Yes, much of this post is a re-run.
But I am posting this story again because last week, 53 ESPN.com experts were asked to pick the top team in the Western Conference. The overwhelming favorite – with 41 votes – was the LA Lakers. Ten votes went to San Antonio Spurs while the Mavericks and Nuggets each received one vote. And the Portland Trail Blazers – the team that finished second in the Western Conference last year in efficiency differential (offensive efficiency minus defensive efficiency) – did not receive a single vote. Yes, even Henry Abbott – a long-time fan of the Trail Blazers – did not give his beloved Blazers a vote.
Matthew Yglesias – who is better known for his political insights than for his sports coverage – argued a few days ago that Portland is the “only team that could beat the Lakers in the Western Conference.” Although I am not sure that’s true, I do concur – as I noted a month ago — with the sentiment that the Blazers can challenge the Lakers.
This argument begins with what happened last year. The Blazers finished last year with a 5.9 efficiency differential. This was the second best mark in the Western Conference, eclipsing what was seen from the Spurs (4.1 differential) and the Nuggets (3.5 differential).
Wins Produced allows us to connect efficiency differential to the individual players. Table One reports the Wins Produced for each player the Blazers employed last year.
Table One: The Portland Trail Blazers in 2008-09
Topping the list is Brandon Roy. Last year Roy produced 15.3 wins with a 0.253 WP48 [Wins Produced per 48 minutes] mark. To put that mark in perspective, Kobe Bryant produced 15.0 wins with a 0.244 WP48 in 2008-09. Yes, Roy is just as productive as Kobe. So if you think Kobe is what separates the Lakers from everyone else, well…the Blazers have a shooting guard like that (at least in terms of overall production).
Unfortunately, Roy’s teammates only produced 39.9 wins last year; while Kobe’s teammates produced 46.1 wins. Because Kobe’s teammates were better, the Lakers won more games than the Blazers.
In the off-season – as I noted last month – the gap between these teams got much smaller. And to make this easier, let me just repeat what I said (yes, as noted, I am going into re-runs).
Portland Misses?
Portland’s objective this summer was to close the gap between them and the Lakers. It appears, though, that this gap has actually gotten bigger. The Lakers were essentially able to exchange a Trevor Ariza (a former second round pick of the Knicks) for All-Star Ron Artest (yes, he did make an appearance in this game in 2004). Meanwhile, the Blazers made every effort to sign Hedo Turkoglu, only to have Turkoglu sign with Toronto at the last moment. Then the Blazers turned to Paul Millsap, only to see the Utah Jazz match Portland’s offer. Finally, in an apparent act of desperation, the Blazers finally got Andre Miller to accept their money. This sequence of events had led Ken Berger of CBS Sportline to list the Blazers as one of the NBA’s losers in the 2009 off-season.
But did the Blazers really fail this summer?
To answer this question, let’s start with where the Lakers and Blazers finished the 2008-09 regular season.
The Lakers in 2008-09
Here are the top 10 players – in minutes played – for the Lakers last season (WP48 = Wins Produced per 48 minutes):
Pau Gasol: 2,999 min., 15.6 Wins Produced, 0.250 WP48
Kobe Bryant: 2,960 min., 15.0 Wins Produced, 0.244 WP48
Derek Fisher: 2,441 min., 2.6 Wins Produced, 0.051 WP48
Lamar Odom: 2,316 min., 10.6 Wins Produced, 0.220 WP48
Trevor Ariza: 1,998 min., 8.0 Wins Produced, 0.192 WP48
Andrew Bynum: 1,446 min., 4.8 Wins Produced, 0.158 WP48
Sasha Vujacic: 1,293 min., 2.7 Wins Produced, 0.099 WP48
Jordan Farmer: 1,192 min., -0.9 Wins Produced, -0.035 WP48
Luke Walton: 1,166 min., 2.2 Wins Produced, 0.091 WP48
Vladimir Radmanovic: 771 min., 1.5 Wins Produced, 0.094 WP48
Totals for Top 10: 18,582 min., 62.2 Wins Produced, 0.161 WP48
The Lakers won 65 games. But their efficiency differential of 7.8 (and correspondingly, the team’s Wins Produced) was consistent with a team that won 61 games (wins that can essentially be connected to the ten players listed above). So the Lakers were not quite as good as their won-loss record indicated.
The Blazers in 2008-09
LaMarcus Aldridge: 3,004 min., 6.7 Wins Produced, 0.107 WP48
Brandon Roy: 2,903 min., 15.3 Wins Produced, 0.253 WP48
Travis Outlaw: 2,246 min., 2.6 Wins Produced, 0.055 WP48
Steve Blake: 2,188 min., 5.3 Wins Produced, 0.117 WP48
Rudy Fernandez: 1,993 min., 6.9 Wins Produced, 0.167 WP48
Joel Przybilla: 1,952 min., 11.7 Wins Produced, 0.288 WP48
Nicolas Batum: 1,454 min., 3.7 Wins Produced, 0.123 WP48
Greg Oden: 1,314 min., 4.2 Wins Produced, 0.154 WP48
Sergio Rodriguez: 1,225 min., 2.2 Wins Produced, 0.087 WP48
Channing Frye: 746 min., -2.3 Wins Produced, -0.146 WP48
Totals for Top 10: 19,025 min., 56.4 Wins Produced, 0.142 WP48
Again, Portland’s Wins Produced for the entire team was 55.1; so the Blazers were about six wins off the pace set by the Lakers.
Evaluating the Changes
Now let’s consider the changes made to each team’s top 10.
First the Lakers:
The Lakers lose…
Trevor Ariza: 1,998 min., 8.0 Wins Produced, 0.192 WP48
Vladimir Radmanovic: 771 min., 1.5 Wins Produced, 0.094 WP48
Total Loss: 2,769 min., 9.5 Wins Produced, 0.162 WP48
The Lakers add to their top 10…
Ron Artest: 2,452 min., 4.6 Wins Produced, 0.089 WP48
Josh Powell: 703 min., -0.6 Wins Produced, -0.040 WP48 or
Didier Ilunga-Mbenga: 181 min., -0.2 Wins Produced, -0.066 WP48
Total Gain: 3,155 min., 4.0 Wins Produced, 0.060 WP48 (with Artest and Powell)
Overall Direction: The Lakers appear to be worse. Artest is simply not as productive as Ariza. And whether Powell or Mbenga takes the 10th slot, the team is really not helped.
Now the Blazers:
The Blazers lose:
Sergio Rodriguez: 1,225 min., 2.2 Wins Produced, 0.087 WP48
Channing Frye: 746 min., -2.3 Wins Produced, -0.146 WP48
Total Loss: 1,971 min., -0.1 Wins Produced, -0.001 WP48
The Blazers gain:
Andre Miller: 2,976 min., 11.1 Wins Produced, 0.178 WP48
Jerryd Bayless: 655 min., -1.4 Wins Produced, -0.104 WP48
Total Gain: 3,631 min., 9.6 Wins Produced, 0.127 WP48
Overall Direction: The Blazers appear to be better. Miller is clearly an upgrade over Sergio Rodriguez at point guard. It also helps that Channing Frye went away.
Once again…when we compare each team’s efficiency differential (and Wins Produced), it appeared the Lakers were only about six wins better than the Blazers in 2008-09. With the moves each team has made, this gap appears to be closed. In sum, if all we look at is what the veteran players on each team did last year, the Blazers are at least as good as the Lakers.
On the other hand…
Of course, all good economists have “the other hand” to look at.
It’s important to note that the Lakers did not have services of Andrew Bynum for much of the 2008-09 season. If Bynum is healthy, he could substantially improve the Lakers.
On the other hand… the same story could be told about Greg Oden.
Then again, on the other hand… Phil Jackson does appear to be one of those coaches who can change a player’s productivity. Maybe he can make Ron Artest better.
Then again, on the other hand…. Artest will be 30 years of age in November, so his production is probably going to slip.
Then again, on the other hand… Andre Miller is already 33 years of age. So how many more years can he be productive?
Then again, on the other hand… we are completely ignoring the changes made by the Mavericks, Spurs, and Nuggets. These teams, like the Blazers, might also be better.
Wow, that’s quite a few hands. Let me try and summarize. Contrary to what Berger argued, I think the moves the Lakers and Blazers have made have actually closed the gap between the two teams. The Lakers were clearly the best team in the West last year. It doesn’t appear to me, though, that the Lakers are clearly the best in 2009-10. So although I can’t guarantee the Blazers will make it to the NBA Finals in 2010 (remember what we found on the other hands), I think Portland fans shouldn’t think their team ranked among the losers this summer. As for fans of the Lakers… well, Phil Jackson really is a good coach so maybe it will still work out.
– DJ
The WoW Journal Comments Policy
Our research on the NBA was summarized HERE.
The Technical Notes at wagesofwins.com provides substantially more information on the published research behind Wins Produced and Win Score
Wins Produced, Win Score, and PAWSmin are also discussed in the following posts:
Simple Models of Player Performance
What Wins Produced Says and What It Does Not Say
Introducing PAWSmin — and a Defense of Box Score Statistics
Finally, A Guide to Evaluating Models contains useful hints on how to interpret and evaluate statistical models.
romalley
September 10, 2009
I don’t really have much to say about this post. It was quality. I just noticed you have posted four times since the 6th. Just want to say thanks for the good work.
Tim
September 11, 2009
Does this mean you are done with the next book? Or that you are trying to review all the teams before the season opens? Or both?
At any rate, I love seeing a new article each day as I wait impatiently for the season to start.
Rumblebuckets
September 11, 2009
Well, there are several tricks, like the one of saying Andre Miller is replacing Sergio. No, he’s replacing Blake. Still an advantage by the numbers. Save when you factor that Blake does one thing that Andre Miller can’t do. Shoot three pointers. And that might have some slight negative impact on other players games.
Then, right now, you also have Martell Webster coming back whose likely going to cut into the minutes of Rudy Fernandez.
And then you still have Lamarcus Aldridge who doesn’t rebound or defend particularly well and makes a living shooting 18-20 foot jump shots.
Maybe over a season they might do as well, but I have a hard time seeing them matching up very well against the Lakers, or even the Spurs, in a series.
They need to make a trade at small forward.
Also, I don’t like the contention that the Lakers (or any team) are aren’t quite as good as their record. The one thing a team is definitely is exactly as good as their record. It’s because they found some way to win (or lose) those games that win differential said they should not.
That, though, is not the same as it being a predictable skill (to be better than your win differential.) At least not right now. There may be things in common that teams that exceed this differential all do, that one would pick up on if one looked closely.
This is still not to say that point differential is not more useful when predicting future performance. Though in predicting future performance, other factors tend to be quite important as well, such as positional match-ups and style match-ups. For instance when Dallas (+.7.2 ppg) lost to Golden State (-.3 ppg), it wasn’t a fluke. Nor was it a fluke when Golden State almost beat Utah.
dberri
September 11, 2009
Tim,
You are correct on both counts. A completed draft of the next book does now exist. There is some editing that needs to be done, but it should be out by next March. And I am trying to review all teams by the start of the season.
wjmooner
September 11, 2009
@Rumblebuckets — Assuming you are talking about Golden State’s 2nd round matchup with the Jazz after they beat Dallas you couldn’t be more wrong. The Jazz destroyed Golden State in five games and, honestly, it wasn’t even that close. Yea, Baron Davis had one nice dunk, but that Jazz team just killed Golden State. What series were you watching?
Nick
September 11, 2009
Rumblebuckets : “Also, I don’t like the contention that the Lakers (or any team) are aren’t quite as good as their record. The one thing a team is definitely is exactly as good as their record. It’s because they found some way to win (or lose) those games that win differential said they should not.”
I disagree. Sometimes, the games come down to a jumpshot. And if someone is a 50% jump shooter, they should be expected to hit half the shots, and thus win half of these games. However, if a player misses 6 in a row. The team would then be 6 games under .500. To maintain the 50%, this player may hit his next 6 game winning shots.
However, at any point, if 6 misses fall into 1 season, and 6 hits fall into the next, the player did not FIND a way lose or win these games. He just hit or missed a jump shot, according to his talent level at hitting jumpshots. People like to attribute this clutch attribute to players, and in some cases, there is probably some correlation. But in general, players perform as good as players can perform, and the better players win. The fact that a series of positive random results occurred in what is a fairly small sample, does not mean the Lakers are a better team, it just means that the Lakers, had a string of positive random results in a fairly small sample size. Now I’m not saying they don’t deserve the wins, but, it’s not the “finding” a way. It’s mostly just randomness.
BTW…I’m a Celtics fan. :)
Italian Stallion
September 11, 2009
I agree with both go you.
A lot of the wins/losses in basketball come down to a random distribution of hit and missed shots.
However….
Anyone that has ever played any competitive game at a reasonably high level knows that there is a psychological component to winning and losing also. The end result is some combination of randomness and those psychological factors.
Assume the Knicks played the Lakers and the score was tied with one minute left to play.
If we calculated the probability of the Knicks winning based upon mathematical models that looked at the Knicks/Lakers stats for a full season, IMO we would sharply overestimate the chances of the Knicks winning.
If we examined how each performed in the last minute of all their games, we’d get closer.
If we examined how each performed in the last minute of a close game, we’d get even closer.
Theoretically, we could probably capture all relevant data inside that model and get it right, but we’d be measuring psycological factors that impact performance.
Nick
September 11, 2009
A jumpshot missed in the first minute of the game, may contribute to the loss in the same way one in the last minute of the game would. It’s just that the a large amount of events that follow the missed shot, are effected, and thus it is harder for people to predict what would have happened if the shot had gone in.
It’s far easier for a person to follow the logical conclusion, for a different shot results in the last minute of a game.
Italian Stallion
September 11, 2009
Nick,
What you are saying is also true, but in those last few minutes of a very close game IMO character issues play a much bigger part in the execution than in the first few minutes.
Sometimes, it seems some mathematically oriented people don’t quite understand that UNLESS they have actually tried to play something under extreme pressure and either peaked or went into a coma because of it.
There doesn’t even seem to be a formula for how you are going to react on any given day. I have played by far my best under extreme pressure, but folded and gone into a coma also. And inside, you KNOW when the pressure is causing that kind of volatility in your level of performance.
Some people seem to be naturally wired better to handle it, but we all seem to get better at it with experience (until your nervous system is shot from years of playing under high pressure).
Paciocco
September 11, 2009
Great preseason reviews so far.
I agree that adding Artest is going to cost the Lakers some regular season victories (but as a Laker fan I hope an improved and healthy Bynum can make up for it), but I think he’ll be very useful come playoff time (I’m sure Mr. Berri won’t agree with that argument, and who am I to argue with the defending stat geek champ?) Also, Shannon Brown is likely to give the Lakers better PG play than either Fisher or Farmar gave them last year.
As for the Blazers, the addition of Miller will help. Sure, he doesn’t hit the 3-pointer Blake does, but he does everything else better. His stats are likely to decline some this year, more due to the fact that Roy often acts as a defacto PG than due to his age (he never played an “athletic” game).
The Blazers though, have almost too much depth. Sure, that’s a stupid idea, but if a coach distributes minutes inefficiently because there are too many “valid” players on the bench, it’s an important idea. Rumblebuckets brings up a good point in Webster’s return. Fernandez is a very effective player. But will McMillan have the courage to give him 30+ mins when Outlaw, Batum, and Webster also will want minutes at the 3? Pryzbilla produced more wins than Oden last year; but who will play more, and who will play more effectively this year?
McMillan should again have one of the most effective benches in the NBA, but will he resist the temptation cut too deeply into his best 5 players’ minutes?
Tom Mandel
September 11, 2009
Isn’t there good data — and didn’t Dave write about this recently — that the closer the game (in any sport) the more randomness affects the win/loss results, and that better teams do worse (and worse teams better) than their overall records in close games?
In that case, what need is there to discuss this issue? No, “winning the close ones” is not a skill.
ilikeflowers
September 11, 2009
Rumblebuckets,
If efficiency differential is a better predictor of whether or not one team will defeat another in the future than win-loss record, then efficiency differential is a better measure of how good a team is. This of course assumes that a team that defeats another team more often than they lose over some number of games is defined as being more ‘good’ than the other team.
Italian Stallion
September 11, 2009
Tom,
Common sense suggests that if a bad team faces a good team and it’s a very close in the last minute, the bad team will win more often than their overall record because typically they are far behind at that point and highly likely to lose. Being close gives them a fighting chance that randomness will overcome their lower level of skill and lower level of psychological fitness to the occasion (and vice versa).
However, IMHO to say winning close ones is not partially related to aspects of ability other than those we typically measure (or to say there is not a psychological component to it) is far from correct.
It may be one of things that is difficult to prove statistically. However, like I said, if you’ve ever played anything at a reasonably high level you would know without any doubt that nerves, confidence, competitiveness, willingness to be the goat in order to be the hero, and a multitude of other personality, psychological, and other traits over which you have only limited control play a part in winning/losing.
We are not all blessed equally in those ways or even the same way each night. Sometimes there’s even a momentum to it and sometimes we learn to relax, become more confident etc.. with time.
Things like this may be overrated by some analysts/media types at times because sometimes a miss is just a miss and a game winner is just a randomly hit shot, but IMHO there should not even have to be discussion about this issue.
ilikeflowers
September 11, 2009
IS,
There are too many anecdotes for and against end of game ability to resolve it without some sort of statistical measurement. Certainly discussion is warranted.
The best teams (and hence most of the best players) aren’t going to have much opportunity to gain experience with these situations since they aren’t typically in them – having blown out the other team already. The 0.500 clubs are going to get the most practice. It might take quite a few seasons to even get a decent sample size for some teams/players.
There are certainly psychological factors as well, but are they typically significant in relation to random factors? How much do they fluctuate? How do randomness and small sample sizes affect the psychological issues? etc…
Italian Stallion
September 12, 2009
ilikeflowers,
I don’t have experience under fire on the basketball court, but I can talk from experience on the pool table. I would approximate shooting pool under fire to putting late in a golf tournament or hitting a critical free throw.
For me, how I was handling the pressure was hugely important to the outcome. I never attempted to weight what percentage was random and what percentage was psychological, I just know that both were an important factor.
The other thing I know is that if I measured performance volatility, high pressure caused much greater volatility. That is, it was more likely to bring out the best or worst of me than no pressure.
The same is true of almost every player I have ever spoken to about the subject. IMO it appears to be true of athletes also.
Let me make a suggestion to anyone that hasn’t competed under heavy performance pressure, but can’t think of a way to get that experience now.
Try playing some poker at a limit that is just high enough to make you uncomfortable.
The only physical activities in poker are picking up your cards and placing the chips in the pot.
Pretty simple right?
The first time you make a bet with a good hand and someone comes over the top “all in” let me know how easy it is to pick up the chips and place them in the pot without fumbling them!
Also, let me know how easy it is to calculate the odds quickly, run through all the actions that took place in the hand to try to determine what your opponent might have etc…
Not everyone is wired exactly the same, but most people are going experience difficulty thinking clearly, shaking hands, rapid heartbeat etc…. at levels that will interfere with their decision making ability and ability to perform some physical tasks.
Now imagine yourself trying to make a pinpoint pass, a super quick decision, a tough shot, etc…. feeling like that. It’s doesn’t work very well.
I think some people are wired much better than others to handle these situations. Some through good fortune or a combination of good fortune and skill gained a lot of experience in these situations and are more comfortable in them etc…
When a basketball team has a lot of players that are experienced and wired to perform well under pressure, I think it contributes to wins (and vice versa).
Some of that will be reflected in the overall stats, but since performance pressure is at it’s highest on only certain occasions, that ability may not be fully reflected in the overall stats.
Rumblebuckets
September 12, 2009
I think I was watching the series where Utah had a four point win (Utah overcame a 4th quarter deficit, late if I remember) and a 10-point overtime win that they could have easily lost in regulation, before killing Utah in Game 3, after which Baron’s hamstring which he previously hurt (in the Dallas series) and had all the time been taking cortisone shots for. Anyone whose had a hamstring injury, no matter how effective the steroid cortisone is at relieving pain, would find it hard to believe such a situation could continue forever.) So, I guess I’m talking about the series which could have just as easily been 3-0, but was actually 1-2.
And in Game 2 and Game 3, Baron Davis put Deron Williams in foul trouble and forced him to sit for large periods of time. When he was healthier, he just wasn’t a good match-up defensively for Williams and managed to take him off the court. Of course, Utah was ultimately the better team which is why they won, and perhaps it is even why they won those close games, but it was by no means as far apart as you would suggest based on 4-1. Those first two games were 50/50 games Utah happened to win. No other game in the series was close.
-What do we mean by how good a team is? “Is” is the strange word there. Do we mean how good a team was? If we mean how good a team “was” than record is a better indicator or at least as good of one, if we mean how good a team will be in the future, then point differential is a better indicator. They are two different questions.
The sentence, “So they were not quite as good as their record,” well that just doesn’t make a whole lot of sense to me. Because they won those games, didn’t they? They did it. A sentence that would make more sense would merely imply that the same team would not repeat their record in the future, which may or may not be true. (For instance, in baseball, the Angels have consistently been well over performing their run differential. Is this a result of just average variance to the norm, or, perhaps they have some small skill to do this?)
I don’t know if anyone has looked into it, but point differential is only an average, and so blowouts are going to contribute to it greatly, I think that possibly that may be a problem as point differential happens to still only be a crude predictor of future events. And while a great team is more likely to blow its opponents by huge sums, when you win games occasionally by 41 points, it skews the average. I would think it would be interesting to see how well the median victory works in its place, and if you could come to a more accurate formula by using that number in conjunction with the average differential.
Also on the question of “Winning the close ones being a skill.” In basketball, the statisticians general definition doesn’t necessarily correspond to what a close game actually is, that is, games decided by 5 or less. That is, I’ve seen dozen of games decided by 10 or more that a team had a chance to win in the last 2 minutes and then just faded, just as I’ve seen dozens of games in which a team loses by 5 or less, and never had a legitimate chance to win. So I’d be interested in seeing that study and how they define what a close game is.
I’d also say that for it to be considered a skill, every team doesn’t have to have it, only one. (There is a misassumption in these types of studies that something that is a skill will be commonly shared, when there’s no reason that has no to be true, and unlike baseball, where there is little clutch effectiveness and that only due to certain players aptitude to walk in end game situations, I think the stats that I’ve seen bear out that certain players are more effective in end game scenarios, so why wouldn’t teams be as well?)
One last thing, and why do we assume it’s a skill necessarily held by good teams? That would seem to be an implicit assumption we have, and know why we have it, but there seems to be no reason, why it must be true. Maybe, it’s actually a skill average teams possess from being in the situation more often. Or maybe it’s a skill that teams with high usage players who can still score efficiently possess. Because at the end of the games, the ball tends to be in the hands of your “best” player an awful lot. (I don’t know about any of this, you know it might not actually be a skill as well.)
http://www.82games.com/CSORT11.HTM
Look at Kevin Durant, I’d bet his team won a greater percentage of close games than you’d guess based on his team’s record. And I looked it up by the crude measure of “within 5 points”, a different measure than 82 games uses, and found they were 7-14 in close games, and 16-45 in games decided by more than 5 points. It’s a small sample, and they still weren’t good in close games, but they were definitely better than the rest of their games would indicate. What you also find is that they are still definitely quite capable of losing these games. And if it was really luck, you would tend to see it be around 50% which it isn’t. Neither I’d guess is it that way with good teams, though I’d bet they don’t seem to be as good as the rest of their record would indicate. (Looked up Lebron’s Cavs as well 11-5 in 5 point games, 15-5 in 6 point games, just a hint at how arbitrary that 5 point number I often see is. Still quite capable of winning, definitely not close to their record in the rest of games. I don’t know about overtime games though.)
And I’d suggest that a lot of what is happening in such a situation is that a good team tends to be in a close game when it is not playing very well (which is still often enough to win), where as a bad team tends to be in a close game when it is playing close to its highest capacity. If you could measure that effect against the clutch trends of the player with the ball in his hands at the end of games, I’d bet you’d come close to solving this question.
Well, that was a ramble.
ilikeflowers
September 12, 2009
IS,
I doubt that anyone will argue that ability is unaffected by pressure – it is a painfully obvious point and a waste of time and a great many words. End-of-game ability’s significance in relation to random effects is however unknown and eminently worthy of discussion since it is entirely possible that it is so small as to be negligible in the game of basketball.
todd2
September 13, 2009
I’m coming from a different perspective. If I’m a coach, I don’t want very many close games. The truly great teams dominate from tipoff. It follows that they would also have less overtime games. The pr0f’s scoring differential is probably a better indicator of playing under pressure—playing at a consistently high level throughout each game. It also occurs to me that the more close games a team has, the more minutes their “superstar” has to play, which isn’t necessarily a good thing. There’s probably an interesting study in there somewhere; an optimal number of people on a roster playing and/or minutes per starter each game and number of wins—or somesuch.
Italian Stallion
September 13, 2009
ilikeflowers,
Understood. The only reason I made such a long winded point of it is that I believe it is more important than my perception of the consensus here.
(waiting on an analysis of the Knicks)
Italian Stallion
September 13, 2009
todd,
I agree with you. I don’t see why anyone would ever want to play a close game. I also think a certain amount of the “ability to play under pressure” almost has to to be built into the point differential because players feel pressure throughout the game.
However, IMO they almost certainly feel more pressure in close games, important games, and in the last few minutes when games are often decided.
I think whatever the point differential suggests in terms of wins, an “experience/ability to play under pressure differential” is probably worth a few extra games a season.
Of course for many match ups it might be inconsequential. I don’t see teams like the Lakers, Spurs, or Celtics having any significant edge on the other in that area. Point differential is probably what matters most.
However, if I was going to make a bet, I would say that all those teams have a slight edge over Portland right now even if the point differential suggests they are equal.
At what point a Portland’s edge in point differential would make up the gap in experience is outside my range. But for gambling purposes it would have to be reasonably significant for me to accept “even money” against any of those more experienced teams.
For gambling purposes, you have to look for significant errors to overcome the rake and to give yourself enough margin of safety to overcome your own miscalculation and misunderstanding.
Italian Stallion
September 13, 2009
ilikeflowers,
Just another thought.
I very rarely bet on basketball. My specialty is horse racing. However, I would guess that the very best Vegas line makers and sports gamblers have studied the “experience and ability to play under pressure” factor in basketball very carefully. So they probably have a very good idea how much it’s worth in a game, series, and over the course of a season.
In fact, I think the work of the best sports gamblers almost certainly has to be worth looking at as compliment to the models that try to evaluate/rate players and their skills.
Man of Steele
September 14, 2009
Italian Stallion,
Being able to consistently “win the close ones” would be worth a few extra wins a season, if a team could control it. No one would disagree on this point. The point of contention is whether or not certain have this innate ability. Your own annecdotes suggest that they do not. You note from experience that playing ability fluctuates much more wildly under pressure. This is not indicative of a special ability; you claim that everyone experiences this fluctuation under pressure. Thus, even some who is best able to deal with pressure might crack in one particular situation. This realization indicates that in fact there is no innate skill for performing well under pressure, at least in any measurable or significant sense whatsoever.
todd2
September 14, 2009
I’m off on another tangent…too much free time. When was the last time (if ever) a player led the league in minutes played and won a title?
ilikeflowers
September 14, 2009
IS,
My own opinion on the matter is that end-of-game ability is likely significant compared to random effects. It would be wonderful however to somehow get a definitive empirical breakdown of how much close games are affected by luck, coaching, teammates, and individual ability. And it wouldn’t surprise me to find out that my opinion is wrong and that random effects dominate or that end-of-game ability is so variable as to be indistinguishable from luck.
todd2
September 14, 2009
For what it’s worth, the great players aren’t “clutch-performers,” they’re consistent. It stands to reason that if a person stinks up the place for three quarters but hits a game winning shot they’re more of a hindrance than a help (which ain’t how the game is marketed).
Italian Stallion
September 14, 2009
Man of Steele,
I was not clear.
First, I did not claim to be wired particularly well to handle pressure. :) I am about average in that regard.
I use the word “wired” because I think a lot of it has to do with your nervous system and other personality traits that we may not have much control over even after we gain a lot of experience.
I would say greater volatility is a factor for everyone, but in different proportions.
Someone like me might produce “dead stroke” under fire 1 out of 5 times and choke badly a similar amount of the time. Some of the best wired people with plenty of experience under their belt might produce their best 3 out of 5 and their worst 1 out of 10.
Both are examples are greater volatility than we would experience on the practice table, but they are different volatilities. Hope that makes sense.