You are on page 1of 6

# Putts Minus Greens: A Simple Aggregate Golf Statistic

## Mark Hachey 3.27.2012

Motivation: Total putts and total greens in regulation do a poor job individually describing a golf round. Any model that tries to predict a score from either statistic doesnt fare very well. For instance, greens in regulation can never go below 0 so it cant really capture the badness of a player who also might be inclined to have a lot of putts. Total putts is much too simplistic and any prediction based on that number fails if we dont consider it in conjunction with GIRs. Can we find a better statistic that captures the combination of putting and ball striking? Putts Minus Greens I created an aggregate statistic that tries to capture ball striking and putting into a single number. I call this statistic "Putts Minus Greens" or "PMG". It's simply total putts minus total greens in regulation. Method I took the statistics from a total of 41 professional and amateur rounds (the Sky Caddie website shows putting and GIRs for players who choose to track them). For total putts, total greens, and PMG, I ran a simple linear regression against score and plotted the fits. With a more easily readable database, I should be able to provide better estimates for all parameters. Discussion Figure 1 shows total putts versus score. The regression coefficient was significant (Pval = .0003, R2 = .2864). Figure 2 shows greens in regulation versus score. The regression coefficient was significant (P-val < .0001, R2 = .7412). Finally, figure 3 shows PMG versus score. Again, regression coefficient was significant with considerably more of the variation in the data explained by the model compared to putts or greens (P-val < .0001, R2 = .8688). PMG gives us a measure of putting AND ball striking together. It doesnt add information to what we already know, but it provides an aggregate measure, like OPS in baseball (on base percentage plus slugging percentage). PMG succinctly captures a snapshot of a whole round. Table 1 shows the predictive abilities of Putts, Greens, and PMG for the data I entered based on my models. More data would make for more accurate models. I included Tiger Woods 4 rounds, and Jeff Overtons 4 rounds for comparison. The models tend to not be as accurate with Jeff Overtons rounds since he had several double bogeys or worse. A nice proprerty of PMG is that a par round should come out to around an 18. For instance, if a player hits 18 greens, and 2 putts them all, he would have a PMG of 18 (36 putts minus 18 greens). Similarly, a player who missed every green but scrambles

successfully on all of them would have a PMG of 18 (18 putts minus 0 Greens). Also, a PMG of less than 18 should be an under par round while a PMG of greater than 18 should be over par.

110 100 90

score

## N 41 Rsq 0.2864 AdjRsq 0.2681 RMSE 10.443

80 70 60 26 28 30 32 34 36 38 40 42

putts
Figure 3 Total putts versus score.

110 100 90

score

## N 41 Rsq 0.7412 AdjRsq 0.7345 RMSE 6.2892

80 70 60 0 2 4 6 8 10 12 14 16 18

GIR
Figure 3 Greens in regulation versus score.

110 100 90

score

## N 41 Rsq 0.8688 AdjRsq 0.8655 RMSE 4.4769

80 70 60 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0 32.5 35.0 37.5 40.0

PuttMinusGir
Figure 3 PMG vs score

## Put ts Overton -1 2 3 4 Tiger - 1 2 3 4 MSE 31 25 30 32 32 30 29 32

Gree ns 12 10 15 10 14 17 11 15

Putting Predictio n 81.25 69.38 79.27 83.23 83.23 79.27 77.29 83.23 94.37

Greens Predicti on 74.92 79.26 68.42 79.26 70.58 64.08 77.09 68.42 19.08

PMG Predicti on 73.76 66.62 66.62 79.12 71.98 63.05 71.98 70.19 7.45

Score 76 70 69 84 69 65 71 70

Table 1. Jeff Overtons rounds, and Tigers rounds from Arnold Palmer Invitational.