I know I'm going to regret writing this one on my 45 minute train ride, because I won't have time to research good enough examples, but I've found in the comment boxes lots of arguments over player evaluations which rely on a faulty use of statistics. In this article, I thought I would try to spell out how I use various statistics when grading out players and looking at their performance.
This isn't meant to take the place of actually watching games. I don't feel any player can be properly evaluated solely by their statistics, but when making an argument about whether a guy is a chucker or a scorer, there are simple measures that can objectively show which one of those is true.
Before anything else can be said, I will point out the best places to get your stats, at least that I know of:
Basketball Reference - good for almost all stats
82games.com - has interesting floor time stats, clutch stats, and adjusted +/-
Now before we get into the next section, let's discuss what it takes to win
a basketball game. In the end, you only need one thing. It can be
said in many ways, but you need to have more points than your
opponent. One way in which I like to say this through basketball
statistics is that in order to win you need greater offensive
efficiency or greater defensive efficiency than your oppponent.
The most additional possessions one team can have vs their opponent in
a game is one. If they get the starting possession and the ending
one. Whether you win or not depends whether you score more points on
your possessions. I know this is obvious, but it's important. It
should drive home the point that it's not how many points your team
scores or holds the other team down to. It's how efficient you score
yoru points or how inefficient you make your opponent.
Some teams employ strategies to speed up or slow down the game. One
ngiht in a fast game you might score 110 and in a slow game score 85.
It doesn't mean that the 110 was a better offensive night. It's a
better offensive night only if you were more efficient with your
possessions on that night.
Thus, in basketball offensive and defensive efficiency is everything.
The rest is smoke and mirrors. Gaudy totals are meaningless if they
aren't creating defensive or offensive efficiency.
In my final precursor, let's talk about per minute statistics vs per game statistics. I used to not be a believe in per minute stats until someone put it to me this way. If player A played 30 minutes in game 1 and scored 20 points and put up 10 rebounds, and his backup player B played 15 minutes in game 1 and 15 and game 2 and scored 20 points and grabbed 10 boards then they both played 30 minutes and had 20/10 over that stretch.
Is player B really playing worse? It's been shown that per minute stats tend to hold up with more minutes. In fact, players with good per minute stats actually tend to improve by playing more minutes. This is likely only true for young players because they're still low on experience and high on energy. Thus playing more time allows them to improve while not suffering from conditioning issues.
There are exceptions to these rules, but in general, when looking at stats, make sure to look at per minute stats.
Individual stats that are useful:
Basketball Reference stats
Player efficiency rating - PER
My favorite overall metric is probably PER developed by John Hollinger. PER gets grief for not always being perfect. It's not. However, if you said I could only choose one stat to sum up whether i wanted a player or not I would choose PER. It's pace adjusted, minutes adjusted, and encompasses all other statistics.
A 15 PER is set to be the exact league average. Why Hollinger choose 15, I'll never know, but PER is based on the total basketball statistics of that year. A PER of 15 this year will not be the same as a PER of 15 last year. It depends on the quality of basketball in each individual season.
Thus how much higher than 15 you're playing is how much better than average that you are.
I think my primary fault with PER is that, in my opinion, it may overvalue rebounders and usage a bit too much. Reboundign is important, but the problem is rebounding is based so much on scheme. It's my feeling that there are far more uncontested rebounds than uncontested dunks in the game of basketball.
Thus players who play in the right system at the right possession collect more free rebounds than anyone else is able to collect free shots or other statistics.
True Shooting Percentage - TS%
True shooting percentage is a sign of how efficiently a guy racks up points when he takes a shot. It's esay to say that player A is a chucker because he shoots 42%, but if he gets to the foul line constantly and shoots tons of threes, then he might be giving you far more points per shot attempt than a guy shooting 52%.
This is one statistic which shows how good Ben Gordon really is.
There are two primary flaws for TS%. First, it doesn't factor in turnovers, so a guy who throws the ball away like crazy will have an inflated TS% relative to his real offensive efficiency. There's some good and some bad in that, because a PG who handles the ball shouldn't have his scoring ability taken away from turnovers generated trying to pass the ball. However, a guy who's making turnovers while trying to score should.
The second probably with TS% is that it doesn't take into account volume of shots. Guys like Tyson Chandler or Joakim Noah can put up great TS% numbers because they're taking a very high percentage of open shots.
TS% is great for comparing the overall offensive efficiency of similar volume players with similar turnover numbers though. Points per possession where a possession is defined by a turnover, trip to the foul line, or shot attempt is a better statistic, but I don't think it's easily found anywhere.
Usage Rate
Usage rate is simply what percentage of the time a player is on the floor that he is the one taking the shot. A high usage rate is not necessarily good or bad, but it's useful in understanding how it relates to other statistics.
For example, Michael Jordan had a huge usage rate, and that's a good thing for him. He's a dominant player and the team was helped by him taking 1/3rd of his teams possessions over his career. Zach Randolph also has a high usage rate, and that's not such a good thing for his team, because Zach Randolph doesn't have a great TS%. His career average hovers around the league averaeg of 51-52%.
I like to look at usage rate and TS% combined to get a feel for how good a scorer a player is. A guy who has a high TS% (over 55%) and a high Usage rate (over 25%) is typically a tremendous scorer. A guy with a high usage rate and a TS% at 52% or under ist typically over utilized in the offense. They may be a chucker, they may be the best option on their team.
Rebounding Percentage
Total rebounding percentage is a great measure of how good a rebounder actually is. If calculated correctly, it shows what percentage of total rebounds available that a player got while he was on the floor. Many sites use estimates for it because they aren't looking at floor time statistics, but game statistics and assume that an equal amount of rebounds were available (per minute) when a player was on the floor as when he was off.
Either way, rebounding percentage is the best way to see whether a guy is truly a great rebounder or whether they're collecting their rebounds merely by playing a lot of minutes and playing in a rebounding position. Obviously, you have different expectations for rebounding percentage based on where a guy plays. A center or power forward should be dominating this category relative to a guard.
Some rebounding percentage notes relative to the Bulls last year: John Salmons struggled with rebounding for the Bulls, particularly in the playoffs. He gave us a lot of scoring, but we really missed Luol Deng's ability to rebound the basketball.
Tyrus Thomas was a good rebounder his rookie year, but he's declined ever since then.
Assist Percentage
Assist percentage measures what percentage of your possessions lead to assists. This stat has limited use in that it really only helps me look at point guards, however, you can look to see how big of a ball hog any player with high usage is. If a player has low usage, than I would ignore his assist percentage all together as he's likely just finishing off plays rather than operating with the ball in his hands.
Derrick Rose's assist percentage was low for a PG. Now, it's a stat that can improve considerably with time in the league, but his 28.8% was lower than Kirk Hinrich back when Kirk Hinrich played the majority of his time at PG. It's radically lower than your great PGs in their good years (Nash and Paul put up over 50% numbers).
It's not something to worry about, but it's something to consider when discussing how much Derrick Rose makes his teammates better. That certainly wasn't born out in the numbers last season, but again, as a young rookie, that's not something to panic about yet. Nor is it something that might ever be his strength. I don't think Rose is the type of guard to ever have an elite assist percentage number, because he'll always be somewhat more of a scorer.
I'm more concerned with him raising his TS% than his assist%, but a prime Derrick Rose should get up to a 57%+ TS% and a 35% or so assist percentage when he's always making the right plays.
82games.com stats
Most of 82games.com stat's are tricky to use. I like to look at them, because they are so unique, but ultimately it's hard to decide how much to gauge out of any of them.
Adjusted +/- shows how well the team performs per minute with you on and off the court. It attempts to remove the unbalance of NBA.com's +/- which just shows how the team does with you on the court. The problem with that is teams that always lose will always have a huge - for any of their players who play a lot of minutes.
Adjusted +/- compares how well the team does with a player on vs when the player is off. Thus if a team is playing even worse with a player off he can have a positive net +/- even if the team is losing consistently with him on the court.
Like many stats +/- can usually pick out the great players, however, I think it's value on other players is often of little use. It can be influenced heavily based on which teammates you play with, how good your backup is, which players you typically go against (starters vs backups), and many other factors.
Andres Nocioni is the poster child of why this stat can be terrible. In his best years, his +/- was horrible and in his worst year it was actually positive. Either that, or Andres Nocioni is the poster child of why I'm terrible at analyzing Andres Nocioni's years.
Clutch stats
The clutch stats show how a player plays down the stretch of a close game. It shows who can deliver when it counts, and who the coaches trust. Overall, I think the sample size of this data is often too small to be of great use, but you'll find that Gordon has been a consistently good clutch player while Kirk Hinrich has been a scary bad one at times.
82games.com has many other interesting statistics, breaking down how players perform vs good teams, average teams and bad teams. You can find out what percentage of jumpers, inside shots, and dunks guys get. They breakdown turnover types into passing, ball handling, and offensive fouls. They have info on how players perform in pairs or lineups on the court as well.
In general, there are many things you can find at the site, but because the typically don't allow you to bring up these stats in sortable lists, they're difficult to use for overall analysis.
Hopefully, this has given those of you who aren't statistical experts some additional understanding of how to look at basketball statistics when making your player judgments, or helped you understand how I arrive at some of my player judgments.
Filed under: Uncategorized
Tags: 82games.com, basketball-reference.com, chicago bulls, PER, statistics, TS%, understanding statistics, usage
Doug, this post almost seems like a direct response to one of my replies from yesterday. Good analysis of some of the most common stat packages. I am old school and trust my eyes more than someone elses stats, many of which are empty.
However, I read a pretty detailed NY times article on Daryl Morey, the Houston GM, and I don't think that they used any of these stats strategically.
To me TS% is massively biased toward 3 point shooting, and essentially proves that the 3 point shot is too easy for NBA players. Since it is virtually impossible for anybody other than dominant centers/post players to shoot over 50% from the field, the target league average for 3 pointers should be 33%.
However, if player A shoots only 2's and hits 50%, and player B shoots only 3's and hits 33%, player A is a starter, player B is out of the league.
A 40% 3 point shooter is a 60% TS%, lots of guys can hit 40% even 45% and sometimes 50% from 3. Nobody but Wilt has ever even dunked his way to 60% shooting, especially with volumn.
Jordan by the way shot over 50% on 2's and about 33% of 3's, much better than Kobe on 2's, slightly worse on 3's, yet in comparisons people try to claim that Kobe is the better shooter.
While I enjoy the 3 point shot, it has to some extent ruined the game. Everytime that I go to the gym, every litlle kid and his brother is shotputting up shots from the 3 point line rather than working on proper form with shots that they can make. Probably where Noahs shot came from.
As for Rose vs Gordon comparisons, only one matters, would anyone on any planet in any universe select Gordon over Rose after only one season in the league, I believe that would be a TS(selection)% of Zero.
Personally, I think that you massively undervalue rebounding, which I vastly prefer to assist numbers. To me the most important thing in the game is the rim, attack it on offense and defend it on defense, find me a stat for that and I'll show you a champion.
I would almost certainly bet the the team with the best TS% has never won an NBA championship, unless it was Wilts 67, 76's or maybe Kareems Bucks(someone in the west(Nuggets) probably beat them out). Now TS% allowed probably has multiple titles.
Pretty much why Gordon and his flashy TS% is as useless as tits on a whore, as the old saying goes. Gordon has stated, more than once that the purpose of the game is to put the ball in the hole, which is why I don't want him on my team, the purpose is to win, just ask Herm Edwards.
Never-the-less a good discussion, even if to me it feels like a not so slightly veiled defense of Gordon, who I hope we can stop talking about soon.
Excellent point, Gordan is gone and we will see if Rose,
Salmons and Hinrich can pick up scoring. Gordon handled the pressure of being the go to guy, will his replacements do the same?
Rose and Salmons are good players and Hinrich is fair at best. Will they put up the numbers and perform at the end of the game? Rose is not an all star but he has great promise!
Hinrich is a career inconsistent streak shooter who shoots in the upper 30% range and manages to get over 40% by the of the year. Will coming off the bench help or hurt him?
Hinrich's assists plummeted last year subtracting his turnovers from his assists, he netted 2.2 assists per game.
Throw in Hinrich's circle dribbling and clock eating and his inability to beat anyone off the dribble and he is the weak link!!!!!!!
I honestly don't know why I bother. I tell myself I'll just ignore your continual ranting about Kirk and/or Joakim, and then it's like I'm physically unable to let it pass.
Of course Kirk's assists dropped last year - he spent most of his time playing off the ball. Why you think it's fair to compare years when he started at PG to a year that he backed up both guard spots, I'll never understand. He's an OK shooter - not great, but far better than you're trying to paint him. He's been well over .400 2 of the past 4 years, and even in the other 2 years it's not like he was in danger of shooting below that or barely made it over the threshold.
And repeating the same thing time after time does not make it true.
I feel your pain, obviously Alex has been trained by the trial lawyers association and the DNC.
Yes it does, Yes it does, Yes it does, Yes it does.
NaNa, NaNa, NaNa
If Rose, Salmons and Deng average 18 ppg that will essentially equal what Gordon, Salmons and Rose averaged last year, and we didn't even have Salmons for most of the season. There we don't miss Gordon already, see how easy that was.
And I expect Hinrich to significantly increase his scoring simply by being healthy, getting more minutes and shots now that Gordon is gone. Now we are ahead, and we haven't even accounted for the rest of the roster, or defense.
By the way, I think that logic would dictate that you can only have one "the weak link".
My mistake, expecting logic and Alex to coexist.
Bigway, why did Vinnie the trainee start Rose and Gordon over the slow circle dribbling clock eating Hinrich? Tell us what is behind Vinnie's choice of Gordon over Hinrich ????????????????
Deng can be a better than average player if healthy but he still must learn to fit into Vinnie's offense. How many games has he missed since he was with the Bulls and how many do you think he will miss this year? Do you think that Deng is fragile?
If Deng goes down again, who should take his spot??
Too bad that they can't put a capability (CPK)calculation on all the data acquired each player and use this number to access a player's worth.
The cold hard facts are usually more relevant, Look at the last deciding playoff game:
Thomas 17 min, 4 floor points/2 Ast, 5 rebounds, 2 PF, 2 Blk "-3"
Miller 28 min, 6 floor points/2 Ast, 7 rebounds, 6 PF, 2 Blk "-8"
Noah 32 min, 2 floor points/1 Ast, 15 rebounds, 6 PF 2 Blk "-9"
The Celts out scored the Bulls 46 to 18 in the paint. Where was the defense? Where was the scoring? These three big guys were pathetic.
Maybe Thomas wakes up and steps up! Miller is as good as he will ever be in a Bulls uniform and Noah is just a junk rebounder and he does not have any offensive skills plus he is too lazy to get stronger.
Maybe Deng, Johnson and Gibson will make a difference this year!
Doug,
I enjoyed your analysis and I know you really like Ben Gordon as a player. However, I have one question for you.
"Was Ben Gordon's selfishness on and off the court good or bad for the Bulls organization?"
Personally, I find that it was more hurtful than helpful. I think the Bulls organization feels the same way or else he would still be here. They will be fine without him.
I'm an eye guy myself, but found this interesting. I think both statistics and ones eyes really tell the while story. Anyway, curious what you think about assists and the offense the Bulls run. It seems to me Rose is immediatley passing the ball away much of the time in half court sets. That is unless the Bulls are running pick and roll, Rose passes the ball to a wing. I would estimate this happens more than half the time. So how would this effect his assist numbers? Conversely Kirk almost always had the ball in his hand under Skiles.
Doug your frequently point to the stats and this helps to define a players capability and his pay scale. But it does not tell the whole story.
In order to improve, you must be evaluate your talent and make changes. The last Celts game completely defined the Bull's front court. Based on their defensive performance Thomas, Noah and Miller are not NBA starters, their offensive performance was a joke!
If Garnett was playing the Bull's would have probably been swept and OUR high hopes for the Bulls would have
dashed! Perkins and Davis embarrassed the Bulls!!!!
I didn't understand why after losing Gordon, why they didn't draft a guard. The last celts game was a big reason why the Bulls drafted two forwards.
Thomas has the potential to get better, if doesn't Gibson and Johnson will push him out the door!
Miller is limited but he can still play. Noah is a LAZY weakling who didn't learn anything from getting pushed around on the court since he got here. Every one talked about Noah bulking up and getting stronger that is a joke. Noah is the weakest link in the front court, the Bulls need every point that they can get! They can't afford a Ben Wallace or a Dennis Rodman type, the team is just not good enough.
Wallace and Rodman worked out to get stronger, watch No-Shot get pushed around the paint because he is a weakling.
Maybe No-Shot Noah trains with Zambrano another LAZY weakling????????
Doug, you are a great stat man:
"There is not a guy who scores more points on less possessions in the league than Gordon last year."
Thanks for the info. How many Gordon haters knew this fact.
Gordon is gone, now we will really see if Rose, Salmons and the clock eating attack killing Hinrich will step up!
Interesting how easily you accept a statement that may or may not indeed be a fact when it concerns a player that you like and ignor mountains of evidence about players you dislike.
If scoring were indeed the ultimate point of the game(as Gordon believes) then this stat might be usefull, but winning is all that matters.
I challenge any and all stat geeks to find a single stat that suggests that Gordon contributes anything more to winning than does flipping a coin.
Gordon is fearless and tough!!!He wasn't afraid to drive to the basket or take a clutch shot. He did his job, he scored! He was not good playing defense but he came up big at the end of many games!He took the heat off many of his team mates!
Rose is not good at playing defense but he "may"
improve. We will see if Rose can make the clutch shots this year. Will Rose be an all-star, maybe? But not this year. He and Salmons must learn to play together.
Hopefully Salmons steps up and scores enough to make us forget Gordon. he is a key player and he appears to be tough, will he be a clutch player like Gordon?
I am so dizzy after watching Hinrich dribble in circles that it is hard for me to express a positive opinion. Actually, he is nothing more than an overpaid backup! But the genius Paxson,
does not have anyone to compete with Hinrich and the third guard spot goes to this clock eater. Hinrich will miss Gordon more than anyone! After dribbling away the clock Hinrich will not be able to pass the ball off to Gordon with 3 seconds left!
Yes Doug,
Discussion begins with the ability to understand
the facts, earlier you stated that the Bull's front court is not good! Are you confused?
The verdict is in, No-Shot Noah is lazy and weak and he doesn't have any offensive skills! What don't understand about those facts.
It us hard to communicate with some one who can't grasp reality. According to Stacey King Joakim has good games against garbage teams! Is you hearing defective?
May be your are disoriented, have you been watching film of Hinrich dribbling in circles again.
In that case, how many circles a minute does Hinrich dribble when he is on the floor????
Stop making excuses for the Bull's lack of talent. Try talking about the positives that Gibson and Johnson bring to the team! Stop defending slugs like Hinrich and Noah the NBA has their number! Every one knows what they bring to the Bulls and it isn't much!!!
Couldn't have said it better myself, which some people might suggest isn't saying much.
Noah and Hinrich are flawed but quality NBA players who would have a role in the 7-8 man rotation of any team worthy of being a champion.
I am sure that Michael, Scottie and Phil would have welcomed both of them with open arms.
Totally agree on Rose. In fact I expect more trips to the free throw line to be one of the biggest improvements in Rose's game. Hopefully it was just a rookie hazing thing last year.
I find the PER stat interesting in a general sort of way but perplexing when I start to compare individual players.
Have you read Hollinger's 09 PER projections. might make for an extensive debate. Among many oddities is that it seems like he is projecting a decreased PER for about 90% of the players, which suggests that he thinks that the quality of play in the entire league will decline noticably.
He has Rose, Noah, Hinrich, Salmons declining, Deng barely better than his worst year ever, and Thomas of all people becoming our best PER player.
That certainly makes me think that the entire system is flawed. Based on what my eyes tell me from watching every game last season I expect Rose(natural growth), Noah(natural growth), Hinrich(injury), Deng(injury) and even Thomas(unatural growth) to improve. Salmons likely stays about the same, although playing the whole year with a better team could improve his rating also. A Decline by Miller would not shock anyone.
when looking at the rest of the league he has guys who can't even get regular playing time ranked in the top 10-20.
whoops almost forgot to ask.
Isn't every bad defensive possesion essentially the equivalent of a turnover. Actually it is worse, after a turnover, most of the time you still have the opportunity to make up for it with good defense.
Certain if not most players have way more bad defensive possessins than anybody does turnovers. Adding this into the equation would indeed be obscene, at least for these players, and their contracts.
Dougie, you stated that the Bulls front line of Noah, Miller and Thomas were not good!!!!!!!!!!! You dont remember what you post do you or is it your comprehesion that fails you?
As configured the Bulls could not afford
Ben Wallace and they dumped him. Rodman was with a great Bulls team. This is not a great Bulls team, have you noticed?
According to the PROFESSIONAL Stacy King, No-Shot Noah only shows up against weak competition! What don't you understand about that accurate opinion.
Since Noah got here he was in NBA shape for less than 1/2 of a year, out of 2 years that he has been here. Joakim's inability to bulk up again in the off season proves that he is lazy and anyone with eyes and a brain can see that he is a weakling without offensive skills!
You comment that Hinrich over dribbles at times proves that you don't watch many Bull's games. Hinrich is a clock eating attack killer who can't pass off the Gordon with 3 second on the clock!
Dougie your double speak is an insult to the posters here. Knock off the tantrums along with the Bull's PR!
Doug your post is 110% correct and I agree with averything you said. Everything!
My favorite targets are Noah & Hinrich. They are not getting any better, ever!
Noah could have improved his presence in the paint in this his third season! But did not bulk up, adding strength and stamina which would have proved that he really wants to win and contribute to the Bulls chances of being winners. Joakim is one lazy weakling and probably not in real good shape. Look at his track record.
The Bulls would be better off with Gibson and Johnson playing instead of Noah. They can rebound and they have skills and they need minutes to grow.
Hinrich is a back up. The Bulls don't have any one to step into his spot as a third guard and that is too bad.
Good teams quickly move the ball, penetrate and drive to the rim.
Hinrich is just to slow, which was all right for Skiles. One on one, he is not a scoring threat!
I will bet that his game suffers even more this year unless he find some one to dump the ball off to in the last 3 seconds after circle dribbling and eating the clock.
Yes, Hinrich circle dribbles so often that he make is team mates and Vinnie the trainee DIZZY!
People ignore the fact Gordon did the job that was assigned to him, score. He did a good job!
By the way, exactly how much defense did Rose play and how many clutch shots did Rose hit? Rose will get better, when is the question?
Do you really beleve that under achievers like Thomas & Deng will pick up where Gordon left off??
Do you really think that a lazy guy like Noah will help! Or how about the circle dribbling Hinrich who can't beat any one off the dribble nor create his own shot will make a big differnce??
Hoope lies in the hands of Vinnie getting better along with major minutes for Gibson and Johnson!