APBRmetrics

[DLew]

Steve,

Since you are new to this board I just want to warn you that Bob Chaikin can at times be very difficult to have a discussion with, especially when he already has his mind made up. He sometimes brings valuable insight to the table, but often answers questions with questions.

[Ilardi]

David,

Good to hear from you! Hey, do you have anything to add regarding the earlier discussion of statistically-augmented adjusted plus-minus? Have you and/or Dan R. ever tried augmenting with non-boxscore stats like eFG% and iFG% allowed, on-off defensive ratings, etc.?

I'm also curious to know what specific method you guys have used to come up with your standard errors of estimate for the hybridized (statistically augmented) adjusted plus-minus ratings. Can you say anything about that without violating your contract with the Cavs?

[Mike G]

[Mountain]

We have various indicators of Chris Paul's average defensive results but not the true individual marginal defense or a good sense of how vulnerable he is to further exploit.

Paul's shot defense is weak ,,, so it would make sense for teams to try to exploit it... but I noted that there were other spots in New Orleans' defense as ripe or riper for attack. Maybe that helps keep attacks on Paul under check ... and maybe Paul (and Scott) also deserves credit for keeping his opponents shots against him in check- a different element of defense.

Looking at 82 games season data Paul's FGAs against him have been at league position defense median every year of his career. And in the playoffs attempts were down by over 10% and in line with a lowest quartile team level.

Ball denial or "ball steering" to other spots by the way he plays defense maybe helping keep the impact of his defensive weakness in check. And Chandler sitting back there. (Looks like team pts allowed with Paul but without Chandler is 5 pts higher than with.)

The weakest defensive players are those where the defense that can't limit the bleeding- has a high marginal FG% and pts allowed and can't limit frequencies of attack. Paul / style of defense / Chandler works well enough. Congrats to Jeff Bower on good team construction.

[bchaikin]

Since you are new to this board I just want to warn you that Bob Chaikin can at times be very difficult to have a discussion with, especially when he already has his mind made up.

as can people who feel the need to make statements like:

...I just want to make sure I know where you stand.

this point was stated about chris paul:

Put simply: his superb offensive contributions have been largely offset by his apparent defensive liabilities.

and this:

his large negative impact on points allowed per possession indicate Paul is almost certainly a defensive liability.

as if this was an open and shut case. then this statement was made:

Nash is still an elite point guard.... makes him the highest-rated point guard....

higher than paul or chauncey billups? then this statement was made:

There is even some evidence Nash is not as bad on the defensive side of the ball as widely believed. For example, opposing PGs had a eFG% of 49.8% when he was on court - just above the league average of 49% and well below Chris Paul's whopping 52% eFG% allowed.

yet the difference in defensive stops due to a 49.8 eFG% allowed and 52% eFG% allowed is much much smaller than the difference in defensive stops due to the steals these two players got...

my question here is this - are you trying to imply that nash was a better defensive player overall than paul in 07-08? on the one hand you are saying nash is not as bad as some may think, but on the other hand are definitively calling a paul a defensive liability, and stating his bad defense largely offset his offensive contributions, when his offensive contributions were very very good. new orleans was 8th in the league in defense, phoenix 16th. are you saying if nash and paul swapped teams and played similar minutes new orleans would have been a better defensive team with nash, and phoenix a worse defensive team with paul?...

finally, you presents a tangible number:

he would be singlehandedly adding +2.8 points per game to the Hornets' bottom line (above the value of a league-average replacement player). Using the "Pythagorean Wins" formula would indicate that this is worth roughly 8.2 extra wins per year.

unfortunately i don't know what an "extra" win is. looking at a plot of final regular season team wins vs team average per game point differential, a +2.8 pts/g point differential is the per game point differential of a team that has, historically speaking, gone about 48-34 to 49-33, or to be more exact about 7.3 wins above a 41-41 record, which is about 1 win less than your stated pythagorean wins, whatever those are...

and i have no idea what a league average replacement player is. if a team has a roster of all league average replacement players, is that a team with a 41-41 record, as that is a league average W-L record?...

i am trying to get a handle on how many wins are generated by the best vs worst PGs by your methodology playing similar minutes. are you saying steve nash generated more wins than chris paul or chauncey billups (playing similar minutes)? what is the difference in "extra" wins between some of the best and worst PGs based on your methodology? between nash and paul? nash and billups? nash and telfair?..

i tend to see a 14-16 wins between the best and worst starters in a typical season (on a 40 min/g and 82 game basis), however this season with paul and billups certainly better than nash, by upwards of 5-6 wins on a 40 min/g and 82 game basis...

Regardless, Bowen's on-off defensive rating (unadjusted) looks reasonably good (compared with that of his most common floormates), whereas his offensive on-off looks poor . . . so I certainly wouldn't want to argue that he's a POOR defender, though I have suspected at times this season that he's no longer the phenomenal lock-down defender he was 3-4 years ago.

is this what your methodology is telling you? i ask as the spurs have been one of the league's top 3 defensive teams each of the past 5 seasons, and bowen has played the 1st-3rd most minutes on the team each season (and not less than 2400 minutes in any season), on nightly basis has guarded the opponents' best SG/SF, and on occasion PGs and PFs, and has been all-D 1st team the last 5 years...

[Mike G]

If facts are described such that the statistical method seems to fit them, is that 'the tail wagging the dog'?
_________________
`
36% of all statistics are wrong

[Ilardi]

[Django]

I really don't understand all of this statistical stuff,but as someone who has actually watched most of the Hornets games I'm somewhat confused by the repeated claim that Jannero Pargo is a good defender or a better defender than Paul.Pargo is probably one of the worst defenders in the NBA and as a player he's utterly useless when his shot isn't falling.

In fact I'm relatively sure that a lot of this discussion is a result of conflating Pargo's defense with Paul's.Before Pargo got hot and began to actually contribute I looked at 82games.com and Paul's defensive PER was around 14.5 or something.His differential PER was in the top five in the NBA because he was producing at better than twice the rate of opposing PG's.At that point Pargo was basically his backup and the Hornets hadn't yet given up on Morris Peterson.After Pargo began to produce he took much of Peterson's minutes,Pargo and Paul would probably play around 15 to 20 minutes per night together.Paul always guarded the stronger player because Pargo is completely incapable of dealing with larger players.That is to say that Pargo nearly always guarded the opposing point guard.Immediately Paul's defensive PER began to skyrocket.If you watched the games you'd see that teams go right at Pargo at every oppurtunity and that he offers very little resistance.

[Mountain]

82games lineup ordering is imperfect and sometimes way off. Thanks for sharing your knowledge of it in this case.

Paul / Pargo in same lineup is worse than either alone. A cost of playing twin small PGs together.

82games has Pargo at pretty close to the same pts and PER allowed at either position and better than Paul's numbers. Mis-accounting when Paul and Pargo are on court together might be helping Paul. Taking the tougher cover is worth noting but what he does there is important too. It could probably be tracked down and summed up from basketballvalue play by play for when they are together if anyone cared enough.


Pargo alone has better raw team defensive marks than Paul alone. And for all minutes- this season by 4 pts, last season by 6.

And every player on the Hornets has a worse team defense when on court with Paul than not, except Butler who was even. Worse with Paul for every player in 06-07 as well. Now Paul on court probably is facing a higher % of starters and that plays into the opponent performance level but this worse with Paul pattern is pretty darn strong. Worse with Paul in about 2/3rd of cases in 05-06 too with all the exceptions being small minute cases.

Still getting the adjusted offensive / defensive splits will be a welcome next stage.

[tenkev]

[Mountain]

Sure, 3 way split is more detail.

[Mountain]

According to Eli's adjusted +/- splits Paul scores the 3rd weakest defensive +/- of all minutes qualified players at any position across the league and is the worst PG on the measure,