Wednesday , Jan , 21 , 2009 C.Y. Ellis

The Ultimate Stat Equivalator II: The Best of Times, the Wurst of Times


The Ultimate Stat Equivalator II: The Best of Times, the Wurst of Times

German statesman Otto von Bismarck once said, “The less people know about how laws and sausages are made, the better they sleep at night,” but he never had to create a fantasy basketball player-rater algorithm.*  Mine is relatively uncomplicated, and explaining it is still taking two columns.  Eat it, Bismarck!  That said, please don’t invade my apartment.  It would not be a valuable strategic addition to Greater Prussia.  It’s a railroad, for Christ’s sake.

In my last column, we introduced my state-of-the-art fantasy rating system, USE(LESS): the Ultimate Stat Equivalator (Linking Equivalated Stats in Summation). Its premise was simple: find out how valuable each type of stat is relative to each other, and adjust stats accordingly.  For example, if the average NBA game has 200 points and 20 blocks, then a block is 10 times as valuable as a point in fantasy terms.  However, I was forced to be coy about USE(LESS) field-goal and free-throw percentages; those clearly can’t work the same way.  Today, we unveil the gruesome inner workings of the machine.

[WARNING: technical, possibly terrifying statistics ahead.  If you want to skip to the USE(LESS) Top 25 Player Rankings further down the page, I will not be offended.  In fact, I probably won’t even notice.  I am definitely not watching you through the window right now as you read this.  There is no need to look up.  Yes.  Good.]

For both percentages, what I sought to do was come up with a single index for each player that took into account 1) their percentage relative to the rest of the league and 2) their attempts per game.  For their percentage, I decided to use everyone’s favorite statisical measure, the z-score:

z-score = (player’s_percentage – league_avg_percentage) / league_standard_deviation

For field-goal percentage, for example, the league average is .452, and the standard deviation is .107.  Therefore, a player with a percentage of .559—one standard deviation above the mean—would have a z-score of +1.

After that, I multiplied the z-score by the number of attempts, and that’s when I thought I was done.  However, I noticed that this gave me insignificant quantities relative to all other stats: Dwight Howard’s horrifying free-throw percentage, in particular, affected his final score very little.  My fix was crude, yet elegant: I took the average of the highest scores across the league for each of the seven non-percentage statistics, divided by the highest percentage scores, and multiplied the index by that figure.  Consequently, the adjusted FG% and FT% numbers have, on average, as much effect on a player’s USE(LESS) index as any other stat.  To quote Emeril:  BAM.**

[END OF STATISTICS.]

For the last ten games, here’s the USE(LESS) Top 25 Player Rankings:

*rank Player Team PS *ppg *rpg *apg *spg *bpg *3pg *topg *fg% *ft% *total
1 granger,danny ind SF 30.20 12.53 19.05 16.00 31.37 44.62 -12.77 -6.45 27.03 161.58
2 iguodala,andre phi SF 24.00 12.77 31.90 28.00 0.00 26.15 -14.89 45.11 2.60 155.63
3 james,lebron cle SF 28.10 18.68 36.67 25.33 33.33 23.08 -30.50 21.17 -1.65 154.21
4 bryant,kobe lal SG 30.20 13.48 32.38 16.00 5.88 30.77 -25.53 28.95 16.95 149.08
5 redd,michael mil SG 26.90 7.09 14.29 20.00 1.96 46.15 -11.35 35.10 2.83 142.98
6 okur,mehmet uta C 21.11 22.06 8.47 11.85 8.71 23.94 -11.82 36.75 19.55 140.62
7 pierce,paul bos SF 23.50 12.29 15.24 14.67 9.80 40.00 -21.28 26.63 17.47 138.32
8 nowitzki,dirk dal PF 26.40 19.15 12.86 14.67 17.65 13.85 -14.89 25.20 21.24 136.11
9 o’neal,shaquille pho C 21.63 23.35 8.33 6.67 36.76 0.00 -15.96 59.17 -5.61 134.33
10 paul,chris nor PG 22.70 13.95 50.00 28.00 0.00 10.77 -18.44 17.16 8.81 132.95
11 wade,dwyane mia SG 29.40 10.17 40.00 33.33 33.33 15.38 -25.53 -0.68 -2.88 132.53
12 lee,david nyk PF 17.80 30.97 11.90 14.67 9.80 0.00 -16.31 56.87 4.83 130.53
13 durant,kevin okl SF 25.90 19.15 16.67 17.33 9.80 24.62 -29.08 26.81 17.15 128.35
14 foye,randy min PG 19.30 7.57 19.05 20.00 13.73 38.46 -14.18 9.18 10.89 123.99
15 bogut,andrew mil C 11.50 19.50 11.90 13.33 34.31 0.00 -15.96 54.69 -7.92 121.36
16 biedrins,andris gsw C 12.30 30.73 11.90 14.67 43.14 0.00 -7.80 33.74 -17.82 120.86
17 howard,dwight orl C 21.10 35.70 6.67 17.33 37.25 0.00 -23.40 54.31 -28.46 120.50
18 bargnani,andrea tor PF 21.00 15.60 7.62 5.33 13.73 41.54 -19.15 26.73 6.62 119.02
19 murphy,troy ind PF 14.00 26.71 11.43 17.33 7.84 41.54 -8.51 1.97 4.68 116.99
20 nelson,jameer orl PG 18.10 7.57 27.62 14.67 3.92 33.85 -11.35 9.33 10.38 114.08
21 lewis,rashard orl PF 17.60 11.82 16.67 13.33 9.80 47.69 -15.60 2.85 8.52 112.68
22 miller,andre phi PG 18.80 11.58 30.00 18.67 1.96 10.77 -19.86 41.33 -1.63 111.61
23 duncan,tim san C 20.00 21.75 21.43 4.00 43.14 0.00 -15.60 23.70 -8.70 109.71
24 kirilenko,andrei uta SF 14.50 15.84 17.62 25.33 31.37 4.62 -7.09 5.10 0.81 108.10
25 ginobili,manu san SG 16.40 9.93 17.62 33.33 7.84 21.54 -12.06 2.07 11.23 107.90

Can we all agree that Danny Granger is an elite fantasy player at this point?  If he is not a first-round pick next year in your league, you should smack all of your friends in the face.  The same is not true of Andre Iguodala, but certainly his current streak needs to be ridden until someone reminds him that he is, when all is said and done, Andre Iguodala.  LeBron and Kobe, at #3 and #4, ought to be especially valuable tonight, and just thinking about it is clogging my arteries with fatty lumps of joy.  Other sneaky value guys turned up by USE(LESS): Randy Foye, David Lee, Troy Murphy, and the incomparable Andrea Bargnani, available in almost 40% of ESPN’s fantasy leagues as I write this.  Please tap that before it comes crashing back to earth.  Please.  Please tap that. 

Other interesting ratings: my personal favorite Brook Lopez at #29, Luol Deng at #36 (in his four games since returning from ankle problems—buy low while you still can), and at #34, Houston’s delicious and nutritiousVon Wafer.  With McGrady out, he’s lighting it up from the two-spot, and in 96% of ESPN leagues, he’s up for grabs.  Ninety-six!  You look hungry.  Have a Wafer.

*Interestingly, he tried to do so in 1866, and ended up with this:

points=invasions_of_austria*sqrt(bratwurst)

The only active player with a non-zero score would be Dikembe Mutombo, who invaded Austria briefly in 1988, and also really likes bratwurst.

**Or: “Pork fat rules!”