Pitcher’s Luck as a Rate Stat

I introduced Luck last week and expanded the work to change hitter’s Luck to a rate stat. Today I am going to look at pitcher’s Luck as a rate stat. Pitcher Luck is mainly linked to a pitcher’s ERA, but it can also be related to their WHIP (extra hits) and Wins (less runs allowed).

When I first looked at pitcher Luck, I added up the entire amount of luck a pitcher had over a season. This method created too much emphasis on playing time. If two pitchers had the same LOB%, HR/FB% and BABIP, they should have the same luck no matter how many innings they pitched. I took the old Luck value and divided it by the total batted balls in play (TBF – K – BB – HBP). After doing the division, I adjusted the values to a -10 to +10 scale (link to spreadsheet with 2010 and 2011 data).

Initially, LOB% and BABIP got the same weightings and HR/FB% was weighted twice as much as the other two. I am keeping the HR/FB% weighting, but only giving a weighting of 1/2 for both of BABIP and LOB%. The reason for this change is that LOB% is directly related to having a higher BABIP. Luck on balls in play will be taken into account, but not as much as before.

With these new adjustments, here is a look at leaders and laggards for both the new and old Luck values (> 100 IP this season):

Name Old Luck New Luck LOB% BABIP HR/FB ERA FIP xFIP
Weaver Jered 9.4 8.9 81.3% 0.254 5.7% 2.44 3.00 3.67
Cain Matt 8.3 7.8 71.3% 0.260 3.9% 2.84 2.87 3.69
Cueto Johnny 7.6 8.5 75.7% 0.248 6.0% 2.36 3.45 3.86
Hellickson Jeremy 7.4 7.7 81.7% 0.229 8.1% 2.96 4.30 4.57
Halladay Roy 7.3 7.2 77.2% 0.307 5.1% 2.44 2.11 2.61
Arroyo Bronson -6.0 -4.2 72.6% 0.288 15.5% 5.28 5.65 4.52
Carmona Fausto -6.1 -4.4 62.0% 0.290 12.9% 5.18 4.52 4.12
Greinke Zack -6.3 -8.1 67.4% 0.316 13.7% 3.93 2.96 2.51
Burnett A.J. -7.7 -6.3 68.4% 0.289 15.8% 5.27 4.82 4.02
Volstad Chris -9.7 -10.0 66.5% 0.317 16.3% 5.37 4.50 3.70
Name New Luck Old Luck LOB% BABIP HR/FB ERA FIP xFIP
Worley Vance 9.1 6.5 77.0% 0.278 6.5% 2.92 3.18 3.60
Moscoso Guillermo 8.9 6.5 68.1% 0.230 5.4% 3.34 4.17 5.17
Weaver Jered 8.9 9.4 81.3% 0.254 5.7% 2.44 3.00 3.67
Cueto Johnny 8.5 7.6 75.7% 0.248 6.0% 2.36 3.45 3.86
Beckett Josh 8.2 7.3 82.4% 0.238 8.1% 2.49 3.37 3.59
Blackburn Nick -5.0 -6.0 70.0% 0.317 14.1% 4.49 4.84 4.30
Burnett A.J. -6.3 -7.7 68.4% 0.289 15.8% 5.27 4.82 4.02
Arrieta Jake -6.5 -4.9 72.7% 0.272 15.0% 5.05 5.34 4.50
Greinke Zack -8.1 -6.3 67.4% 0.316 13.7% 3.93 2.96 2.51
Volstad Chris -10.0 -9.7 66.5% 0.317 16.3% 5.37 4.50 3.70

Luck can be seen as the differences between the pitcher’s ERA and their FIP and xFIP. With each lucky pitcher, their ERA is less than their FIP and xFIP. The unlucky pitchers are the other way with their ERA being more than both their FIP and xFIP.

Additionally, I went back and looked at how Luck stood up last season. I took all the pitchers with the bottom and top 10 Luck values from 2010 and saw what their ERA did in 2011 (min 40 IP in 2011). Here is how 2010’s top 10 luckiest pitchers have fared so far in 2011:

Name Luck 2011 ERA 2010 ERA Difference
Johnson Josh 9.8 1.64 2.30 0.66
Buchholz Clay 8.8 3.48 2.33 -1.15
Zambrano Carlos 8.4 4.82 3.33 -1.49
Anderson Brett 8.3 4.00 2.80 -1.20
Jimenez Ubaldo 7.6 4.61 2.88 -1.73
Kershaw Clayton 7.0 2.36 2.91 0.55
Wood Travis 6.9 4.92 3.51 -1.41
Duensing Brian 6.9 5.34 2.62 -2.72
Price David 6.8 3.40 2.72 -0.68
Wilson C.J. 6.5 3.01 3.35 0.34
3.76 2.88 -0.88

Seven of the pitcher’s ERAs have increased with an average overall increase of 0.88.

Here are the ten unluckiest pitchers in 2010 and how their ERA changed in 2011:

Name Luck 2011 ERA 2010 ERA Difference
Harang Aaron -3.2 3.74 4.53 -0.79
Nolasco Ricky -3.5 4.40 5.32 -0.92
Leake Mike -4.1 4.00 4.27 -0.27
Blackburn Nick -4.4 4.49 4.23 0.26
Shields James -4.4 2.70 5.42 -2.72
Karstens Jeff -4.8 3.32 5.18 -1.86
Francis Jeff -4.8 4.88 4.92 -0.04
Duke Zach -5.8 5.00 5.13 -0.13
Correia Kevin -7.0 4.79 4.84 -0.05
Beckett Josh -9.7 2.49 4.22 -1.73
Average 3.98 4.80 -0.82

This group on average saw their ERA drop by 0.82

The effects of luck can also be seen with Wins and WHIP. Here is a comparison of the chances of a win per start and WHIP for 2010’s luckiest pitchers and how they performed in 2011:

Stat 2010 2011
% chance of a Win per Start 50.0% 40.8%
Wins per 32 Starts 16.0 13.1
WHIP 1.20 1.26
BB/9 3.3 2.9

Per 32 starts, the pitchers averaged 3 less wins in 2011. Also, their WHIP was 0.06 higher even though their BB/9 went from 3.3 to 2.9.

For now, I like the results of pitcher’s Luck. It easily finds pitchers that may have their stats inflated by being unlucky on balls in play, home runs and sequence of hits (LOB%). It is a little late in the season to track pitchers for this season, but I will have of a final 2011 list and will be using it during the off season to help find under valued players. Again, let me know what you think.





Jeff, one of the authors of the fantasy baseball guide,The Process, writes for RotoGraphs, The Hardball Times, Rotowire, Baseball America, and BaseballHQ. He has been nominated for two SABR Analytics Research Award for Contemporary Analysis and won it in 2013 in tandem with Bill Petti. He has won four FSWA Awards including on for his Mining the News series. He's won Tout Wars three times, LABR twice, and got his first NFBC Main Event win in 2021. Follow him on Twitter @jeffwzimmerman.

3 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
evan
12 years ago

are the 2011 ERA’s adjusted for run environment?