Kyle Weidie – at Truth About It – has taken issue with my discussion this past summer of John Wall. (you have to scroll down a bit to Weidie’s discussion of John Wall and stats):
It’s hard to talk about John Wall’s stats without mentioning Dave Berri. Berri, of The Wages of Wins Journal and the books “Wages of Wins” and “Stumbling On Wins”, is not a fan of Wall. Or rather, he’s not a fan of Wall’s stats and thus, he is overall indignantly unimpressed with the number one overall draft pick — so much so that Berri took the time this past summer to write a post, “Are We Allowed to Say that John Wall Has Yet to Produce?,” criticizing a Las Vegas Summer League recap I’d written about Wall for ESPN’s Daily Dime, essentially because I didn’t mention Wall’s faults until the ninth paragraph. Perhaps stats on other people’s writing are the new frontier. Either way … good one, Dave.
Weidie’s attack doesn’t get much better after this beginning. And here is how it ends:
Mr. Berri really doesn’t have a calculator to stand on. But in this country, irrational outlandishness sells. And when someone combines intelligence with a pompous nature like Berri, he might as well be smugly selling snake oil.
In between these two points we see that Weidie essentially failed to understand the actual argument I was making in the following three posts:
Is John Wall the Obvious Number One Pick for the Wizards?
John Wall, Derrick Rose, and the State of the Wizards
Are We Allowed to Say that John Wall Has Yet to Produce?
Let me clarify by just repeating what I said last summer:
So at this point we can say this about Wall. His physical skills suggest that he has a great deal of potential. But so far, that potential hasn’t translated into actual production. Wall was not particularly productive in college. And he wasn’t productive in summer league.
Now it’s very important to emphasize what I am saying. I am not saying – and I repeat, I am not saying – that Wall will never be a great basketball player. What I am saying is that in college and summer league he was not a great basketball player (again, I am differentiating what Wall has done from what he might do in the future).
And I am trying to emphasize that ignoring missed shots and turnovers for eight paragraphs paints a misleading picture of Wall’s actual performance. Missed shots and turnovers really matter in basketball.
The person who ignored Wall’s shortcomings for eight paragraphs was Weidie. Again, I think this is a story that deserved a bit more coverage. Wall is supposed to be a great basketball player. So far, though, he hasn’t been great. Again – and I am saying this again (maybe Weidie can understand this point if I say it enough) – I am not arguing that Wall will never be a great player. I am just arguing that so far he has not been as good as advertised.
And launching personal attacks at me is not going to change this basic observation.
– DJ
Wages of Wins Journal 
Italian Stallion
October 3, 2010
When people hear the same things from supposed experts in the media over and over again, I think it’s virtually impossible for them to reexamine what they think they know, admit they were wrong, and then relearn everything.
That’s seems to be true in investing, economics, basketball, horse racing and just about everything else I’ve ever taken a serious interest in.
I think it’s obvious you are correct about Wall.
Schermeister
October 3, 2010
The obsession with John Wall continues. Like numerous other 1st picks that turned out to be merely average or worse then average players. He hasnt done anything nor proven anything.
jimbo
October 3, 2010
I think the problem most folks have is that your statement is NEVER that “John Wall was not a productive player in college, per my metrics.” Instead you always say that “John Wall was not a productive player in college, period.”
There are other metrics out there that produce differing results to yours and your unwillingness to acknowledge them and put forth yours as the best metric is what rankles folks…
robbieomalley
October 3, 2010
Jimbo,
No, they have a problem with him contradicting their notions. No one ever says jack about the entire ESPN team only ever using PER because it’s a scoring based metric. Why would Berri use someone elses metric if his has been proven to be better? Models can be tested you know. It’s not this is my number so it’s better.
Michael
October 3, 2010
This is a total non-issue. Weidie may not have mentioned Wall’s ‘shortcomings’ fast enough for you, (apparently such things must be mentioned within 8 paragraphs), but he did mention them:
“Turnovers also have been an issue. Wall surely doesn’t want to be like his childhood idol, Allen Iverson, and give the ball away at a rate of 4.4 times per game during his rookie season. Turnovers came up when Wall was asked to grade himself. “Probably a ‘B’ average,” Wall said. “The last two games I really turned my turnovers down, and that was the biggest plus for me.”
Wall averaged 9.4 turnovers per 36 minutes during the first two games and 2.6 per 36 in the last two. Once Wall adjusts his speed to the pace of the NBA game, he should be able to reduce his mistakes and improve his decision-making. Getting better accustomed to the more physical defenders he will see in the league is also a must.”
jbrett
October 3, 2010
John Wall has an excellent rebuttal at his disposal; all he has to do is score enough points. There may be some who look at his assists, so he can help himself there, as well, but mainly he needs to show he can shoot enough to score enough to ‘prove’ the doubters wrong. Lucky for him, it doesn’t matter how few games the Wiz win; other players will get the blame–probably those whose scoring declines. In the meantime, the nimrod offering this piss-poor defense of him will certainly feel vindicated, because some people never change, and idiots least of all.
nerdnumbers
October 3, 2010
I think 30 Rock has gotten in on the Dave Berri bashing (it’s a spoken quote so I can change the spelling.)
“You want to make god laugh? Make a plan. . . or read him a Dave Berri book.”
– Tracy Jordan, 30 Rock Season 5 Episode 2 – ‘When it Rains, It Pours’
Also if quoting the stats, saying it does not agree with conventional wisdom and then admitting their potential shortcomings is snake oil, put me down for 2.
arturogalletti
October 3, 2010
nerdnumbers,
Probably Dave Barry.
PER is by the way an ESPN exclusive metric from an ESPN employee.
I read the original piece and it painted a misleading picture of Wall’s performance. The stats did not match perception (and the piece pushed perception). I’m on record as to how I feel about the pick and the Wizards chances. Wall is young and athletic and can grow to be a good player. He isn’t one now and to push that as truth is disingenuous.
fricktho
October 3, 2010
Wall isn’t a bad player, but his production doesn’t say he’s a star is basically the jist correct? He still does a lot of positive things on the court. Creates assists and steals, and rebounds well. He’ll get to the FT line and scores efficiently enough from it. The turnovers are a production killer as is his shooting efficiency. I would guess a season much like Brandon Jennings rookie season, but with more points, shots, rebounds, assists and most importantly turnovers. So basically he’s Jennings on speed or something. .067 wp/48.
Blue
October 3, 2010
College stats are useful in projecting NBA stats on average but many individual cases will under and over perform significantly. It is a very different context including because of the rules. Strength of opponents faced varies among teams and even among positions (I’d guess fewer bigs in college are relatively near NBA quality than for perimeter positions).
Accounting for age and years of experience in college in reviewing those college stats and projecting NBA stats would likely be an enhancement over a simple approach where that is not done.
Similarity systems used to project NBA performance probably are helpful compared to not considering them but they are still just a better informed baseline guess. Players are unique and can follow different paths in ways that I am not sure stats alone can fully pick up.
Summer league stats may be somewhat interesting and possibly useful if they are really good or really bad but probably not a very precise basis for general projection.
EntityAbyss
October 3, 2010
John Wall shoots a low percentage and turns over the ball a lot. These are big things in winning games. He doesn’t rebound too exceptionally or steal the ball a lot, and I expect him to have a rookie season worse than derrick rose did and derrick rose is overrated too. If he’s the derrick rose “clone”, derrick rose produced better than him in college, and derrick is overrated too, it’s hard to think he’ll do well immediately.
dm
October 3, 2010
Oh oily snakesman David Berri, could you be anymore irrational? Everyone knows you propagate these views for the fame and money. I’ve seen you cruising around Utah in your red hot convertible jingling your lavish jewelry at the homeless and laughing heartily.
Don’t you know it’s the preseason, a time for optimism, a time for hope, a time to turn over a new leaf and reserve front row tickets for the playoffs.
The off season is a time for puff pieces, that’s what we the people want to read. NBA bottom feeders are undefeated, they are no longer “bad” teams they are young and promising. With proper coaching and the right chemistry things might “click” and then anything is possible!
So you just keep your blue pills to yourself Morpheus, cause we don’t want to hear it! Not only that if I even hear your telling a different tale i’ll attack you personally.
Is that clear mister?
Robbo
October 3, 2010
Good comparison fricktho. The issue is Jennings was taken with the 10th pick, and Wall the 1st. If Wall was taken lower, we wouldn’t be having this conversation.
Robbo
October 3, 2010
Guys, you’ll be interested in Cuban’s latest comments about using stats to build a team:
http://blogmaverick.com/2010/10/03/building-teams-using-quantitative-analysis-in-the-nba-the-last-rites/?utm_source=feedburner&utm_medium=twitter&utm_campaign=Feed%3A+blogmaverick%2FtyiP+%28blog+maverick%29
notherbert
October 3, 2010
draft projections are a bit shaky and the number one pick is highly amplified, thus whirling words will be.
i just did a quick glance on DraftExpress where they list WS40 for the last few years. i refined the search to freshman point guards and JohnWall comes up with a WS40 of 7.2. looking back through the 6 years of stats they have listed, there are not that many players who in their freshman year, at point guard, posted a WS40 of 7 or more, 2-3 a year, sometimes none rated at a 7 or higher with a WS40 of 10 being the highest (StephCurry and GeorgeHill).
glancing at all that, Wall doesn’t look like a sure fire star but he does look like he’s in the top group of players to excel in the WS40 metric as freshman playing the point in the last 6 years.
attempting to project a 19 year old point guard into the NBA feels much like guessing the density of clouds as you lay on the ground. it definitely helps to know the metrics used in measuring cloud density but it would be better if you had a mechanical device to just shove into the cloud and take a measurement.
Wall is probably the wrong #1 pick to go touting his greatness and any freshman PG is probably best viewed with skepticism concerning their future stardom. i’d say Mr. Berri wins out with the more appropriate tone.
Robbo
October 3, 2010
How’s this for inflammatory:
“Yes, there are stats that are out there that could be used to better build an NBA team, but No, they can’t be used for building an NBA team, because the stats that most likely most correlate to a player and team’s success are not being collected.
I’m not here to say I know exactly which variables independently, or collectively equate to getting a competitive and financial advantage. Only actual testing will determine what works.”
marparker
October 4, 2010
The real problem is that DB’s method is boring. The fun for most fans of basketball is in the minute details. Simplifying the game takes all the fun out of the game. Its more fun to list all of Carmelo’s go to moves as reasons why he is awesome than to than to list the two more rebounds and assists he needs to collect per game to truly live up to that billing.
Joe
October 4, 2010
http://www.ritholtz.com/blog/2010/10/asymmetrical-motivation-the-299-hitter/
Wonder if basketball has a 20 PPG thing like this… I can imagine Brandon Jennings crunching the numbers the night before game 82 and finding out he needs 25 points to average 20 PPG for the season and taking 30 shots in that last game.
nerdnumbers
October 4, 2010
Joe,
Remember the scoring battle with Shaq and Robinson? The last game of the season Robinson had an amazing game to take the title. Also Wilt’s 100 point game was similar.
An interesting this is this. Dan Ariely did some research on clutch shooting and found in clutch scenarios players percentages did not increase. I do not know if this applies in baseball though. It it does then I wonder if in baseball that means players are batting with a wrong strategy normally and by changing it (when chasing a record) they increase their odds of success?
Awesome read, thanks for sharing it.
brgulker
October 4, 2010
Quoted from the article by Widie,
lolwut? Ignorance is bliss, apparently.
The irony, it’s so thick.
Regarding the Cuban quote, that sounds a lot like something an owner of a team that’s failed to win a championship would say … or maybe someone who’s trying to justify paying all that money to Dirk, a player who’s clearly on the decline.
reservoirgod
October 4, 2010
@marparker:
I agree w/ you completely! It’s amazing how many ppl want to look at diagrammed plays at Synergy or break the game down to its minutae as if they get paid to be a scout.
Italian Stallion
October 4, 2010
I don’t think Cuban said much that super controversial.
This metric measures productivity. It’s not supposed to tell us how to build a team.
We already know that we generally a PG, SG, SF, PF and C and can’t just take the 5 most productive players.
We know there are diminishing returns for some skills etc…
robbieomalley
October 4, 2010
If todays game is any indication the Minnesota Timberwolves are going to run their offense through Michael Beasley at the starting small forward position similar to how the Nuggets use Carmelo Anthony.
Beasley played pretty well today, Ron Artest forced him into committing several turn overs but he shot lights out. Love, as usual, played like a beast.
robbieomalley
October 4, 2010
This might actually be a smart move by the Wolves. If Beasley moves over to the small forward position and maintains basically the same production he has been having, he would be around a ~.125 – .130 WP48 type player.
fricktho
October 4, 2010
How about Darko with 5 assists showing off those Webber like skills.
some dude
October 4, 2010
it’s pre-season. nothing matters.
arturogalletti
October 4, 2010
So use Beasley like Odom? That’s …. pretty freaking brilliant actually. he’s a tall jumpshooter. Might actually work out.
robbieomalley
October 4, 2010
I don’t get why some dude always has to have a problem with everything people say.
Italian Stallion
October 4, 2010
arturu,
IMO Beasley is another case where a player’s skill set does not match his official position designation well. IMO he IS a SF even when he’s playing PF. If it turns out he can’t defend SFs well enough, it creates a bit of a problem. But if the team can find a player that can defend SFs well and plays more like the typical PF, it can work well.
Positions are one of the things I think we have to remain more flexible about in out thinking. Beasley is Beasley. He isn’t suddenly a more productive or better player simply because he changes position to SF. It’s up to the coach and management to fit the right players together no matter what the official position designation. I think that’s one of the things Cuban is talking about.
some dude
October 4, 2010
robbie, I ain’t your mommie. I’m not here for positive reinforcement. I’m not here to cheer lead. I’m here to help improve analysis. That usually requires critiquing.
But I’m not all sour grapes. I’ve said good things to arturo. :D
reservoirgod
October 5, 2010
Whoa… lets keep Beasley in perspective & give Spoelstra some credit. They tried him at SF & PF – he’s below average at both. Beasley doesn’t have the handle to initiate the offense like Odom does for the Lakers and he’s too slow to stay in front of NBA SFs. The other problem he has is that he was below average at SF last season. Using his SF stats at 82games.com, his Win Score per 48 mins at SF was 6.7 (.139 per min, average SF Win Score per min was .170). That gives him an est. WP48 of 0.049. Unfortunately, that’s still better than what the Timberwolves got from Ryan Gomes (who got the bulk of the minutes at SF last season w/ a .020 WP48) so I guess you can say they’ve upgraded the position by playing Beasley there – but it’s not a great upgrade.
Robbo
October 5, 2010
After looking at some of the things Mark Cuban wishes he had stats for, it seems one of the benefits of Wins Produced is its use of already available stats, and relative simplicity. MC’s asking for so many detailed relatively minor/unimportant stats, it’s hard to see the woods for the trees, and then he doesn’t make full use of them…Wins Produced seems superior in many ways.
You can’t win them all over Dave, if you did there’d be very little difference between clubs, coaches and owners. So those who use WP get the biggest advantage, before the many distortions/inefficiencies in the market are evened out by accurate analysis.
Darrin Thompson
October 5, 2010
Prof. Dave,
Someday when you’ve tuned WP to the point that you can accurately predict season records to within 0.05 games, and still no one listens, and your life’s work has been dismissed without consideration, your career has been a failure, and you’re 80 years old muttering to yourself in your straightjacket in a padded room, remember this video about Donald Knuth and basketball. It’s not trying to be funny but the irony of how WP has been received in the real world is just too sharp.
—
Darrin
marparker
October 5, 2010
My question as always is do we really want people to “listen” to WP. If it really is superior why do we want to even out the knowledge? When it helped me start predicting WNBA games I stopped talking to anyone about it. I will come on the board but I don’t generally point anyone towards this information. Not that it would matter. But, if Michael Lewis ever writes a book about the Prof I will be really f-ing bitter.
Let the idiots be fascinated by crossovers and the wanna be stat geeks keep calculating charges and picks and floor spacing. I’m happy with the status quo.
My two cents.
arturogalletti
October 5, 2010
Darrin,
That’s awesome. Anyone else feel like they want to take a look at Knuth’s model?
marparker,
I think the problem for some of us is that we find inefficient systems annoying. The system is also out there for review.
Darrin Thompson
October 5, 2010
marparker:
I care because I want the Pacers to succeed since I’m a homer. I made a lot of good memories watching the Pacers-Knicks playoff series in the 90’s and the regular seasons during the “Beasts of the East” era. At the end of the day I like to watch my home team play solid basketball.
Your interests sound more, shall we say… economic. Not sure I’d have the skills to do that but it sounds completely brilliant.
Michael
October 5, 2010
Can we please have exact, win total predictions for each team in the upcoming season.
I’ve already seen predictions from Arturo for the Bulls (58 wins conservative, 67 optimistic) and the Wizards (10 or 22 wins), but can we get a table made for the entire league? Preferably with a single number prediction.
Michael
October 5, 2010
Using wins produced only.
nerdnumbers
October 5, 2010
Michael,
Just for the record that is not an easy task. That still requires estimating minutes for every player in the league and trying to predict for rookie play. In the near future some of the writers are going to put up our season predictions and I’m sure some of them will use Wins Produced and rookie models (I will be flipping coins and rolling dice, to giveaway my secret strategy)
Michael
October 5, 2010
I realise that it’s difficult, not knowing minutes etc. But since guys are starting to now say that this system predicts records within 0.05 games, I really feel that this needs to be tested before the fact.
arturogalletti
October 5, 2010
Michael,
I’ve got an exclusive contract with an octopus.
Seriously though, I will put this up (once rosters are final).
nerdnumbers
October 5, 2010
Michael,
This is an important stats thing to mention. The Wins Produced Metric has a linear regression of between 0.93-0.95 with the wins. What this means is 95% of the variance of wins in the NBA can be explained using the Wins Produced metric. It is also dependent on what has been seen. A player’s per-minute production has around an 80% correlation with previous years. So players stay stable, but minutes and team moves can drastically change things. (LA, Boston and now Miami all went from bottom seeds to championship contenders quickly)
The prediction thing is nigh impossible due to weird stuff happening. No one can expect a player to go down with injury, or if a trade will happen, or if a coach will be insanely stupid and bench his good players.
To sum up, you’ll have some predictions soon and I expect many of them to be good. In fact it is my belief they will be the best publicly available ones on the net! I’m just making sure to temper people against thinking that somehow being good at math let’s us know a coaches minute alignment or which players will be injured, or what trades will happen.
dberri
October 5, 2010
And no one said we can predict the future with 95% accuracy. We can explain 95% of wins with the box score data. But that is after the season has been played. As Andres (nerdnumbers) notes, predicting the future requires that someone know much more than how productive a player has been in the past.
harri
October 5, 2010
nice info….check it out blog saya
arturogalletti
October 5, 2010
Michael,
It is a hard exercise to build a model for forecasting. There are many,many variables involved. That’s what makes it interesting problem to tackle. We will get many of the predictions wrong but hopefully we can also get a lot right. The point of the exercise for me is to learn something though and improve my understanding of the math and the theory.
comixtakoma
October 6, 2010
Well any system that has a 95% ability to predict the past must be a REMARKABLE system, I’ll tell you what…
http://www.bulletsforever.com/2010/10/3/1728598/wages-of-wins-and-the-value-of-skepticism-more-thoughts-on-john-wall
jbrett
October 6, 2010
Holy crap, is there a bigger bunch of fatuous whiners anywhere than these Wizards fans? Your team went for the beauty queen; maybe she can learn to run the family business, maybe not. What’s done is done; you can stop lashing out now.
You want predictions? You’re already here; go up to the Search box and type in Allen Iverson, or Chauncey Billups, or Andre Miller. Read for a few minutes–just a few; it won’t take long. If you don’t come back to eat your words, you’re a waste of oxygen.
Tom Mandel
October 8, 2010
The problem with any roll-up number, no matter how well it correlates with the desired results, is that by nature it obscures its component numbers. This is independent of *which* metric it represents and even of how good the metric itself is.
In the case of John Wall, the 7.2 ws40 roll-up number obscures something that jumps out at you if you look at his boxscore numbers. *One* outlier stat — turnovers — has a huge impact on the ws40 number. Yes, he could also shoot better, but if his turnovers were average, his ws40 number would be among the best for a freshman point guard.
Thus the interesting next research step would be to look at how many guys w/ excessive turnovers in college were able to lower them significantly in the league. John Wall is an *individual*, so the research would offer a guide but not certainty. It would be *relevant* however in a much more specific way than a simple look into ws40 numbers.
Note that in 76 minutes in his first 2 preseason games, Wall totaled 6 turnovers and 7 steals. 76 minutes isn’t a big sample, but Dave wrote *extensively* about Wall based on a similar sample size from Summer League. Yes, he qualified his conclusions but first he made them with some rhetorical force.
As to sample size, it’s also worth mentioning that his college sample size is @1200 minutes. That’s not very big either.
As a thought experiment, plug the 2 game results from above into his ncaa numbers in place of what’s there. Now you have a ws40 of 10.3.
It’s easy to see what John Wall has to do to be as good as his fans would like him to be.
Tom Mandel
October 8, 2010
As I’ve written many times, and supported w/ the kind of analysis I can’t take the time to reproduce right now, a statistical measure like wp48 is “good” to the extent that it correlates with team wins. It is “better” than an alternative statistical measure (e.g. EFF or PER) to the extent that it delivers a better correlation. WP48 is better than anything else out there in this important sense.
Achieving a high correlation with such a measure is a *substantial* achievement, and Dave (and his confreres) deserve enormous accolades for their work. Anyone analyzing basketball talent should use this measure. Denver’s GM should have *grabbed* the deal he seems to have been offered for Carmelo Anthony and could have figured that out in 5 minutes using Dave’s work.
What wp48 does *not* do is rank college players in order of their likely productivity in the NBA. If it did, Bryan Zoubek would have been drafted in the top 6 in June. Ditto Artsiom Parakhouski. Above all what it does not do is give the analyst a magic tool to assess an individual player. For heaven’s sake, that is *not* what statistics does as a discipline!
You may illustrate this fact to yourself by asking: “Will Dave’s analysis have been ‘wrong’ if John Wall turns out to be outstanding?” Obviously the answer is “no.”
Analyses that cannot be falsified are not scientific. Moreover, if the answer above is ‘no,’ then how about this one: “Will Dave’s analysis have been ‘right’ if John Wall turns out not to be terribly good?”
The answer here has to be ‘no’ as well — and that makes my point succinctly, which otherwise I seem not to have done (but wrote a novel instead — sorry!)
doclinkin
October 10, 2010
I have no issue with Berri’s attempt to quantify basketball skill with a simplistic metric. It can be a useful stat, if perhaps overly reductive. I do get sand in my shorts over would-be Berriens blindly defending the dogma of the ‘Wins’ system, without a nod to context. Especially when the stataddicts put blinders on against all evidence to the contrary.
And I’m doubly irked when these assertions attempt to be predictive especially about guard play.
And that’s key: One glaring hole in the system is betrayed by the ‘position-adjusted’ nature required by comparisons. Bigs must be compared to bigs etc because the calculation weights rebounds heavily. And frontcourt players by the nature of their position, tend to scoop up more rebounds.
But rebounds are a zero-sum game. If your teammate snatches boards at a prodigious rate, your ‘win’ score drops. This handicaps players who happen to share the court with dominant frontcourt boardsmen. Or players in systems where their coaches prefer them to keep out of the way of the rebounding duties.
Derrick Rose is a dominant player at his position. Difficult to contain, unselfish, speed that’s unexceeded. This is acknowledged by scouts and opponents who have to gameplan to stop him. ROY voters tapped him for the honor. His team plays significantly better with him on the floor. Despite the handicap of a sub-par long range jumper, and despite his elusive nature at avoiding contact (hence opponent fouls) he manages to score at a remarkable rate. Remains durable because of the lack of contact.
But in the keyhole view of the ‘Wins’ stat [and the sheep who bleat about it’s virtues] Derrick Rose is not a good player. Really? Get the [flock] outta here.
In fact the key reason why DRose suffers in this metric lies elsewhere: Jannick Noah’s boy is a pretty good boardsman despite the unruly hair and pretty-boy eyelashes, bitching and pouting.
Joakim Noah posted above a 20% total rebound rate last year. One fifth of every free possession landed in his hands when he was on the floor. A greedy two whole shares, considering the ten players on court. Now did this hurt the team? No it did not. Chicago was a pretty good rebounding team, top ten, even with DRose as a non-participant in the muscle tussle.
Riccocheting among flying giants scrabbling for scraps was not Derrick Rose’s role. His role was to put up almost 800 more points on the season than any other teammate. That role he did rather well.
Fact is, evidence suggests that having high rebounding % in your backcourt is actually a problem. For those who do understand and are interested in coaching, as opposed to stats without context, the reason for this is evident. You have court-balance issues and transition defense issues (for would-be backline offensive rebounders).
Consider Golden State last year with a high proportion of high% rebounding guards — and the dead-last overall rebound rate. Now because of that your pet metric would take Kelenna Azubike over Kobe Bryant. Why? Because Pau Gasol and Lamar Odom took the lions’ share of the loose balls and Kobe didn’t have to. Really. You’d rather build a team around say Mike Miller than Kobe. Shaking my mfin’ head…
And with rearline rebounding emphasis you expose your smaller players to injury, which from a real ‘Wages’ perspective is key. A player on the bench posts a zero in total game possessions, while costing a significant portion of your cap. Notice players like DRose are actually out there playing, while rebounding PGs like Gil Arenas sit the bench with a busted knee hinge after Crash Wallace taa-daahed in from off-stage to crush him to the floor (benched him for 2 years and he got to watch helplessly while the media savaged him for accepting a big contract, offered while he was already injured).
What’s irksome is that the various dingle-Berris assert their own stat as evidence of proof of their stat’s utility. ‘My stat predicts that Derrick Rose will not be a good player– as measured by my stat.’ Funny quote above– ‘95% ability to predict the past’. (And thanks for linking to me, BTW.) I’m no mathemagician but seems to me all that means is if you use box score stats and run a few changes on them you will get something very similar to: box score stats.
So yes, from a nerf coaching standpoint: when you look at the Coach Cal dribble-drive motion offense you will tend to see two things high usage guards with a ton of responsibility (which means opponents will load up to stop them on them, increasing TO probability) and a single high-rate rebounder underneath (since the spread for attack draws the secondary bigs out to the perimeter). Thus: ‘Joey Dorsey is more important than Derrick Rose’ and ‘John Wall suxxors, the 25min/game Big Cousins was the entire team’. It’s not rocket surgery here, be prepared to challenge your own orthodoxy every now and again– even if it’s fun to be contrary and you know, smarter than everyday good old common sense.
dberri
October 10, 2010
doclinkin,
Have you read Stumbling on Wins? The issue of rebounding was addressed in that book.
Michael
October 10, 2010
doclinkin, regardless of whether I agree with everything you are saying there were some excellent turns of phrase in that comment. A few of them made me laugh out loud. Kudos!
doclinkin
October 10, 2010
Typo correction: Dribble-Drive Motion attack = ‘Spread Four’ attack. That is: one low-post player, four potential perimeter attackers. No ‘plays’ run; the primary attack guard initiates the attack, slashing to the basket and if impeded then either kicking out (restarting the attack at the perimeter) or dumping to the big.
Often the Post player is sent weakside to clear the paint, but Coach Cal also likes to use them rolling downhill to the basket.
@ Michael. Thank you sir.
@ dberri
Not for nothing but: reading these comment threads, often when a sticky point comes up that nibbles at the roots of the system, the answer seems to be: ‘Buy my book.’ Fair enough. Dude ought to be able to earn a dollar or two.
But being a cheap bastid, I’d happily take the cliff notes version. Heavy rebounding-rate front courts have no effect on the stats of same-squad rearline players because why. (I know, I know: ‘buy my book…’ sigh).
Fine, I’ll go back to the hinterlands of the internets now, to root for my quite possibly all-time worst team ever now, thanks.
dberri
October 10, 2010
doclinkin,
It does seem like you have some very strong opinions about my work. But my sense is — and your comment confirms — that you haven’t actually read the entire story. I really don’t want your money. Perhaps you could visit a library.