Saturday, August 27, 2005

Lean On Me: A Primitive Study of Sacramento's Players' Impacts on Teammates' Shooting

More analysis on the player pair stats made available by 82games.com was promised, so here we go.

We looked at True Shooting percentage (abbreviated as TS%) plus-minus figures for players on the Kings. True shooting is a measure of shooting efficiency that accounts for free throws, as well as field goals. You can learn more about why it's used at Sonics.com writer Kevin Pelton's stats primer. He makes a much better agrument than I can. But in short, the formula is:
TS% = Points/(2 x (Field Goals Attempted + (Field Throws Attempted x .44)))


So, we ripped the stats from 82games (thanks, Roland!) and calculated TS% for each player when a certain player is also in the game. True shooting wasn't one of things Roland included in the release, necessitating that calculation.

Then, we compared each player's TS% with a certain teammate with that player's overall TS%.

For instance, for Mike Bibby, we calculated his TS% with Brad Miller, Chris Webber, Peja Stojakovic, Doug Christie, Cuttino Mobley, Darius Songaila, Kenny Thomas, Brian Skinner, Erik Daniels, Eddie House, Bobby Jackson -- every player he played a minute with last season.

Then we subtracted Bibby's overall TS% (which happens to be 54.3%) from those pair TS% figures (for example, Bibby's TS% when Miller was also in the game was 57.4%).

In our example, Bibby has a TS% 3.1% better with Miller in the game. But Mobley, for instance, had a negative effect on Bibby's TS% -- dropping it to 53.9%, or a net -0.4% effect.

This tells us a lot on an individual basis -- Doug Christie may not score as much or as well as Mobley, but his being on the court makes Bibby's TS% much better than does Mobley being on the floor.

But it can tell us something about players overall, too -- not just tandems. Looking at the data, some players seem to have a positive effect on their teammates TS% more than others. Can we put a number on this effect?

Yes! We resorted our data by the 'player2' -- the teammate that correlates with specific rises and falls of TS%. Instead of simply averaging the net +/- impact the teammates had on TS%, we weighted the values by multiplying the net +/- by the minutes the tandem played together.

For example, when Christie was in the game, Maurice Evans's TS% an astronomical 27.1%. Sadly, the pair played only 64 minutes together, negating the seemingly huge impact. Christie did play a significant number of minutes along Peja, though -- 864 minutes. Peja's TS% went up 2.2% in that time.

Multiplying the minutes and the net +/- magnifies the actual impact to the player's TS% with a certain teammate on the floor. We added all those up for each 'teammate' and divided by the total minutes as a 'teammate' (or four times the teammate's actual floor minutes -- each player is four players' teammate every minute on the court).

This gave us a weighted TS% impact for each player. Or put another way, this is the net impact a player had on his teammates' TS%. Here's the list, sorted from postive to negative (including the three new additions based on their own team's stats):

Surprising in places, isn't it? I expected Miller at the top -- he's a big guy that opens up the floor for everyone. I have to admit Webber was a minor shock -- the assumption is that he clogged up and choked the offense, making everyone else stand around and not look for good shots. But if not for Miller, CWebb would've been the best in this category.

Cuttino having a positive impact on his teammates' TS% was also surprising. Evans was lower than I would've guessed, as was Skinner. But injuries and the trades really provided a lot of turbulence in the rotation, and roles were pretty jumbled all season.

It's not encouraging that none of SAR, Bonzi and The Hitman (new nickname alert! new nickname alert!) positively affected their teammates' TS%. Shareef's stats showed a positive impact on other bigs (like Zach Randolph and Theo Ratliff) and a negative on guards (especially Mighty Mouse, for some reason). He did have a positive effect on Sebastian Telfair's TS%, though.

Bonzi's stats looked awful. Gasol, Watson and Miller saw their TS% increased a decent amount with Bonzi on the floor, but the TS% for every other major rotation player (including White Chocolate, Battier, Lorenzen Wright, Stro Swift and James Posey) saw significant decreases with Bonzi alongside them.

The Hitman helped Primoz Brezec and a host of role players. He hurt Gerald Wallace, Brevin Knight, Steve Smith and Jason Kapono substantially and Emeka Okafor and Keith Bogans a bit.

And remember, this only measures shifts in a player's teammates' TS% when said player is in the game. It does not account for the other three players on the floor (which is the biggest challenge to effectively measuring teammate impacts), the defenders on the other side of the ball and garbage time.

But I think it starts to tell us a little bit about what players have a positive impact on their teammates, at least on the offensive end. I haven't touched the rebound, assist, points or steals data, though, because I'm nervous about the whole "three other players on the court" thing.

Anyone have ideas how to further dig into this data? Comment or email me.

1 Comments:

Blogger pookeyguru said...

ok first off...the only 3 surprises was that webber was that high...that barnes was that high & that thomas was that low...not surprised by anybody else...for instance mobley did make extra passes to open shots for peja alot of times...that will increase anybody's % in that scenario..bibby doesnt do that..which will naturally decrease it...first off i wasnt surprised by evans..he really didnt have a chance to make anybody 'better'..same with skinner & williamson..wasnt their roles..i guess considering the way its calculated i assume webber does get some credit for what he did do @ times..but that still doesnt take into account that he did alot of negative things on both ends...oh well...its an interesting chart & interesting numbers...but like anything numbers mislead you without a real context in which they're set under to give a real guideline to the impact they make. & like many things...these numbers may point to webber being higher than u may have thot...but he also had the ball in his hands alot so that may account for such a high %..the real question i have is which is the more valuable %...webber's or stojakovics....in my humble opinions its pejas..which is why he's still a king..& webber is not..

2:53 AM  

Post a Comment

<< Home


Editor:
Tom Ziller

The Sacramento Kings are a tough act to follow, literally.

SKB does the dirty work so you can forget about the blood, sweat and tears.

Oh, you'll still need the tears in April, though.

Powered by Blogger

Weblog Commenting and Trackback by HaloScan.com