»
S
I
D
E
B
A
R
«
Tracking Defensive Plays in the Magic’s 106-98 Victory Over the Pacers
Dec 15th, 2009 by Jon Nichols

The following is part of a weekly series at the Orlando Magic blog, Third Quarter Collapse.

For my previous piece on the Magic, I charted hustle stats such as deflections, loose balls, missed blockouts, etc.  While these things are all important, perhaps the area in which hustle is most important is defense.  Although it takes more than just good hustle to be a good defender (as a certain Mr. Howard will show us later), effort is one of the keys to being a good defensive team.  Therefore, I decided to track defensive plays in last night’s contest between the Magic and the Pacers (this time, I only kept track of the Magic’s stats).  I imagine most (if not all) NBA teams track these on their own, as well as companies such as Synergy Sports.

To see the numbers, click the link below:

http://spreadsheets.google.com/ccc?key=0AvNKNGJ_AHijdFVaanJkQ2xSSG5DU3ZNcUVPRHpfSkE&hl=en

The rest of this article will explain what those numbers mean.  I will also provide a few observations and notes about the contest.

The first column is “Forced Misses.”  This is pretty self-explanatory, although I should explain a few things.  First, forced misses don’t only occur on an individual’s man-to-man assignment.  A help defender that forces a missed shot would receive the credit.  Second, I conservatively rewarded a few forced turnovers as forced misses.  These were situations in which a player caused his opponent to turn the ball over (through traveling, bobbling the ball, etc.) by applying good pressure and staying in good position.

The second column is “Baskets Allowed.”  This is also self-explanatory, with one caveat.  If a player made a bad defensive play that eventually led to someone from the opposing team scoring, he was the one credited with an allowed basket.  For example, on one play, Jason Williams allowed his man to penetrate into the lane with ease, forcing the Magic to help and rotate.  The Pacers swung the ball and ended up with an easy three-pointer.  Although Williams’ man did not receive any points or assists for the play, Williams was penalized for allowing the basket.

The third column is “Good Help D.”  This occurred when a player came off his man to either help a beaten teammate or to make a good play such as causing a turnover.  When Dwight Howard met the opponent in the lane and forced a miss, he was credited with a “Good Help D.”  When Ryan Anderson strayed from his man for a second and reached in and knocked the ball loose from another player, he also received credit for good help defense.  As you can see by now, many of these statistics are subjective (which is both a great thing and a bad thing).

The fourth column is titled “BB/MD.”  This stands for blow-bys/middle drives.  This occurred any time a Magic player allowed his man to drive right past him without the use of a screen (in certain cases, when there was a switch on a screen and the new defender allowed the opponent receiving the screen to drive by, a BB/MD was assessed).  A BB/MD did not have to result in a made basket to be counted.

The fifth column is titled “Lost Man.”  This was recorded every time a player failed to stay on his man, resulting in score.  This occurred most frequently in one of two ways: either a player simply wasn’t paying attention and allowed off-the-ball movement (such as a backdoor cut) for a score, or a player failed to chase his man quickly enough through screens.

The sixth and final column containing raw data is “Silly Fouls.”  While obviously the most subjective of the six categories, it was generally pretty easy to determine.  Fouls that occurred off the ball and away from the play were the biggest culprits.

The rest of the data is computed based on those six categories.  I included each player’s minutes played to serve as a reference point.  “FG% Allowed” was calculated as follows: Baskets Allowed / (Baskets Allowed + Forced Misses).  This statistic does not mean the field goal percentage of the man-to-man assignment.  Because Forced Misses and Baskets Allowed are not always credited on a man-to-man basis, FG% Allowed is a bit more complicated.   Obviously, a lower percentage is a better percentage.

The final six columns are the six raw categories calculated on a per-minute basis.  Like last time, I calculated them in the form of “minutes per stat” as opposed to “stat per minute.”  This is to avoid presenting very small numbers.  For positive stats such as Forced Misses and Good Help D, a lower number is better (in other words, a player achieves these stats more frequently and therefore in less minutes on average).  For negative stats such as Baskets Allowed and BB/MD, a higher number (or no number at all) is better.

Finally, some observations:

  • As amazing as this might sound, Dwight Howard’s impact was possibly even stronger than the box score indicates.  He altered a ton of shots in a night in which many of his teammates seemed to struggle defensively.  He was able to help his teammates and alter the Pacers’ offense considerably.  It didn’t hurt that the opposing centers weren’t looking to challenge him very often.
  • Other than Howard, Anthony Johnson and J.J. Redick had the two best defensive performances.  Despite his age and perceived lack of speed compared to T.J. Ford, Johnson was able to play solid one-on-one defense and force a lot of misses.  It’s not surprising he ended up playing 27 minutes despite coming off the bench.  Redick, although not challenged as often as Johnson, also made a few good defensive plays and forced some missed shots.
  • The only blemishes on Redick’s defensive stat line were the silly fouls.  He certainly did not agree with the calls, but three times he was called for defensive fouls away from the ball while fighting for position.  He was the only Magic player to record a silly foul.
  • The man Johnson replaced, Jason Williams, struggled on the defensive end.  He had a difficult time preventing Ford from penetrating and generally played poor pick-and-roll defense.  The Pacers built an early lead partially by attacking him.
  • The Magic’s two highest-paid players were mediocre at best on defense.  Vince Carter’s performance, which wasn’t terrible, can possibly be excused because of his great offensive play.  But Rashard Lewis complimented his offensive struggles with some defensive struggles.  He allowed seven baskets while only forcing three misses, and he let his man blow by him into the lane four times.  Lewis made some defensive strides last season, and he must be careful not to relapse this year.
  • Ryan Anderson struggled a bit on defense, although not because of a lack of effort.  He allowed five baskets (second-most on the team behind Lewis), but did have a nice steal.
  • Marcin Gortat won’t be happy about the measly six minutes of playing time he received, but he was generally pretty effective on defense when he was on the court.

I enjoyed tracking the individual defensive performances of the Magic, and I think this type of information sheds a great deal of light on what’s going on beyond the box score.  Although I don’t know what Synergy Sports or the Magic track defensively about their players, I imagine data such as this is of interest to them simply because it’s so useful.  I plan on doing this again soon.

New “About Me” Page on the Left
Dec 15th, 2009 by Jon Nichols

Just some personal information that I thought should be out there.

-Jon

Tracking Hustle Plays in the Magic’s 126-118 Victory Over the Warriors
Dec 10th, 2009 by Jon Nichols

The following is part of a weekly series at the Orlando Magic blog, Third Quarter Collapse.

A week ago I tracked the hustle plays in a game between the Los Angeles Clippers and the Memphis Grizzlies.  Tracking hustle plays is presumably something most, if not every, NBA team does.  After all, box scores are pretty limited.  Even if we use the play-by-play data to do thorough analysis, it still doesn’t include things such as diving for loose balls, deflections, missed blockouts, etc.  But teams would like to know these things, so they must track it themselves.

I decided to track the hustle plays during last Saturday’s game between the Magic and the Golden State Warriors.  During the game, I kept track of five things.  First, I tracked players going for loose balls.  In my experience with a college team, we only record plays where a player dove for a loose ball.  But since this is the NBA, and effort is often lacking, I include all plays in which a player ends up with the ball, regardless of whether or not he dove.  A second thing I track is drawn charges.  You can somewhat glean this from the play-by-play data, but it is much easier to just record it yourself.

Thirdly, I kept track of good sprints.  I define these as plays in which a player creates a play for himself or others by sprinting the floor and forcing the defense to adjust.  For this game featuring the fast-paced Warriors, I had to be more selective in my criteria or else we’d have a lot of good sprints.  A fourth thing I tracked for this game was deflections.  This is relatively easy to define and track.  Basically it includes any deflection that is not recorded as a steal, rebound, etc.  Finally, I kept track of missed blockouts.  These were most noticeable when they led to an easy offensive rebound, and they were much more rare in this game than in my first one.

Of course, these aren’t all the hustle plays that players can make.  Traditional box score stats such as offensive rebounds and steals often reflect hustle plays.  Defense is also largely a product of effort, but that is something I will track another time.

Below is a link to a spreadsheet that contains the hustle stats for the Magic-Warriors game.  On the left side of each tab is the raw numbers.  On the right side is the per-minute numbers.  Instead of presenting them as “statistic per minute,” they are presented as “minute per statistic.”  I did this because the numbers are so low.  As it turns out, this method is not too difficult to grasp conceptually.  For positive statistics such as deflections, a lower number is better (a blank number means the player did not record any deflection at all, which obviously is bad).  For negative statistics such as missed blockouts, blank numbers are the best and low numbers are the worst.

http://spreadsheets.google.com/ccc?key=0AvNKNGJ_AHijdHo3UEZHLUtnMVlUY2NDQ1Q1MEpSa0E&hl=en

I have a few observations about the data:

  • Marcin Gortat, Brandon Bass, and Matt Barnes showed great effort off the bench.  All three were active on the boards and had active hands.  Per minute, they were all special in terms of the number of deflections they had.
  • The Magic did not draw a charge all game, and they certainly had some opportunities.  Vince Carter in particular had three opportunities in which he failed to draw a charge.  In the first, he was just late and recorded a foul.  In the second, he turned his shoulders and avoided contact (Monta Ellis was coming like a speeding bullet).  In the third, he watched Ellis blow by him.  A charge on any of the three would have been an excellent play, but they certainly were possible.
  • Overall, though, Carter was not lazy.  He recorded a few deflections and went for loose balls.
  • Dwight Howard created a few opportunities by running the floor.  Barnes had an easy dunk after he sprinted down the floor after a dead ball and caught the Warriors sleeping.
  • Anthony Randolph had a great game and killed the Magic with his effort.  He crashed the boards, ran the floor hard, got to loose balls, and even drew a charge.  My only complaint was the amount in which he flopped.  He was forced to guard Howard for extended periods of time, though, which is something he shouldn’t be asked to do.  Perhaps he figured flopping was his only option.
  • Vladimir Radmanovic showed more hustle than I thought he would, but his defense certainly wasn’t spectacular at all times.
  • Except for a mistake by Stephen Curry, both teams did a good job of putting a body on each man when going for the defensive rebound.

The Magic did not win this game because they outhustled the Warriors.  In terms of effort, both teams were solid and about even.  The Magic won, obviously, because of a huge run late in the fourth quarter in which they hit their shots and the Warriors made silly plays.

Up next I’d like to track the defense of the Magic.  With a few games in the data set, we may be able to rate the defense of Magic players in other ways besides Defensive Rating, plus-minus, etc.

Comparing the 08-09 Seasons of Vince Carter and Hedo Turkoglu
Nov 9th, 2009 by Jon Nichols

The following is part of a weekly series at the Orlando Magic blog, Third Quarter Collapse.

Making major changes to your team when you are already very, very good appears to be the thing to do in today’s NBA.  The Lakers essentially swapped Trevor Ariza for Ron Artest, the Magic swapped Hedo Turkoglu for Vince Carter, and the Cavaliers added Shaquille O’Neal.  Each of these teams was among the best in the league last year, and we’ll see how messing with a good thing turns out.

Of course, since this is a Magic blog, I will look at Orlando’s decision to let Hedo Turkoglu walk and trade for Vince Carter.  I will be using a number of advanced statistics that, thankfully, I didn’t have to calculate myself.  There is a wealth of basketball statistics available on the Internet these days, and everything I will discuss today is publicly available.  The numbers I will be using were gathered from BasketballValue.com, my own Composite Score statistics, Basketball-Reference.com, 82games.com, and a new site called Hoopdata.com.

Overall Player Rating Statistics

carterturkogluoverall

Let’s start with a cursory glance at overall player ratings for Carter and Turkoglu.  With these numbers, Turk fares better in adjusted plus-minus and Composite Score, while Carter has the upper hand in PER and Win Shares.  The difference in Composite Score is the most dramatic, and that is mainly due to differences in their Defensive Composite Scores (which I will get into later).  There is also a pretty substantial difference in PER, and I think that is a reflection of Carter’s overall production per minute being more high volume than Turkoglu’s production.  The fact that Carter had more Win Shares than Turkoglu despite playing on a pretty bad team is quite impressive.

Offense Stats

carterturkogluoffense

According to the numbers, this is Carter’s biggest advantage.  In just amount any offensive metric you use, Carter looks better.  He is more efficient and produces more total offense than Turk.  Similarly, he had a greater impact on his team’s offense in terms of plus-minus.  Offensive Composite Score reflects all of these things.

Defense Stats

carterturkogludefense

Here, Turkoglu strikes back.  Carter looks below average in just about every category, and this supports his reputation.  Turk, on the other hand, recorded numbers well above average in every category.  The trickiest part about these comparisons is team context.  It is something I’ve mentioned constantly when talking about my Composite Score numbers.  Because of the way stats are tracked (at least publicly), it’s very difficult to separate a player’s individual contribution to his defense.  How much of this is Hedo’s own doing, and how much of it is due to the fact that Orlando featured a very strong all-around defense?  It’s hard to say, but I do think Turkoglu was probably a better defender than Carter.

Playmaking Statscartertukogluplaymaking

One of Turkoglu’s biggest benefits to the Magic, and something I thought they may miss, was his ability to create looks for others.  This was magnified in the playoffs when the Magic dominated the Cavaliers behind the creativity of Turkoglu.  Of course, Carter is no slouch in this area either, and the numbers above reflect this.  His Assist Rate was actually higher than Turkoglu’s, and he was able to take better care of the ball in the process.  Despite this, 82games.com gave Turk a better “Passing Rating,” although a worse “Hands Rating.”  Regardless of the tiny differences on each side, I think it’s safe to say that Turkoglu’s playmaking abilities are no better than Carter’s.

Assist Stats

carterturkogluassist

These numbers, which are available at Hoopdata, show what types of shots that the two players assisted on.  They are pretty similar across the board.  I think it’s interesting that Carter assisted on slightly more shots that were converted at the rim than Turkoglu did, despite the latter playing with one of the best (if not the best) finishers in the game in Dwight Howard.  It’ll be interesting to see how these numbers look after this season.

Clutch Stats

carterturkogluclutch

Partly because of his success in the playoffs, Hedo Turkoglu developed the reputation of being a clutch scorer and player.  Carter has been a go-to guy late in the game for much of his career, so how do the two compare?  Last year, Carter was actually more productive and more efficient shooting-wise than Turkoglu.  Both were great from the free throw line and reasonably good playmakers, but the difference in effective field goal percentage was pretty dramatic.  Carter’s was above average, while Turkoglu’s was well below average.  Most players find it more difficult to hit their shots in crunch time when defenses tighten up, so the fact that Carter actually became more efficient with the game on the line is quite impressive.

Conclusion

Ignoring all of the other players involved (although we definitely should not understate them), did the Magic make the right move by switching from Turkoglu to Carter?  VC is better offensively, and two of Turk’s most famous skills, playmaking and clutch play, are performed as well or better by Carter.  The only concern is defense, especially since the Magic lost Courtney Lee.  However, we don’t know for sure how great of a defender Turkoglu is when he isn’t playing in front of Dwight Howard, so that aspect remains to be seen.  All in all, considering Carter’s potential to put them over the top, the other players they acquired, and the amount of money Hedo was demanding, it appears to have been the right move for Orlando.

Each Player’s Impact on His Teammates’ Three-Point Shooting
Oct 21st, 2009 by Jon Nichols

In my last two pieces of research, I took a look at the impact of superstars on their teammates’ three-point shooting.  Specifically, I looked at how often and how efficiently teammates shot three-pointers when a particular superstar is on the court versus when he is off of it.  The theory has always been that superstars create open looks for their teammates, and my research for the most part confirmed that.

Unfortunately, I only took a look at ten players.  What about the rest of the league?  How good do you have to be to have a positive impact on three-point shooting?  What kind of impacts do role players have?  Do weak offensive players make three-point shooting harder for their teammates because defenders don’t pay them attention?

Today I will provide the data that should answer many of those questions.  My methodology for generating this data was the same as before, only on a much larger scale.  For each player, I calculated the three-point attempt percentage (three-point attempts divided by total field goal attempts) and three-point percentage for each of his teammates when the given player was on the court and off it.  Each of those teammates was then weighted based on their total number of attempts.  Finally, I calculated a weighted average for each player.  A player’s own three-point shooting was excluded from teammate effects.

In case you didn’t care about any of the stuff in the last paragraph, all you need to know is the following: the table below contains two numbers for each player who saw action in the NBA last year.  The first number is the average increase (or decrease) of all teammates’ three-point attempt percentage when the given player is in the game.  The second number is the average increase (or decrease) of all teammates’ three-point shooting percentage when the given player is in the game.  In both cases, a positive number is generally good.

Before you take a look at the numbers, beware of small sample sizes.  Some players may appear to have very positive or very negative impacts on their teammates, but it is most likely due to that player receiving very little playing time or some other confounding factors.

The numbers, courtesy of Google Docs:

http://spreadsheets.google.com/ccc?key=0AvNKNGJ_AHijdFJSNVJ1R3dzNTJsbEU3aDBCejhsY0E&hl=en

Below are some leaders and other interesting categories:

Top Impacts on Three-Point Attempt Percentage (Outliers Removed)

  1. LeBron James, 11.42%
  2. Chris Paul, 8.53%
  3. Kobe Bryant, 7.2%
  4. Rajon Rondo, 7.15%
  5. Hedo Turkoglu, 6.89%
  6. Joe Johnson, 6.65%
  7. Jason Terry, 6.63%
  8. Dwight Howard, 6%
  9. Dwyane Wade, 5.36%
  10. Aaron Gray, 5.3%

Nothing too shocking as there are a lot of great playmakers at the top of this list (except for Gray).  We’ll see if the Magic miss Turkoglu’s playmaking ability, although he benefitted by often playing with Dwight Howard and Rashard Lewis.

Top Impacts on Three-Point Shooting Percentage (Outliers Removed)

  1. Desmond Mason, 12.27%
  2. Will Solomon, 12%
  3. Jamaal Magloire, 11.85%
  4. Kyle Lowry (MEM), 8.92%
  5. Mike Taylor, 8.61%
  6. Lou Williams, 8.4%
  7. Caron Butler, 8.07%
  8. Rashard Lewis, 7.7%
  9. Aaron Gray, 7.54%
  10. Reggie Evans, 6.86%

Unless these ten seemingly random players have some special abilities we’re not all aware of, I think this list is strong evidence that perhaps a player’s “impact” on his teammates’ three-point shooting efficiency can often be a matter of luck or hidden factors such as common teammates.  There are some offensive stars with high impacts in this area (such as Shaquille O’Neal and Chris Paul), but this measure appears to be rather inconsistent.

Top Big Man Impacts on Three-Point Attempt Percentage (Outliers Removed)

  1. Dwight Howard, 6%
  2. Aaron Gray, 5.3%
  3. David West, 4.64%
  4. Brandon Bass, 4.51%
  5. Boris Diaw, 3.83%
  6. Marreese Speights, 3.44%
  7. Chris Bosh, 3.36%
  8. LaMarcus Aldridge, 3.35%
  9. Elton Brand, 3.07%
  10. Kevin Garnett, 2.89%

Some randomness, although I think Howard at the top is no fluke.  Orlando’s offense often runs through him and he’s more than willing to kick out the ball for an open three.

Top Point Guard Impacts on Three-Point Attempt Percentage (Outliers Removed)

  1. Chris Paul, 8.53%
  2. Rajon Rondo, 7.15%
  3. Roko Ukic, 5.09%
  4. Tony Parker, 4.8%
  5. Deron Williams, 4.45%
  6. Kyle Lowry (HOU), 4.43%
  7. Will Solomon, 4.42%
  8. Goran Dragic, 4.14%
  9. T.J. Ford, 3.77%
  10. Derrick Rose, 3.77%

Top Swingman Impacts on Three-Point Attempt Percentage (Outliers Removed)

  1. LeBron James, 11.42%
  2. Kobe Bryant, 7.2%
  3. Hedo Turkoglu, 6.89%
  4. Joe Johnson, 6.65%
  5. Jason Terry, 6.63%
  6. Dwyane Wade, 5.36%
  7. Courtney Lee, 2.98%
  8. O.J. Mayo, 2.85%
  9. Linas Kleiza, 2.81%
  10. Manu Ginobili, 2.75%

I plan on doing more studies like this in the future, as well as working more with the data I presented here today.  There may be some randomness to the numbers but there also may be some relevance, so I’d like to dig deeper.

TPS Reports: Workhorses, Defenders/Outside Shooters, and Interior Defenders
Oct 8th, 2009 by Jon Nichols

Today I’m going to unveil some more player ratings using the TPS system I developed recently, which produces customizable rating systems that are adjusted for position.  To see last time’s ratings, click here.

The first rating today is called the Workhorse Rating.  It consists of three components: PER, Usage Rate, and Assisted Rate (the percentage of a player’s shots that were assisted).  It is a player rating system that focuses especially on a player’s ability to generate his own offense without the help of others.  This isn’t always a good thing (see: ball hog), but it’s often useful to have a player who can create his own shot.  PER is a great all-around player rating system that factors in Usage Rate.  The reasons for including the other two components are pretty obvious.  The top 10 players in this category are:

workhorsetop10

And the bottom 10 are:

workhorsebottom10

The next rating is called Defender and Outside Shooter Rating (not very creative but very self-explanatory).  It consists of three components: Defensive Composite Score, Three-Point Attempt Percentage, and Three-Point True Shot Percentage.  The leaders in this category are (remember they’re adjusted for position):

defenderoutsideshootertop10

There are some names you’d expect (Bowen, Battier, Artest, and even Wallace) with some others you wouldn’t expect (Ilgauskas, Lewis, Nelson).  The bottom 10 in this category are:

defenderoutsideshooterbottom10

The third and final rating I will discuss today is my Interior Defender Rating.  This one is also self-explanatory.  It consists of Defensive Composite Score and Rebound Rate.  You can be a good rebounder without being a good defender (cough, Carlos Boozer, cough), but someone who is good in both usually is a good interior defender.  Notice that I didn’t include blocks.  It is not one of the 25 components in the TPS system, and I’ve done studies that question the importance of them.  Keeping the position adjustments in mind, here are the top 10:

interiordefendertop10

I don’t know about Chris Paul in there, but I suppose he is a good rebounder for his position.  The bottom 10 in the league are:

interiordefenderbottom10

To see the numbers for every player, go to:

http://spreadsheets.google.com/ccc?key=0AvNKNGJ_AHijdEs4cW0wdkQ5U1o3dC1wRDc5d0NUYnc&hl=en

Stay tuned to Basketball-Statistics.com for more TPS Reports.

TPS Reports: Paint Presence, Perimeter Threats, and Super-Efficiency
Oct 3rd, 2009 by Jon Nichols

Last time I introduced my TPS (Talent Plus Style) system and a basic statistic using that system called Playmaker Score.  Today I’ll show a couple more statistics I’ve developed using TPS.

The first rating is something I called Paint Presence Rating.  Before I go any further, remember that all ratings are adjusted for position, so some point guards rate higher than centers, even though they are obviously not greater “paint presences.”  Paint Presence Rating is determined by a player’s Composite Score, Close Attempt Percentage, Close True Shot Percentage, free throws/field goal attempts, and Rebound Rating.  Composite Score reflects a player’s all-around ability while the other four are all affected by a player’s skills/tendencies around the basket.

The second rating is Perimeter Threat Rating.  This rating reflects a player’s tendency to shoot three-pointers and his ability to make them.  It consists of three components: Composite Score, Three-Point Attempt Percentage, and Three-Point True Shot Percentage.

The third and final rating I will release today is called Super-Efficient Rating (it sounds silly, I know).  It incorporates three elements: Offensive Rating, True Shot Percentage, and Turnover Rate.  Offensive Rating is a good catch-all for every efficient thing a player can do offensively, while the latter two measure a player’s ability to hit shots when he needs to and not turn the ball over.  Things such as assists and offensive rebounds are also part of offensive efficiency, but they are only partly incorporated, through the use of Offensive Rating.

To see all the numbers, go to: Link.

A New Method for Evaluating Players, Part II: Explanation of TPS
Sep 30th, 2009 by Jon Nichols

As I alluded to in my last post, something I’ve wanted to do for a long time was to make a player rating system that is totally customizable.  Every team has different needs at different times, so the one-number-fits-all style of player rating systems, which has been the norm, seems inappropriate in certain situations.  Today I would like to explain to you something I’ve created called the TPS (Talent Plus Style) player rating system (movie reference: when I produce TPS reports in the future, just know that I did see the memo and I won’t forget to include the new cover sheet).

Why TPS?  As I said in my last article:

Let’s say Shaquille O’Neal rates better than Rashard Lewis in Player Rating System X.  The Magic should try to swap Lewis for Shaq, then, right?  Obviously not.  Orlando needs a big man (calling Lewis “big” is a stretch, but go along with it for now) that can stretch the floor and give space for Dwight Howard down low.  Suddenly, we’re doing so much contextual research for Player Rating System X that the player rating itself isn’t that useful anymore.  Instead, we’re relying on shooting percentages, shooting tendencies, rebounding ability, defensive ability, etc.

Having the flexibility to specify what you are looking for in particular from a player makes things easier for talent evaluators and more fun for fans.  There are a couple of things about TPS that make it very useful:

  • Every variation of TPS ratings is adjusted for position.  A center who can shoot three-pointers at the league average is much more valuable than a shooting guard with the same efficiency.  Likewise, a point guard with a league average rebounding rate is probably pretty good in that respect.  Not only does adjusting for position level the playing field, but it also makes it easier to find players who are unconventionally good at things.
  • TPS can include my original Composite Score rating system if desired.  As you may or may not recall, Composite Score is a combination of the Offensive/Defensive Ratings developed by Dean Oliver, PER /counterpart PER, and offensive/defensive plus-minus.  Additionally, if you only like some of those components, you can pick and choose which ones to include.  Composite Score often serves as a base for many of the ratings I develop when I’m playing around with TPS.  It provides a nice “all-around” measure for players so that even when you want to focus on specific skills, you don’t ignore everything else that occurs on a basketball court.
  • There are literally an infinite number of possibilities for how you can rate players.  Up to 25 different variables can be included, and each can be assigned a different weight, depending on what’s important.
  • Every variable is adjusted either per-minute or per-shot-attempt.  Players who perform admirably despite limited playing time will still rate well.

Which variables can be included in TPS player ratings?  They can be broken down into three categories:

Composite Score Stats:

  • Composite Score
  • Offensive Composite Score
  • Defensive Composite Score
  • Offensive Rating
  • PER
  • Offensive Plus-Minus
  • Defensive Rating
  • Counterpart PER
  • Defensive Plus-Minus

Shooting Stats:

  • Close Attempt Percentage (=close attempts/total shot attempts)
  • Close True Shot Percentage
  • Midrange and Post Attempt Percentage (=midrange and post attempts/total shot attempts)
  • Midrange and Post True Shot Percentage
  • Three-Point Attempt Percentage (=three-point attempts/total shot attempts)
  • Three-Point True Shot Percentage
  • Fouled Attempt Percentage (=fouled shot attempts/total shot attempts)
  • Fouled True Shot Percentage
  • Assisted Rate (=assisted field goals/total made field goals)

Other Advanced Stats:

  • True Shot Percentage
  • Free Throws/Field Goal Attempts
  • Pure Point Rating
  • Assist Rate
  • Turnover Rate
  • Rebound Rate
  • Usage Rate

As you can see, I did not develop all of these statistics (for some I just adjusted them for position and use them as components in my ratings).  Offensive and Defensive Ratings were developed by Dean Oliver and can be found at Basketball-Reference.com.  PER and many of the other advanced stats were developed by John Hollinger.  I used Basketball-Reference and Knickerblogger.net to gather these.  For player names and teams, I used Dougstats.com.  For positions, plus-minus, and Counterpart PER, I used 82games.com.  Additionally, although I calculated the shooting stats myself, similar numbers can be found at 82games.  To calculate the shooting stats, I used the play-by-play data available at BasketballValue.com.

The ratings I have come up with so far are all based on a 0-100 scale, with 50 being average.  Remember that these ratings are adjusted for position, so a player with a 50 is average for his position.

To give an example of what TPS can do, I have created a rating called Playmaker Score.  This number rates players on all of their abilities, but especially their ability to create shots for others.  It considers a player’s Composite Score, Assisted Rate, Pure Point Rating, and Assist Rate (remember that Assisted Rate measures how many of a player’s shots were assisted by others while Assist Rate estimates how many of teammates’ field goals an individual player assists on).  To see the numbers, go to:

Link

This is just one rating that I threw together pretty quickly.  Still, it gives some interesting results and shows what TPS is capable of doing.  Over the next few weeks, I will come out with some similar player ratings using TPS.  If you have any questions or suggestions, feel free to comment below.

P.S. If you’re wondering what movie I referenced earlier, check out: Link to video

A New Method for Evaluating Players, Part I: Rating Shooting Abilities
Sep 25th, 2009 by Jon Nichols

One concern that many people have is that player rating systems are often too general.  I’ll be the first to tell you my Composite Score rating system needs a bunch of contextual information to truly be useful.  It’s simply too hard to sum up all of a player’s abilities with a single number.  One major problem is all the things that go unmeasured, although that’s outside the scope of our abilities until we start tracking new things.

A second major problem, one that I’m trying to find a solution for, is that different teams have different needs for different situations.  Let’s say Shaquille O’Neal rates better than Rashard Lewis in Player Rating System X.  The Magic should try to swap Lewis for Shaq, then, right?  Obviously not.  Orlando needs a big man (calling Lewis “big” is a stretch, but go along with it for now) that can stretch the floor and give space for Dwight Howard down low.  Suddenly, we’re doing so much contextual research for Player Rating System X that the player rating itself isn’t that useful anymore.  Instead, we’re relying on shooting percentages, shooting tendencies, rebounding ability, defensive ability, etc.

It still would be nice to have one number when we’re trying to evaluate players, if for no other reason than to save time.  But we’ve already proven that one number is useless without context.  What can we do?

Create multiple sets of player ratings.  Better yet, create an organic player rating system that adjusts based on whatever is important to us at the moment.  The Magic need a power forward that can shoot three-pointers efficiently and create his own shot from time to time?  Ok, let’s rate power forwards based on that.

The next step is to calculate all of those little components and adjust them by position.  Why adjust for position?  If we made a player rating system based on three-point shooting ability and shot-creation alone, without adjusting for position, our numbers would tell us the Magic should acquire someone like Roger Mason and put him at power forward.  That doesn’t seem like a wise suggestion.

Once we have all the position-adjusted components, we can then decide which are important based on our needs.  Today is the first step.  Similar to how I broke down individual players by quarter, each player in the league will be rated based on how he performs from four shooting locations: close (dunks and layups), midrange (including post shots), three-pointers, and getting to the free throw line.  Each rating is adjusted for position, so a center with a 90 rating on three-pointers is still very likely worse overall than a shooting guard with an 80.

The ratings will combine both frequency and efficiency.  In other words, if a player rarely shoots from midrange but is efficient at it, he won’t rate that well.  Similarly, if he shoots from midrange all the time but is highly inefficient, he also won’t rate well.  Ratings are on a scale of 1 to 100, with 50 being average for that position.

Frequency is measured by the player’s attempts from that shot location divided by his total attempts.  Efficiency is measured by his makes divided by his total attempts from that location.  The only situation that is included in this efficiency measure is when a shot actually goes up, so things like turnovers are ignored.

Before I release the numbers, I should say that these shooting tendencies and efficiencies are nothing new.  82games.com has had this data available for a while now.  My methods for extracting these tendencies and efficiencies from the play-by-play data are slightly different, but they are similar.  The new step I am taking is adjusting these numbers by position and creating a rating system off of these adjustments.  The numbers are available through Google Docs below:

http://spreadsheets.google.com/ccc?key=0AvNKNGJ_AHijdE5RWnZVcG9vS1VaQ1B5VFdBZG5tMHc&hl=en

If you’re angry because a certain player does not rate the way you’d expect, allow me to explain.  First, remember these ratings account for efficiency.  Superstars may be excellent shot producers (a skill I will rate in the near future), but they are not always the most efficient.  Second, these ratings also account for a player’s tendencies.  If a player is extremely likely to take a certain shot, his rating will be high for that.  However, if he balances his shot attempts, he will not rate extremely high in any of them.

A simple way to look at it is that these ratings are attempting to describe players as much as they are attempting to evaluate them.  LeBron James may only get an 80 in close shots (which is still quite high), but that’s because he mixes up his attempts.  He clearly is one of the most frightening players in the world when he’s near the basket.

These ratings do evaluate to an extent, but the bulk of evaluation for my new rating system will come from other components.  Shooting ratings will be a big part of the context I mentioned at the beginning of this article.

This is just a first run, so changes will inevitably be made.  If you have any suggestions, feel free to comment below.

Banks, Hooks, and Fades Article Fixed
Aug 6th, 2009 by Jon Nichols

Just a quick note: some of you may have had trouble viewing the tables on the Banks, Hooks, and Fades article.  The article is fixed now.  My apologies.

-Jon Nichols

»  Substance: WordPress   »  Style: Ahren Ahimsa
© Copyright 2009 Basketball-Statistics.com