
> :It does matter -- D is tougher to play, and results
> :in more fouls. Both result in softer D later on, and
> :the latter disqualifies people or brings on lesser
> :quality players for a time.
>
> If this is true, it will result in a higher
> offensive efficiency later in the game, and so
> will be reflected in the numbers.
True. I'd just reward it all the time.
--Tedward
George W. Harris
<T...@tkennedy.mc.duke.edu> wrote:
:George W. Harris wrote:
:>
:> On Fri, 07 Dec 2001 13:38:51 -0500, "Edward M. Kennedy"
:> <T...@tkennedy.mc.duke.edu> wrote:
:>
:> :"George W. Harris" wrote:
:> :
:> :> If you have possession, it generally doesn't
:> :> matter how long your possession lasts if the number of
:> :> points you score is the same.
:> :
:> :It does matter -- D is tougher to play, and results
:> :in more fouls. Both result in softer D later on, and
:> :the latter disqualifies people or brings on lesser
:> :quality players for a time.
:>
:> If this is true, it will result in a higher
:> offensive efficiency later in the game, and so
:> will be reflected in the numbers.
:
:True. I'd just reward it all the time.
Why reward it in a game where it provides
no advantage?
Edward M. Kennedy
"George W. Harris" wrote:
>
> On Sun, 09 Dec 2001 17:46:11 -0500, "Edward M. Kennedy"
> <T...@tkennedy.mc.duke.edu> wrote:
>
> :George W. Harris wrote:
> :>
> :> On Fri, 07 Dec 2001 13:38:51 -0500, "Edward M. Kennedy"
> :> <T...@tkennedy.mc.duke.edu> wrote:
> :>
> :> :"George W. Harris" wrote:
> :> :
> :> :> If you have possession, it generally doesn't
> :> :> matter how long your possession lasts if the number of
> :> :> points you score is the same.
> :> :
> :> :It does matter -- D is tougher to play, and results
> :> :in more fouls. Both result in softer D later on, and
> :> :the latter disqualifies people or brings on lesser
> :> :quality players for a time.
> :>
> :> If this is true, it will result in a higher
> :> offensive efficiency later in the game, and so
> :> will be reflected in the numbers.
> :
> :True. I'd just reward it all the time.
>
> Why reward it in a game where it provides
> no advantage?
_That_ is where we disagree.
--Tedward