How Unusual Was the Finals Foul Disparity?
Posted by Neil Paine on June 29, 2010
There's a good discussion brewing in the comments of yesterday's thread, centering on whether the big disparity in FTA between L.A. and Boston during the NBA Finals was something we could have expected based on the two teams' season-long tendencies. Here are the facts we know right now:
- Los Angeles received far more free throw attempts per shot in the immediate basket area than Boston did.
- Per possession, Boston was the 3rd-most foul-prone team in the NBA during the season; L.A. was the 2nd-least foul-prone team.
Basically, Boston fouled a lot more than L.A. during the Finals, but they also fouled a lot against everybody, and L.A. was a team that didn't foul much at all. The question is whether the disparity during the Finals was bigger than we would expect based on the teams' known fouling tendencies. Fortunately, I can estimate a team's expected fouls per possession and free throw attempts per possession by comparing their regular-season rates to the league-average and multiplying by their opponent's regular-season rates:
BOS Expected PF/Poss = ((BOS PF/Poss) / (Lg Avg PF/Poss)) * (LAL Opponent PF/Poss)
Boston committed .240 PF/Poss during the regular-season in a league where the average was .223; the Lakers' opponents committed .227 PF/Poss. Therefore, we would expect Boston to commit (.240 / .223) * .227 = .244 PF/Poss against Los Angeles, and in reality they committed .291 PF/Poss... But before you get conspiratorial, consider that the Lakers, who could have expected to commit .219 PF/Poss vs. Boston, actually committed .261 PF/Poss -- so the officials actually called more fouls on everyone. If you scale those numbers back down from the Finals-wide average of .276 to the regular-season league average of .223, the numbers would be .236 for Boston and .211 for L.A., which are actually very much in keeping with the .244 and .219 figures we would expect from their regular-season numbers.
That's only half the story, though -- free throws are where the conspiracy buffs really make their case, since the Lakers shot 51 more FTA than the Celtics did despite Boston taking more shots in the immediate basket area. Using the same methodology outlined above, except with FTA/Poss instead of PF/Poss, we see that Los Angeles took .259 FTA/Poss in a league where average was .263. Boston conceded .285 opponent FTA/Poss during the season, so we would expect L.A. to take (.259 / .263) * .285 = .281 FTA/poss in a series with the Celts; instead, they took .335 FTA/poss. On the other side, we would have expected Boston to have .244 FTA/poss against the Lakers, and in reality they had .249.
This result is unexpected -- Boston essentially had just as many FTA/poss as we would have expected based on team tendencies, but Los Angeles' rate of FTA/poss is dramatically higher than we would have expected. However, neither team's rate of fouls per possession was unexpected, so the only real unexpected outcome of the Finals in the fouling department was the number of free throws Los Angeles shot per foul drawn.
June 30th, 2010 at 2:34 pm
#50
Last edit: I promise, these are my final words; I realize that you can't cite fouls as a reason for losing if you win the series, but they become blown out of proportion when the team you are rooting for loses, and this happens with every fan base.
June 30th, 2010 at 2:51 pm
One thing that seems to be overlooked is whether or not the Lakers knew about the Celtics tendencies and used it to their advantage.
The stats for fouling would be just as available to the Lakers as they are here. And as the Laker scouts are getting paid big to provide information for coaches to use anything to their advantage, I would expect the Lakers (and Celtics) to change their game plan.
What I'm getting at is that the statistics point to how the teams played in the past which can be a good indicator of play in the future. However, statistics don't dictate that a team will play like they did before. If you believe that Jackson is a coaching genius, would it be difficult to assume that he would create a game plan that takes advantage of the Celtics tendencies? And yes you would expect Rivers to do the same. Then it becomes a question of which team is better able to execute the game plan.
I personally would expect to see a team do things a bit differently in the Finals depending if they can impose their will upon the other team. The West is soft and uptempo (stereotype) and the East is hardnosed, grind it out. Up until the Finals they can play one style for different opponents. But in the Finals, you have to impose your style or counter the one you're playing against.
All I'm saying is that stats don't take into account that people are involved and can behave differently than stats indicate.
June 30th, 2010 at 3:02 pm
#51
Probably the last post for me as well. You are still really missing my central point which I'll state for the last time:
Fouls are subjective. They can influence a game or a series in general, but they become a larger influence on the result of a game/series the closer the teams match up.
Because of this, a seven game series can end upon the whims of the officials (if you are a conspiracy believer which I am not), or on the ineptness of the officials (of which I wholeheartedly agree with). I wish this could be fixed, but it appears as though the NBA doesn't want to.
So essentially, a seven games series that ends with a four point differential can equate to something of a coin flip given the variance of expectation that the subjective officiating adds to the equation, which irks me a little because the final say in this championship may not be determined by the players, but by the referees. I would think that this is a large problem that would upset more people, but I guess it isn't as long as you are on the winning end of it, which is what I gather on this forum as well as your blog.
You reference to Cleveland vs Boston doesn't hold water because those two teams weren't all that evenly matched in the playoffs. This is obvious if you watched the games...Boston was simply a better team in the playoffs than Cleveland in the same way that L.A. was simply a better team than Phoenix, Utah, and OKC. Going back to my central thesis that:
Fouls influence the outcome of games/series that are closely matched.
June 30th, 2010 at 3:09 pm
#52
Stevie,
I think the Lakers were better at playing to the referees tendencies, which is what I really believe is the source of the FTA differential. If you notice, the Lakers went into super-flop mode in every game that Joey Crawford and Joe DeRosa officiated. Those two guys call games on the tighter side of the spectrum, particularly Crawford...who calls it super tight usually in the first quarter and the last, from all of the games I've seen.
If I can see these tendencies just from watching games, and consulting a few gambling websites that contain stats on officials, then it would be reasonable to believe that the teams are also scouting for this as well, and admittedly the Lakers were better prepared in this regard. Whether or not this is good for the game, is an issue that I've been trying to convey to everyone to no avail. Playing to garner favorable calls from officials is not better for the game than just playing basketball. Just watch any FIFA game and you can see where this trend will lead you.
June 30th, 2010 at 3:14 pm
The difference can be explained by two assumptions that Neil used (perhaps without thinking about them) that are not valid.
First: By using the regular season team numbers, Neil is assuming that the distribution of possesions used by each player in the finals will be identical to the distribution during the regular season. This is obviously false.
I took a look at FTA/MIN for each player (I would love to use possessions but they aren't provided on this website ) as a proxy for FTA/Poss. Looking at the Lakers it is obvious that Kobe Bryant and Pau Gasol lead the team by a wide margin in FTA/min. During the regular season they played only 71% and 61% of the minutes each game (partly due to injury) but in the Finals they played 86% and 87% of the minutes. Obviously with Kobe and Gasol playing more minutes they are likely to earn more foul shots as well. In addition to just minutes, Kobe's usage rate was higher in the finals than in the regular season as he took an even larger load of the offense so the expected free throws for the Lakers would likely be even higher yet.
Using just FTA/MIN and the minute distributions from the regular season and the finals I calculated that the Lakers would have expected a 8.8% increase in FT/MIN just due to the shift of minutes between the players. Similarly, looking at PF/MIN the Lakers would have expected a decrease of 4.1% in the number of fouls called per minute do to the shift in minutes between the players. Also, given that Kobe and Gasol's usage increased in the finals these numbers are likely understated.
Second: The team's numbers don't contemplate match-ups (Neil has no way to account for this anyway). Looking at the FTA/MIN by player between the regular season and the playoffs some interesting numbers emerge. Kobe saw only a slight uptick per minute (probably due again to using more possessions per minute), but Pau Gasol saw a huge increase and the rest of the Lakers were close to their season average (even slightly down). Gasol saw quite a bit of time against Perkins and Wallace and both are significantly above league average in fouling. It isn't a surprise to see Gasol get more FT's when going against those guys.
On the other side of the ball, the two Celtics who suffered the biggest decline in FTA/Min were Paul Pierce and Rajon Rondo. Pierce was matched up against one of the best players in the league at defending him. Artest has the strength to prevent the post-up and quickness to move his feet. We would expect to see Pierce's FTA/Min drop against certainly an above average defender. It also shouldn't be too surprising that Rondo saw his FT's drop as well. I don't see too many teams having the guy guarding Rondo play 10 feet off of him and it is tough to get a foul from 10 feet away. Also, as the series went on I thought Rondo became hesitant and afraid of contact because of his inability to hit a free throw when he did get to the line (26% for the series).
Given the change in minutes which put the Lakers best players at drawing FT's on the court more and the match-ups that limited Boston's top two guys in FTA during the season, is the free throw descrepancy really that surprising? I don't think so.
June 30th, 2010 at 4:18 pm
Neil has uncovered a statistical oddity where "Los Angeles' rate of FTA/poss is dramatically higher than we would have expected." I for one would just like to understand if this is due to officiating, or if it is something more player related."
The problem is the analysis is based on league wide stuff and doesn't tell us anything about matchups. That's the thing about sports is that all the stats league wide can never account for matchups.
And as I said, game 7's discrepancy seemed plainly obvious as a result of Boston tiring out.
Personally, I thought the officiating favored Boston slightly more than L.A. throughout the series, though not purposely.
I agree with you that officiating in the NBA is a problem. Inconsistency, age, what I perceive as grudges, and other things play a factor. And it sucks.
But just as easily as you say a "4 point margin," i can point out that L.A. seemed to be robbed in game 2 and had they won that one the series is probably over then and there. It is what it is. Officials can change outcomes; what is most important is that they don't intentionally change outcomes. Incompetence should then be rooted out.
Anyway, the NBA Finals was at times reffed too tightly, but I don't think it heavily favored either team much (with the exception of game 3 and Kobe being taken out).
June 30th, 2010 at 4:24 pm
#55. That's a good point about the distribution of possessions being more concentrated among the high-usage players in the playoffs. But as far as matchups go, you can account for that the same way I did (comparing one team's fouling tendency to the opponent's tendency to be fouled), you would just use amended numbers to account for more possessions being used by high-FTA/poss players.
June 30th, 2010 at 4:53 pm
#57. Neil, I agree that you did make an adjustment for match-ups on a team level, but what I was getting at is match-ups on an individual level. The Lakers putting Kobe on Rondo and having him stay 10 feet off of him is prime example. I doubt the league on average did this same thing to Rondo and this type of scenario would not be present in either the Lakers or Celtics season stastics.
June 30th, 2010 at 5:20 pm
Well, first, that's been the book on Rondo for years. Did any team do it as successfully as the Lakers? No. But every team in the league at least knows to try to force Rondo to shoot jumpers... Anyway, the point of what I was saying was that matchups at the the team level are just a combination of a lot of individual matchups. If the Celtics give more playing time to players who draw fouls, and the Lakers give more PT to players who avoid fouling, you can simply create a new team "tendency profile" to reflect that. Every team employs different strategies, but none of those strategies is so much different from the league average that it taints the methodology.
June 30th, 2010 at 7:08 pm
But you're matching things up based on statistics, not how to say this...the way the game was played.
For instance:
let's say Bynum is prone to fouling in help situations a lot, and most specifically on point guards and centers who penetrate into the lane looking to attack the basket. I think this is actually reasonable as it happened vs Westbrook, Amar'e, and Deron in the playoffs.
Now, Bynum's foul numbers dropped big time in the finals? Why?
A. Garnett never attacked him when matched up and Perkins has been too scared to shoot for 4 months since his wrist injury.
B. Kobe played so far off Rondo, he almost never got into the paint (outside of fast breaks), so Bynum was rarely in position to take a charge.
So while Bynum usually fouls little dudes coming into the paint, this rarely happened in the series because of the matchup. League wide number analysis will never figure this out.
It doesn't "taint" the methodology, but it at least suggest that this could be a situation outside the confidence interval.
There will also be certain matchups where long term numbers won't matter. Perhaps they're uncommon, but they certainly exist.
And while many teams play back off Rondo, none played as far back as Kobe and this is because the Lakers are the only team with the inside length and ability to play their taller SG on Rondo. Name me another NBA team that could pull this off? Portland when everyone is healthy, perhaps?
June 30th, 2010 at 7:09 pm
"to take a charge" was meant to be "to take a foul"
June 30th, 2010 at 7:12 pm
This has been a great discussion. We need to account for fatigue when calculating fouls, especially in the NBA Finals. Anyone who has played basketball knows that you are far more likely to foul (especially on the defensive end) when you tire. Throughout the Finals both Boston and LA looked tired. Both teams had played a lot of games and were dealing with injury and fatigue. This was especially true during the second half of Game 7 (by far the most physical game of the entire playoffs). Both teams looked drained in the fourth quarter. Kobe admitted as much after the game. Boston's front court players were especially gassed. Each of them played more than they were used to because of Perkins' injury. A number of the fouls (for both teams) occurred when players were reacting slowly to the ball. Players were reaching and grabbing instead of playing good defense or rebounding cleanly. I think this was especially true of Boston. I think this makes it difficult to make too many assumptions based on this quarter.
July 1st, 2010 at 9:45 am
Nothing works better than the eye test. Most of the boston fouls were fouls but in the fourth quarter the refs swallowed the whistle on obvious fouls commited by the Lkaers. Gasol fouls Garnett on almost every shot he takes but becuase he fades away it's easy to miss,but garnett beat gasol to the basket and Gasol fouls him hitting him on the wrist but gets credited with a block, they show the replay and announcers say,hand part of the ball(lol). In real time it was a obvious foulcuz he got beat and garnet had the ball up high, Glenn davis goes to the rim shoots draws lots of contact(body at that ) no foul( this happened numerous times in this game and throughout series no call). That crazy looking play by pierce on Gasol was actually a headsy play cuz the big man (gasol) was dribbling with his head down blocking call on pierce. The whistles went one way in final quarter. They called anything close for LA while ignored the obvious for Boston.
July 2nd, 2010 at 11:33 pm
I didn't read through all the comments since there are so many but has anybody factored in the issue of teams leading at the end of a close or semi-close game? The team in the lead is going to shoot a lot more free throws when the other team intentionally fouls. This is something else that the statistical analysis might not have accounted for.
July 2nd, 2010 at 11:36 pm
Following up with my previous comment, one would expect finals games to be tighter than the average regular season game of the top teams in the league. So the issue of fouls at the end of a close games might have some significance for foul and free throw attempt stats.
July 7th, 2010 at 5:32 am
If only Kobe had received the same treatment from refs like MJ DWade and LJames and hell even Durantula..
July 8th, 2010 at 4:06 pm
The problem is, L.A. got the early whistle in the 4th qtr (game 7) putting them in the penalty VERY early. It appears that after BOS went up 13 pts, the fouls started going L.A.'s way.
I'm not sayin'...
I'm just sayin'...
July 26th, 2010 at 5:52 pm
Stern fixed the finals and helped LA, its the only way they would have won.
September 6th, 2011 at 2:49 am
How can anybody talk about the 4th quarter of Game 7 of the NBA Finals 2010 without talking about Pau Gasol? and I find it odd that nobody brings up Gasol's plays in the 4th qtr.
Gasol was not whistled for 1 violation in the entire 4th quarter. Gasol knocked two balls out of bounds, travelled twice, charged into Pierce, knocked down Rondo, hit Davis in the head and knocked his head backwards, got a cheap and 1 on the first play of the 4th with Davis, and got another bailout on a play with Garnett, ALL in the 4th qtr.
Gasol was involved in 10 plays in the 4th quarter, ALL 10 plays gave the ball to the Lakers. ALL 10 PLAYS. Gasol was untouchable in the 4th qtr of Game 7. Gasol was called for nothing, no foul on Gasol, no travel on Gasol, no ball knocked out of bounds by Gasol, no up and down by Gasol, nothing. Absolutely nothing.
The refs used their whistle 10 TIMES in thie 4th qtr of Game 7 on plays involving Pau Gasol. ALL 10 PLAYS gave the ball to the Lakers. How does ANY player get 10 calls in their favor, with 0 against, in the last quarter of the entire season? How does that happen? How can you talk about Game 7 2010 without bringing up the plays Gasol was involved with. Even when you talk about the 4th qtr, that doesn't include Garnett's basket taken away on and And1, Rondo's layup taken away, Gasol hacking Garnett then the refs call it a block, Kobe falling down while dribbling no travel called then the Lakers score on that posession, etc.
Go watch the 4th qtr again. Go watch the 10 plays that Gasol was involved with in the 4th quarter where play was stopped because of a referee whistle. Go watch as ALL 10 PLAYS give the Lakers the ball, and tell me how can that happen in the last quarter of an entire season without something funny going on.