Rodgers Report: Bang for Buck

There are numerous available metrics for measuring the worth of a sports athlete. Bang for Buck (BFB) is a simple metric derived from two objective measures. Bang is the total of points, assists, and rebounds the player has posted. Buck is the player's salary in millions. The idea behind the stat is that a team is paying for a player's output so one way to measure the payoff on the investment is to look at some measure of output divided by money spent. I opt for total points, assists, and boards instead of per game totals because a player who is elite when he plays but who misses a good deal of time may not be as good an investment as a reliably above average player.

Through 25 games, the Lakers' best investments are as follows:
Matt Barnes 208.5
Derrick Caracter 178.7
Shannon Brown 155.9
Devin Ebanks 108.5
Lamar Odom 84.3
Derek Fisher 81.4
Steve Blake 55.8
Ron Artest 48.1
Pau Gasol 47.5
Kobe Bryant 35.5
Theo Ratliff 13.3
Luke Walton 6.3
Sasha Vujacic 5.3
Andrew Bynum 0.0

A quick look at the offseason moves by LA reveals a good deal of shrewdness on GM Mitch Kupchak's part. Nabbing Matt Barnes for a pittance (1.77m) looks pretty brilliant right now. Resigning Shannon Brown and Derek Fisher look like good moves, too. Meanwhile, the rookies (Caracter and Ebanks) are already paying off reasonably well. On the other hand, oft-injured Luke Walton and perennial pineman Sasha Vujacic look like horrible investments.

Limitations: BFB is a season-long metric, which will take into account playoff performance as well. Obviously, some players do not make the playoffs so their opportunities are limited to 82 games provided they are healthy. At the other extreme, Kobe Bryant played in 23 playoff games last season, posting more than 900 combined points, assists, and rebounds. Those games count at least equally to regular season games. I'll be working on some thoughts for the playoffs.

interesting.. if you kinda average out the fact that superstars get paid ENORMOUS amounts and that rookies get paid minimal...


you could say that LO is probably our best investment...

and those that say to get rid of RON.. should look at this.. for what we are paying him.. hes doing decent.. although i too am infuriated by his offense these days

Bynum sux.

Killer B's represent!

000000 wrote:Bynum sux.

this

i do like your total approach as opposed to averages or per minute stats, but couldn't you get a better picture by using adjusted +/-, ewa, roland, etc. and dividing by salary in millions? it would at least attempt to take into account impact not measured by traditional stats.

i'd also consider some kind of transformation on the salary data, as i don't believe they're very normally distributed. in the end, these sorts of analyses will undersell the top players, especially older ones.

needless to say, bang for buck be damned, you wouldn't want to field an entire roster of players who rank highly in this stat.

OKC looks pretty impressive on this list. I think the Clipps do, too.

This is very cool. It would be interesting to see what Kobe would need to contribute to justify his salary. If you feel comfortable posting your Excel sheet that would be cool to see. I guess you could goal seek it to see what Kobe needs to average to have the best "Bang for the Buck."

Good work!

^ Good question.

Last Season
Shannon 479
Farmar 435
Powell 333
Odom 260
Artest 241
DJ 208
Fisher 195
Gasol 130
Bynum 127
Kobe 118
Vujacic 61
Walton 30
Ammo 23

Postseason
Farmar 84
Brown 81
Fisher 71
Artest 68
Odom 62
Gasol 48
Kobe 41
Bynum 29
Powell 17
DJ 11
Vujacic 11
Walton 8
Ammo 3

I guess the question is, "What justifies a certain salary?" Based on regular season performance, it looks like a rating of about 200. Of course, there are some issues here. First, some salaries are simply set by the league. It's unfair to call rookie contract players "bargains". Second, none of this take into account winning.

abeer3 wrote:i do like your total approach as opposed to averages or per minute stats, but couldn't you get a better picture by using adjusted +/-, ewa, roland, etc. and dividing by salary in millions? it would at least attempt to take into account impact not measured by traditional stats.

Good question. 82games.com does a "Salary" stat similar to that. I was looking for quick and dirty :)

Maybe a fix would be this: each team's P+A+R is divided by their wins and then multiplied by 41 (a .500 season) or maybe whatever a .500 season is including all games in the playoffs. That way, good teams get more bang for their buck...as they should. I'll try that.

Okay, if I do what I just said (LA had 73 wins), Kobe's value is 274. At the same time, it increases Shannon's to 964. The only players not worth their salary (200) on LA were:

Vujacic 123
Ammo 45
Walton 66

That sure looks reasonable.