Wilt Chamberlain: The Numbers Don't Lie
Many of the things asserted by basketball "stat gurus" offend my ears (and mind) more than the sound of long fingernails scraping a chalkboard but one piece of nonsense that particularly bothers me is when a "stat guru" attempts to "normalize" one player's numbers to supposedly determine how that player would have performed in a different era. For instance, a popular "stat guru" declaration is that Michael Jordan's 37.1 ppg average in the 1986-87 season--when "adjusted" for pace--is actually superior to Wilt Chamberlain's record-shattering 50.4 ppg average in the 1961-62 season. There are many problems with this deceptively simple comparison:1) There is no way to accurately "adjust" for the relative competition that Chamberlain and Jordan faced; Chamberlain played in a smaller league with less players per team, so it could be argued that he played against tougher competition, the very best of the best--but Jordan played in an era with superior knowledge about nutrition and training and he faced players from a greater number of countries thanks to basketball's global expansion so perhaps Jordan played against tougher competition. A good case could be made for either side of this argument but the point is that no one knows for sure what the correct answer is. Another related issue is the question of whether the very best athletes in the world were more likely to play pro basketball (as opposed to another sport or as opposed to seeking out another occupation entirely) in the 1960s or in the 1980s; there is plenty of room for intriguing speculation about this but no way to draw definitive conclusions.
2) Regardless of whether or not 37.1 ppg scored at a slower pace is mathematically equivalent to 50.4 ppg scored at a faster pace, human beings are not machines; making extra field goals and extra free throws over the course of an 80 or 82 game season requires a tremendous expenditure of energy and increases the likelihood of fatigue and/or injury. In other words, the fact that Jordan scored 37.1 ppg at a slower pace tells us nothing about his capability to score 50.4 ppg at a faster pace, even without factoring in possible differences in competition level and definite differences in diet, nutrition, scheduling and travel arrangements.
3) The NBA has been around for six decades and during that time pace has gone up and down but no one has even come close to doing what Wilt Chamberlain did statistically--not just in scoring but also in rebounding and even in terms of passing from the center position (Chamberlain is the only center to lead the league in assists). If pace were the only factor affecting individual scoring averages then one would assume that in higher pace eras someone else would have at least come close to matching Chamberlain but, while Chamberlain exceeded 40 ppg in four different seasons, no other player has even come close to averaging 40 ppg in one season.
There is a big difference between saying that Jordan's 37.1 ppg is proportionally greater than Chamberlain's 50.4 ppg based on pace and definitively asserting that Jordan's 1986-87 scoring feat was greater than Chamberlain's--but basketball "stat gurus" have no qualms about making extraordinary claims without providing extraordinary proof, which is the very opposite of the approach that authentic scientists and researchers take; that is why physicists are still running experiments to test Einstein's Theory of Relativity--arguably the most successful and influential theory in history--while many "stat gurus" refuse to even acknowledge that basic box score data is flawed and that therefore the so-called "advanced basketball statistics" are skewed even if the "advanced" formulas are sound (which is far from a proven proposition).
In my pro basketball Pantheon I did not attempt to rank players from different eras but simply selected the 10 players who excelled when compared to the players from their own eras; how much a player dominates his own time is a significant indication of true greatness. Fran Blinebury's recent Wilt Chamberlain tribute notes that Chamberlain dominated his peers in breathtaking fashion (in reference to the first point in the passage quoted below from Blinebury's article, it is worth noting that Blinebury's larger point is correct even though he failed to mention that Elgin Baylor averaged 38.3 ppg in 48 games in 1961-62):
• Consider that after Wilt's 50.4 mark for the 1961-62 season, the second-highest scoring averaged in NBA history by a player not named Chamberlain was Michael Jordan's 37.1 in 1986-87. That makes Wilt's number 36 percent higher than Jordan.
• The highest batting average for a season in Major League Baseball over the past 70 years was George Brett's .390 in 1980. To exceed Brett by 36 percent, a batter would have to hit .530.
• The all-time single season rushing record in the NFL is 2,105 yards by Eric Dickerson in 1984. To exceed Dickerson by 36 percent a runner would have to gain 2,863 yards.
• The NHL single-season record for goals is 92 by Wayne Gretzky in 1981-82. To exceed Gretzky by Chamberlain's pace, a skater would have to pump in 125 goals.
The truth is, in American sports, only Babe Ruth transcended and transformed his sport like Chamberlain.
Pace alone is not an adequate explanation for how far Chamberlain's records are ahead of not just what any other pro basketball players have accomplished but also how much more dominant his performances are than the record-setting performances of all-time greats in other sports.
Oscar Robertson recently penned an eloquent plea urging that the NBA's great history--including the incredible 1961-62 season in which Chamberlain averaged 50.4 ppg and Robertson averaged a triple double--should be remembered and celebrated. I wholeheartedly echo Robertson's complaints and laments and I am proud of the opportunities I have had to interview Robertson and other greats of the game. Robertson is right that it is important not just that NBA history be told but that it be told by competent people; my contribution to that effort is displayed in the right hand sidebar of this website and I truly hope that someday my hard work and dedication to preserving and telling these stories will reach the widest possible audience, supplanting the gossip and nonsense that poses as journalism today at far too many magazines and websites.
Labels: "advanced basketball statistics", Fran Blinebury, Oscar Robertson, Wilt Chamberlain
posted by David Friedman @ 10:19 PM
14 Comments:
Pretty spot on analysis.
I have been battling many detractors of Wilt for a long time, both here and In Real Life, and their common criticisms are the following:
Wilt played against short white stiffs. (easily debunked when the fact that the average height is pretty consistent over 40 years is demonstrated, that Wilt played better competition at the center spot nightly)
Wilt played in a fast paced era. (easily debunked with the fact that only he dominated to that extent, notwithstanding Oscar's five years of triple doubles, or Baylor's unreal 38 and 19)
He only won two titles. (despite the fact he was the greatest opponent of the greatest dynasty in NBA history, that he performed great in Game Sevens, and came within a few points of winning more titles).
He sucked on the free throw line. (can't deny that)
Your point about pace factor easily refutes Eliot Kalb's contentious claim that Shaquille Oneal is the greatest of all-time. There were many other factors (style of play, no 3 point line, fewer anally retentive coaches who call plays every possession, etc, etc)
I'd like to see the data that shows the average center height in Wilt's era, as I believe it was much shorter than it is now (the white stiffs argument).
I know this is a statistical fueled article, but there are other factors at play beyond stats. For instance, the fact that most people believed Wilt to be a bit of a jerk and a pain to manage.
Comparing players of different generations is always going to be a difficult thing to do. A lot of today's basketball fans are also swayed by the fact they watched Jordan play frequently, and haven't spent a lot of time watching Wilt.
It was a different game at that point...so many factors...knowledge of training/nutrition, defense schemes & rotations, the abilities/height/desire of the average athlete, living conditions, creative double teams, etc...
Basketball Training:
This article is not about personalities but rather about the flawed attempt to use "advanced basketball statistics" to allegedly "prove" that 37.1 ppg is in fact a greater accomplishment than 50.4 ppg. However, if we are going to talk about personalities/coachability it should be noted that Chamberlain was coachable enough to be the dominant player on the two greatest single season teams in the first five decades of the NBA's existence (1967 76ers--voted the greatest team of all-time in 1981--and 1972 Lakers). Michael Jordan often clashed with coaches/teammates and when Tex Winter--trying to convince Jordan about the value of the Triangle Offense, which Jordan initially mocked as an equal opportunity system--said that there is no "I" in team Jordan replied that there is an "I" in win. Jordan's personality can either be viewed as exceptionally competitive or as difficult to coach (the same thing is true of Kobe Bryant, which is why it is so fascinating that the rough edges of Jordan's personality are incessantly praised while Bryant is criticized for having very similar personality traits).
I agree that because of the factors you mentioned--most of which I mentioned in my article--it is difficult to make objective and fair comparisons of players from different generations.
In a critique of ESPN's ranking of Lebron James as the best player in the league, Bill Reiter from Fox Sports wrote this : "If this logic held, Wilt Chamberlain would be the Michael Jordan of the NBA. No one ever has, or will, compete with Wilt on the statistical plane. Yet there’s a reason serious basketball people look at Jordan, Magic, Kareem, Russell and a slew of others with a higher level of respect."
I swear I don't make this stuff up.
Matt:
I spoke with several former players/coaches and Dolph Schayes, Oscar Robertson and Warren Jabali are among those who would take Wilt Chamberlain over Shaquille O'Neal. I guess those guys are not "serious basketball people" in Bill Reiter's eyes.
Also, Tony Gwynn hit .394 in 1994, which he fails to recognize. I'm not sure why people overlook that season for Gwynn. It was a strike-shortened season, but he played in 110 games, which is a greater pct. of games played than baylor's 48 game 38.3ppg season. Also, Brett only played in 117 games in his .390ba season, with only 40 more PAs than gwynn had in 1994.
That's an interesting point about Jordan. He probably was extremely hard to coach, though any coach would certainly take him. I am also perpetually mystified but how Kobe usually is bashed for the same qualities/personalities than jordan gets praised for.
It sounds like normalizing wilt's 50.4ppg season to jordan's is another attempt to elevate Jordan to some type of supernatural power, asserting that nobody ever is anywhere near as good as jordan was. I don't quite understand this.
@BasketballTraining
Example: The height of the 2006 Miami Heat was indistinguishable from that of the 1972 Lakers.
Fact: in the 60's and 70's players' height were taken barefoot, contra to today.
I will look up the average height later.
Keep it up, David. Your work is truly excellent, and it never fails to add much needed perspective to any NBA discussion. My only wish is that you also wrote about the NFL.
-David-
Before I even get to the numbered lines I do want to touch on the first part of what you said. I don’t try to normalize numbers to determine how players would play in other eras. I don’t believe placing players in other eras even makes argumentative sense. We don’t own time machines, Jordan will never play in ’64, Wilt will never play in 2002, Steph will never play in ’70, and so on. Players develop skills and traits that make them more effective in their era against their competition, for the way the game is played at that time under those time specific rules and general ideologies. Each player’s greatness is only relative to how good he is relative to his competition. This is why things like true shooting % 30 years apart is a bad comparison because the league average true shooting % for each period likely isn’t the same. People claim Steph might not be as good of a player playing before the 3pt line existed, which might be true but he didn’t play then, he plays now, and his obligation to himself and team is to be the best player he can be right now. He has no reason to focus on things that would showcase he might be really good in 1975 if he’s never going to paly in 1975.
1.) I don’t try to adjust for competition. I have no interest in trying to determine which players played the best competition because at any given time the NBA is the best basketball players in the world. As you said “no one knows for sure what the correct answer is” people just pick their favorite era and inject bias.. I don’t know what I’ve said to make you feel like that’s what I’m doing, but I find those discussions to be largely pointless.
2.) I don’t know that this is true. Keep in mind playing at a “faster pace” isn’t literally playing faster, it just means the ball changes hands more. If one teams dribbles up the floor and shoots an air ball early in the clock, the other teams walks to the ref, in-bounds the ball, the other team walking back on defense, say that team then makes a bad pass for a turnover and the player who jumped the passing lanes makes a lay-up while everyone is hanging back watching. So then we inbound the ball again… In that instance this next in-bounds pass would be the 4th possession in a matter of say 22 seconds with only player truly exerting a ton of energy being the player who jumped the passing lane and shot the lay-up… In another example a team could dribble the ball up the floor, run a set play take takes the same 22 seconds and it was only 1 possession for the same amount of time. Pace is literally just the number of possessions. In the first example the pace of that game is 2 times higher than the pace of the second example, but in the second example every defender on the floor is likely running off screens, helping, recovering, everything that goes in to forcing a team to use 22 seconds of a shot clock. In the other example mostly walking end to end, watching, in-bounding, etc.
Anonymous:
I believe that you posted comments about this article in a different thread first, and you are now continuing the conversation here, which will confuse anyone who has not seen your prior comments. In the future, please post your comments to the relevant article.
Regarding your comment above, (1) I don't know who you are, so I don't know what numbers you prefer or what you are trying to prove. If you state who you are and which numbers ("advanced" or otherwise) you prefer then we can productively continue the conversation; (2) for the reasons you stated, the reasons I stated above, and other reasons, "pace" is not a very meaningful or useful metric and it is worthless regarding the specific comparison we are discussing (Chamberlain versus Jordan).
Part 2/3
There is no way of knowing who exerted more energy, had more injury risk, all of that without tracking every single possession. We have that data now, but obviously not for eras prior. Keep in mind though even in just comparing bigs then vs now big men largely jogged up the floor from post to post. Wilt was obviously well conditioned to play all 48+ minutes, but running block to block and rarely leaving his post from under the hoop on defense to the block on offense is not really indicative of him having the cardio endurance of an elite marathon runner. Even comparing to a bigs now they run up the floor, maybe start in the post, come high to set a ball screen 30ft from the rim, start to roll to the rim, come back to run a hand off, then pop for a jumper. Defensively they have to defend in the post, come high to hedge, sometimes switch and have to chase small guards, recover back to the post, and then maybe lunge out at a guy shooting a 30ft jumper. For all we know a player who plays 30 minutes now might exert as much effort as Wilt did in 48 minutes then. I’m not claiming that to be true, we can’t even come close to measuring that with so few full Wilt games to track, but I’m more so just trying to make the point that “faster pace” doesn’t really have direct correlation to effort exerted. They aren’t the same thing, running a race at a fast pace is literally running faster. Playing a basketball game at a faster pace is just how many times the possessions changes. Pace is just number of possessions, a game with 100 possessions has a pace stat of 100.
You’re right that Jordan averaging 37 playing 80 possessions doesn’t with 100% certainty tell us what he could average playing 135 possessions. But Wilt averaging 50 at 135 possessions also doesn’t tell us with 100% certainty what he could average if he only had 80 either. That’s sorta the whole reason for attempting to normalize the stats to begin with. “per game” numbers are just a counting measure of tendencies. For whatever reason everyone views that as a knock on Wilt, it’s not, it’s just trying to get a more realistic comparison as doubling the number of scoring opportunities skews counting stats no matter how good/bad the player is. It’s apples to oranges. Nobody knows the exact figure, that isn’t really the point.
3.) Pace has gone up and down but it’s never came close to eclipsing 130+ possessions that it was at in those years. Remember the league was so primitive then, tons of rules have changed since, the NBA as a professional basketball league had only existed for 15 years. You’re saying that someone should have averages close, but pace has never been close, and players don’t play 45+ minutes anymore.
Anonymous:
I am not advocating the value of adjusting for "pace," so you don't have to convince me of the limitations of such an adjustment. One of the main points of my article is that the vast limitations outweigh any minimal value that might be derived from such an adjustment.
It has become popular in some quarters to mock the value of per game numbers, but the value of per game numbers is that a game is the basic measuring unit for basketball statistics--each game is a complete entity, and NBA games have always been 48 minutes long (except for games that went to overtime). A team that wins enough games qualifies for the playoffs, and the team that wins the most playoff games wins the championship. A player who consistently performs at a high level in games helps his team to win games.
Part 3/3
Wilt could play every single minute of a game played at today’s pace and it would only be 98 possessions. So even Wilt himself couldn’t come close to himself. If that makes sense. Saying “If pace were the only factor affecting individual scoring averages” misses the point. Pace doesn’t directly change scoring average, but the amount of shots do. If you can shoot more shots you can score more points. If two teams playing at the exact same level get a different amount of shots they will average a different amount of points. Scoring is dictated by shooting attempts, shooting attempts are dictated by how many possessions a team gets, the number of possessions a team gets is “pace”. You can’t physically score without shooting, a team can’t shoot if they don’t have the ball. Wilt’s 36 percent higher average was achieved largely because he had like 60 percent more offensive possessions to create scoring attempts from.
Look at your baseball example as a comparison. Batting average isn’t relative to attempts. If at-bats is 200 and a player hits on 100 his average is .500.. If at-bats are 50 and a player hits on 25 his average is also .500… But one player has four times as many hits. The hits in this example is ppg in the basketball example, not the batting average. And the amount of hits is determined by the number of at-bats. The second player could bat 1.000 and it wouldn’t matter, with being limited to 50 at-bats he can never reach 100 hits. The possibility doesn’t exist.
To your last point, “pace alone” isn’t the point I’m making. It’s like LeBron didn’t break the scoring record solely because he played forever. He broke it because he played forever while simultaneously being a great scorer. More than one factor can exist at the same time. But at the same shot attempts and FT attempts Jordan would have had to shoot 75% from the field to reach Wilt’s 50ppg. Players don’t shoot 75%. It’s not completely impossible like the baseball example but it’s highly unlikely because the opportunity for output isn’t even remotely similar. It’s not about attempting to discredit Wilt, it’s about “this player never averaged 50”, “this player couldn’t score 40+ 60 times in a season”, as if all those counting stats should be taken at face value and all those players had the same opportunity for output. It’s massively unfair to those players every time that argument is brought up, it’s not even remotely close to apples to apples.
Anonymous:
My point in this article is that it is, to borrow your phrase, "massively unfair" to Chamberlain to crunch some numbers and then conclude that Jordan's scoring average is superior to Chamberlain's scoring average. You made a point earlier that you don't make comparisons between eras because each era is so different. I think that some comparisons can be made, but the larger point is that you can't just take what happened on the court, put it in a spreadsheet, and then declare that Jordan is superior.
In today's game that is less physical and less focused on defense, Chamberlain might average 60 ppg. I am not saying that he would, but in an era during which teams supposedly hunt the most efficient shots what shot would be more efficient than Chamberlain catching the ball in the paint and dunking over undersized centers? Hacking Chamberlain would not work, because that would put Chamberlain's team in the bonus.
I am not saying that Chamberlain would average 60 ppg. The point is precisely that we don't know, and therefore Chamberlain's average stands as the record until someone beats it by averaging more points per game--not more points per possession, not more "pace-adjusted" points, and not more anything other than more points per game.
Post a Comment
<< Home