Stat Inflation

by Jeff Fogle 27. March 2011 23:42

It's one of the biggest migraines statheads have to deal with in all sports. To what degree is the statistical record a reflection of reality? To what degree is the record disturbed by pollution?

In football, quarterbacks can pad their passing yardage and TD totals when facing prevent defenses after they've fallen way behind. In baseball, hitting in Colorado guarantees stat inflation...while pitching in the poor visibility and heavy marine air of San Diego will guarantee ERA deflation. In college basketball, tempo...strength of schedule...when the head coach calls off the dogs in a blowout...(and probably a dozen others we could think of after watching March Madness this year)...can muddy the waters significantly.

In the NBA, stat inflation can be at the mercy of these elements:

*Tempo: which is why so many methodologies try to pro-rate to 100 possessions or other equalizers.

*Role on Your Team: which can wreak havoc in either direction depending on which stat you're talking about.

*Whether Your Games Matter or Not: meaning that it's easier to compile stats in games where one or both teams are just kind of playing out the string rather than battling with peak intensity. The lack of parity in the NBA leads to several teams that are out of the playoff picture by midseason.

There are a lot more we could think of. But, I want to stop right there because I think those are the three that are at the heart of the Kevin Love conundrum right now.

Kevin Love of the Minnesota Timberwolves is a hustling 22-year old power forward (who plays like a center) with a great future ahead of him. He certainly is somebody you'd want on your team. Is he one of the best players in the league? Is he THE best player in the league?

If you were watching Portland/Oklahoma City on ESPN Sunday night, you saw the scroll at the bottom of the screen listing John Hollinger's top five players in the NBA. Love ranked fifth out of everyone. The mainstream audience might find that odd. But, the stathead audience may not argue. In fact, Love ranks very high pretty much any which way you can measure anything. On the individual players page here at HoopData, Win Shares and Adjusted Win Shares actually rank Love as FIRST in the NBA. Best of everyone. Better than LeBron James, Dwyane Wade, Dwight Howard, and any player on either All-Star team.

It's worth noting that Love is right in the sweet spot for stat inflation.

Tempo: Minnesota currently ranks as the fastest team in the league, meaning that point and rebound production are not quite what they seem in their games. Some methodologies account for this influence. If you're the type to just see 20 points and 15 rebounds per game and draw a gut conclusion...be sure you're making a mental adjustment for Minnesota's very fast pace.

Role: Love's a key player on a horrible team...meaning he's an important part of the offense, and the man in the heart of the defensive scheme. The ball is going to find his hands a lot. The easiest things to measure in basketball are the things that happen when a ball is in somebody's hands. Also, current research on rebounding is looking into the role that choice plays within a defensive structure. If there's a ball that a few guys can get, do some players defer while others grab the ball? Is Love getting his share of rebounds, PLUS more than his share of the "optional" rebounds because his teammates just let him grab everything? There's a theory that this played a role in David Lee's big numbers in New York.

Meaningless Games: Minnesota's horrible. Their games don't matter. Opponents who need the win will usually get the win. They don't have to go all out for four quarters to take care of business. They can trade baskets much of the night, and have the one big quarter that gets the job done. Could Love be benefitting from the equivalent of a football's "prevent" defense much of the time? There's no reason to clamp down on Love if the game's already won...or if you know that you can trade baskets now and clamp down in the fourth quarter if needed.

I won't pretend to offer up certain answers tonight. I will point out what's happened to the players who have replaced Love (and his role) for the T-wolves the last three games while he was recovering from a groin injury.

Kevin Love's per game numbers:
20.3 points, 15.4 rebounds, 36 minutes per game (99 possessions)

Anthony Randolph vs. Dallas:
31 points, 11 rebounds, 36 minutes (96 possessions)

Anthony Randolph vs. Oklahoma City:
24 points, 15 rebounds, 37 minutes (101 possessions)

Anthony Tolliver vs. Boston:

16 points, 15 rebounds, 36 minutes (mid 80's in possessions)

Randolph started Sunday vs. Boston, but committed 3 fouls in 3 minutes of action. That brought in Tolliver to fill the role of the big man for the night...and he basically did what the others were doing after you adjust for tempo. Fifteen rebounds in a slow game is pretty darned good!

The three replacements pro-rate to roughly 25 points per game and 14 rebounds in 36 minutes using the 99 possession benchmark. Hey, it's only three games, and anything can happen in three games. Those three games do provide some evidence, though, that it's not that hard to compile Love-type stats if you're playing in the paint for Minnesota at this stage of the season. Even for horrible teams, somebody has to score the points in losses...and somebody has to grab the rebounds off shots the opponents are destined to miss.

Is Love truly one of the best five players in the league? Or, is he hitting clean-up for Colorado? Or passing for almost 4,000 yards on a 4-12 team like Carson Palmer did for the Cincinnati Bengals last year?

Fun player to root for. Extremely bright future. Would he strike you as one of the best players in the league if he played on one of the best teams in the league? It's a bit tougher to make that case based on what Randolph and Tolliver just did the past few nights. Let's see if we get additional data points before Love returns to the lineup.

Transition Points

*Boston beat Minnesota 85-82 Sunday night, blowing a huge lead before finishing off the job late. Once again, the bench was a problem for Boston. Nobody on the Celtics bench had a positive plus/minus even though they were facing a very bad team. All five starters were positive. Note though that Rajon Rondo didn't play because of a bad thumb. So, the team was even more shorthanded than usual.

*Houston and Miami played a wild one, providing some evidence for the "teams will trade baskets until it's time to decide a winner" line of thinking.

Miami led 38-37 after one quarter
Miami led 71-68 at the half
It was tied 93-93 heading into the fourth quarter

So much for Miami having a playoff caliber defense. Of course, we KNOW that they do have a playoff caliber defense...so you almost have to throw this one out when thinking about what will happen in the playoffs. It was more of a 125-119 scrimmage with a serious ending rather than a game matching playiff contenders.

*Kendrick Perkins is having an immediate impact in Oklahoma City. He had a plus/minus of +13 in 31 minutes on Sunday's 99-90 victory over Portland. Mentioned earlier that was an ESPN game. Great energy. It definitely had a playoff feel to it much of the night.

You get the sense Russell Westbrook is tired of hearing so much about Derrick Rose. Westbrook had 28 points and 7 assists. As I mentioned the other day, it's very hard to separate Westbrook from Rose in a meaningful way in the stats. The case for Rose seems to be that Westbrook has teammates who can score, so he doesn't have to carry all the load. Is a player's value based on what he accomplishes? Or, who his teammates are? Age old questions.

*Virginia Commonwealth had a great first half from behind the arc in a 71-61 upset of Kansas today. They join Butler in the final four as a team scoring more than 30% of its points from long range this season. I didn't think my "no #1 seeds in the Final Four" prediction was going to have a chance with Kansas such a heavy favorite. Congrats to VCU for rising to the occasion.

Three-Point Shooting Today:

VCU: 12 of 25
Kansas: 2 of 21 (!!!)

Kentucky: 12 of 22
N. Carolina: 3 of 16

Four Weekend Losers: 12 of 72 (17%)

All four games were certainly dramatic. I don't think I'll ever be a fan of the impact trey randomization can have in small sample sizes in this sport. The drama of watching a ball spin around a roulette wheel loses its zing after awhile. That can wreak havoc with efficiency data too...as a string of five blacks will look more efficient than a string fo three reds...but that won't influence what color comes up next.

There will be a mid-major in the finals for the second straight year because VCU and Butler are in the same half of the bracket. Gotta be driving the power brokers nuts!

Back late Monday to talk about Orlando/New York, Philadelphia/Chicago, Boston/Indiana, and maybe Portland/San Antonio. Manu Ginobili only played 19 minutes Sunday in Memphis after suffering a "left quad contusion."

Bookmark and Share DotnetKicks dotnetshoutout

Tags:

Comments

3/29/2011 1:25:03 PM #

Jimmy

One way of taking both pace and team and opponent field goal percentage out of the equation is to figure out the percentage of missed shots a player grabs while he's on the floor.  Given all the factors you mentioned (and many others), I question the real value of this stat but it could still give a little bit better indication than rebounds per game.

For players with at least 500 minutes burn this season, Kevin Love ranks 3rd in Total Rebound %, behind only Reggie Evans and Marcus Camby, and ahead of guys like Dwight Howard, Zach Randolph, and Blake Griffin.  (Interesting side note:  Kris Humphries and Pau Gasol are both at 10.3 rebs/game but Humphries plays 10 minutes fewer per game.  Humphries' TRB% is 22.0 -- good for 4th in the league -- vs. Gasol's 15.7 -- good for 50th.)

Still, Love's good #s might not be an accurate reflection of his ability but I think what's so impressive is that among the top 30 guys or so, he's 5th in True Shooting %, owing to his great 3-point shooting ability.  There's nobody else in the league who's even close to him in terms of both his TRB% and his 3P% (=42.3)!

Jimmy United States

3/29/2011 1:42:22 PM #

Michael

I would disagree with the comparison you are trying to make between what, combined, Anthony Rondolph and Tolliver have done the last three games to what Kevin has done this whole season. which is to average 20 and 15. in no way can you draw a parrellel from only three games to entire season. In november Darko had three games in which he had scored 23, 21 and 22 points. All that now points to is Darko had three really good games in a row, nothing more. if it was easy to to average 20/15 on a bad team, then whith all of the bad teams over  last 20 or 30 years why hasnt it been done? even some of the bad Timberwolves teams that KG was on, who is a very good rebounder in his own right, never did what Kevin Love did.

Michael United States

4/2/2011 6:25:32 PM #

Jeff Fogle

Thanks for the comments guys. Agree that Love is a great talent with a bright future. Regarding drawing parallels...the point isn't that other guys can have good games that look like Love's numbers in the stats. It was that...when Love was out with an injury...the guys that became his cog in the team wheel posted stats in line with what Love usually gets. Even accepting Love's unique and meaningful talents, it's tough to know what his cog in the wheel would normally get on an up tempo lousy that defers to one guy to get loose rebounds that could be grabbed by anybody.

Definitely not easy to be a 20/15 guy in any circumstances. Not suggesting he's a stiff who's only posting good numbers because of stat inflation. Just that there's potential that such inflation is influencing perceptions. Is he the best player in the league, as some metrics say? Is he 5th best? Or, is he 20th to 30th best but getting a boost. Is he 40th best and getting a big boost. If everyone became a free agent today, and there was a redraft before the playoffs, where does Love get taken? what would his impact be on a playoff team...in the playoffs...at 90 possessions per game...rather than on a horrible team at closer to 100 possessions per game?

KG was on very few bad Minnesota teams. They were 40-42 the year he became a starter, then were .500 or better over the next 8 years through his prime. Pace was different. And, we don't know how the team handled the coin flip rebounds (defer to KG or just grab it yourself). Still a lot to work to do in terms of understanding diminishing returns and the internal dynamics of teams. Not saying Love is a stiff. He's a special talent. Hope nobody's thinking "he's better now than KG ever was" because of the rebounding stats though.

Jeff Fogle United States

4/6/2011 10:36:59 PM #

Jamie


I had one problem with this article.

You wrote "It's one of the biggest migraines statheads have to deal with in all sports. To what degree is the statistical record a reflection of reality? To what degree is the record disturbed by pollution? "


I assume that you were using the term " Migraine " to explain the ferocity of a headache.

Migraine is a genetic neurological disease, which can have a headache as a symptom of a Migraine attack. However, the headache stage is just a part of a Migraine,  nausea, vomiting, temporary loss of  motor skills, visual auras, including losing parts of your vision, hearing things that aren't there, etc.

When you use the word "Migraine" so casually, you're shortchanging your writing and your readers- some of whom may themselves be Migraineurs. As a writer, you know how priceless word choice is in conveying your thoughts to your readers. If you didn't mean to  imply that the statisticians all had Migraine disease, then the word "Migraine" did not belong. Stating that they had "awful headaches" would have made the same point.

Respectfully,

Jamie United States

Comments are closed
Archive | FeedSubscribe | Log in