Freaky Pitching Leaderboards
We keep track of a number of unusual baseball stats at The Hardball Times, based on the data sent to us by Baseball Info Solutions, including batted ball stats (ground ball, fly ball, line drive) for each batter and pitcher. We think these stats provide a little extra insight into what’s happening in the baseball world.
You can find many of these stats in our stats section. We’re working on a sortable version (similar to our new Win Shares format) so you can construct your own leaderboards. In the meantime, however, I’ve constructed some special THT-only leaderboards, including some new stats. Pitchers today, batters Thursday.
These leaderboards include only pitchers who have faced at least 149 batters—a total of 152 pitchers or about five per team. Stats are based on games through last Friday, which marked the end of the first third of the season. There are four Leaderboards in all. (If you like, you can start with more conventional leaderboards on the Major League Baseball site).
Expected DER
Lately I’ve been reading a lot of articles claiming that you can judge a team’s fielding prowess by its Defense Efficiency Ratio (DER). I’m not so sure about that.
DER tracks the number of batted balls that are caught for outs by fielders, not including home runs, so it seems as though DER should be a useful fielding stat. Pitchers, however, can have a lot of impact on whether a batted ball is caught or not. In fact, I’ve conducted research in the past that indicates that fielders are responsible for at most 50% of variance in DER between teams. How can this be?
There are two issues:
- The batted ball type yielded by a pitcher can have a big impact on DER. Line drives are converted into outs only about 25% of the time. For ground balls, the percentage is 72%; that figure is 75% for outfield flies and 97% for infield flies.
- Beyond that, batted balls fall into different zones on the field, some of which are within a fielder’s normal range and others that aren’t. Without zone data, it is impossible to use statistics to judge whether or not a batted ball might have been fairly caught by a fielder.
I’ve constructed a stat that attempts to address the first of these two issues. It’s called Expected DER.
To create Expected DER, I looked at all 2004 pitchers with at least 50 innings pitched, and ran a simple regression analysis of batted ball types against each one’s DER. The best formula I found was
Expected DER equals .738 + .068*OF% – .349*LD% + .23*IF%
where OF% equals the percent of batted balls that were outfield flies (not including home runs), LD% the percent that wereline drives, and IF% the percent that were infield flies. This formula has an R squared of .195, which means that about 20% of an individual pitcher’s DER can be explained by the type of batted ball he allows.
As I said, this is a simple analysis; I didn’t attempt to correct for team fielding or ballpark. I will improve this formula in the future, but I think that even this simple form is useful. So here’s a leaderboard of the pitchers whose actual DERs most exceeds their expected DERs, based on the above formula. These pitchers either have some fine fielders behind them or are getting batters to hit into good zones.
Player Team ERA IP DER xDER Diff Ohka T. WAS 3.20 50.67 .805 0.695 0.110 Clemens R. HOU 1.30 76.00 .784 0.696 0.087 Myers B. PHI 2.06 74.33 .765 0.693 0.072 Martinez P. NYN 2.62 79.00 .789 0.722 0.067 Santos V. MIL 2.73 62.67 .751 0.685 0.067 Prior M. CHN 2.93 58.33 .774 0.710 0.064 Rogers K. TEX 1.65 76.33 .747 0.685 0.062 Hampton M. ATL 1.83 59.00 .766 0.709 0.057 Contreras J. CHA 3.27 66.00 .789 0.734 0.055 Obermueller MIL 3.00 36.00 .739 0.687 0.052 Ishii K. NYN 4.79 35.67 .790 0.741 0.049 Zambrano C. CHN 3.22 72.67 .770 0.722 0.048 Redman M. PIT 3.14 71.67 .750 0.705 0.045 Estes S. ARI 4.16 67.00 .722 0.677 0.044 Heilman A. NYN 3.99 49.67 .721 0.682 0.038
Many of the best pitchers so far this year are on this list, which of course makes absolute sense. If you achieve a DER of .750 or higher, you are going to have a very good year.
The opposite is also true. Here is the Expected DER un-leaderboard (or laggardboard, I guess):
Player Team ERA IP DER xDER Diff Ortiz R. CIN 5.23 43.00 .634 0.717 -0.082 Brown K. NYA 5.14 56.00 .635 0.704 -0.069 Wilson P. CIN 7.77 46.33 .644 0.709 -0.065 Schmidt J. SF 5.08 51.33 .680 0.740 -0.060 Glover G. MIL 6.19 48.00 .660 0.712 -0.052 May D. SD 4.96 32.67 .667 0.718 -0.052 Perez O. PIT 6.92 53.33 .703 0.752 -0.049 Kennedy J. COL 6.51 56.67 .655 0.703 -0.048 Young C. TEX 3.03 65.33 .704 0.744 -0.039 Hendrickson TB 5.37 55.33 .682 0.720 -0.038 Washburn J. LAA 3.80 68.67 .674 0.711 -0.037 Escobar K. LAA 2.97 36.33 .697 0.733 -0.036 Waechter D. TB 5.43 53.00 .700 0.730 -0.030 Lilly T. TOR 7.60 45.00 .653 0.682 -0.029 Vazquez J. ARI 3.65 74.00 .677 0.706 -0.029
It’s interesting that there are some pitchers having very good years on this list, such as Javier Vazquez, Chris Young, Kelvim Escobar and Jerrod Washburn. I don’t know what that means, but just think how good they’d be if their actual DERs were as good as expected.
By the way, Oliver Perez’s .752 Expected DER is by far the best in the majors (Chris Young is second at .744) due to an extreme fly ball ratio and low line drive rate.
Dominance
Because infield flies are caught 97% of the time, they are just about as certain an out as a strikeout. So I’ve created a stat that I call Dominance, which measures how often a pitcher induces either a strikeout or infield fly as a percent of batters faced. Here are the Dominance Leaders:
Player Team IP ERA K IFFly Dom Martinez P. NYN 79.0 2.62 92 11 0.354 Santana J. MIN 83.3 3.67 105 8 0.348 Burnett A. FLA 68.0 2.91 64 21 0.309 Escobar K. LAA 36.3 2.97 43 3 0.305 Myers B. PHI 74.3 2.06 76 9 0.294 Prior M. CHN 58.3 2.93 62 7 0.292 Peavy J. SD 76.0 2.37 78 6 0.287 Vazquez J. ARI 74.0 3.65 67 19 0.282 Harang A. CIN 66.7 2.97 57 17 0.280 Clemens R. HOU 76.0 1.30 76 6 0.279 Fossum C. TB 40.7 3.54 40 9 0.275 Zambrano C. CHN 72.7 3.22 68 12 0.275 Sheets B. MIL 37.3 4.34 39 4 0.274 Perez O. PIT 53.3 6.92 51 15 0.266 Schmidt J. SF 51.3 5.08 49 14 0.265
As you can see, Pedro and Johan (yes, we’re on first name bases here) are the most dominant pitchers in baseball, measured in this manner. Thirty-five percent of the time, they have been 100% responsible for getting an out.
Here’s the Dominance un-leaderboard:
Player Team IP ERA K IFFly Dom Drese R. TEX 67.0 6.04 18 5 0.076 Saarloos K. OAK 55.0 4.75 14 5 0.078 Erickson S. LAN 42.7 6.75 9 7 0.083 Silva C. MIN 67.0 3.09 21 2 0.087 Rueter K. SF 65.3 4.27 14 13 0.095 Ohka T. WAS 50.7 3.20 16 5 0.100 Blanton J. OAK 48.7 6.66 16 8 0.108 Wang C. NYA 37.7 4.06 14 3 0.109 Robertson N. DET 59.7 3.17 27 3 0.114 Lima J. KC 55.3 8.13 25 6 0.118 Ramirez H. ATL 58.0 5.43 20 10 0.120 Glavine T. NYN 62.3 5.05 31 4 0.122 Rogers K. TEX 76.3 1.65 30 8 0.123 Hampton M. ATL 59.0 1.83 22 6 0.124 Sele A. SEA 63.0 4.43 29 5 0.125
Once again, there are some very good pitchers on this list, such as Carlos Silva, Kenny Rogers and Mike Hampton. The key to Silva’s ERA is his outstanding control. I’ll have more to say about Rogers and Hampton in a bit.
Home Run/Outfield Fly
In general, the number of home runs a pitcher gives up depends on the number of outfield fly balls he yields and the park in which he pitches. This year, 11% of outfield fly balls have been home runs overall, though there are a lot of differences among pitchers. Here is a leaderboard of the pitchers who have given up the fewest home runs, as a percent of outfield flies, adjusted for ballpark. For reference, I’ve also listed each pitcher’s Ground ball/Fly ball ratio.
Player Team IP ERA HR/F G/F Wang C. NYA 37.7 4.06 0% 2.67 Rusch G. CHN 55.0 1.96 2% 1.02 Patterson J. WAS 47.3 2.85 2% .79 Rogers K. TEX 76.3 1.65 2% 1.25 Saarloos K. OAK 55.0 4.75 2% 2.26 Brown K. NYA 56.0 5.14 4% 2.00 Lidle C. PHI 69.7 3.88 4% 1.54 Thomson J. ATL 50.0 3.42 4% 1.36 Willis D. FLA 78.0 1.85 4% 1.69 Robertson N. DET 59.7 3.17 4% 1.97 Young C. TEX 65.3 3.03 4% .87
It pays to be a fly ball pitcher who doesn’t allow home runs, because fly ball pitchers have high DERs. This has been the secret of the success so far this season for John Patterson, Chris Young and Glendon Rusch.
Here’s a list of the pitchers who have yielded the most home runs as a percent of outfield flies (adjusted for ballpark again):
Player Team IP ERA HR/F G/F Erickson S. LAN 42.7 6.75 25% 1.86 Westbrook J. CLE 69.7 5.30 24% 4.05 Burnett A. FLA 68.0 2.91 24% 2.41 Perez O. PIT 53.3 6.92 23% .57 Lowry N. SF 58.7 5.37 23% 1.05 Lima J. KC 55.3 8.13 23% .79 Maroth M. DET 64.3 5.04 21% 1.40 Suppan J. STL 63.3 4.41 21% 1.63 Belisle M. CIN 39.7 4.76 20% 1.73 Pineiro J. SEA 48.7 6.66 19% 1.57 Lowe D. LAN 78.0 3.58 19% 2.92
For groundball pitchers like Jake Westbrook, Derek Lowe and A.J. Burnett, the high percentage doesn’t hurt nearly as much because they don’t give up a lot of fly balls. For fly ball pitchers however, such as Oliver Perez and Jose Lima, these percentages are killers.
Expected FIP
So far, these leaderboards have been retrospective, uncovering the keys to each pitcher’s success or lack thereof. For our last leaderboard, let’s turn to a prospective stat called Expected Fielding Independent Pitching, which I introduced a few articles ago.
Expected FIP is just like regular Fielding Independent Pitching, except I’ve normalized each pitcher’s home run rate to the average of 11%. As Ron Shandler disclosed in his most recent Baseball Forecaster, most pitchers will regress to the mean of 11% over time.
In last month’s article, I reviewed a list of all pitchers whose ERA most exceeded their Expected FIP and stated that they could be expected to improve. Here’s the list, with their May 10 ERA and June 4 ERA, so you can see how many actually saw their ERAs drop in the last month:
Previous Current Player Team ERA ERA Brown K. NYA 8.25 5.14 Down Wright J. NYA 9.15 6.39 Down Lilly T. TOR 7.77 7.60 Down Anderson B. KC 6.91 6.75 Down Elarton S. CLE 7.20 5.53 Down Bell R. TB 8.28 ---- Backe B. HOU 6.81 4.67 Down Kennedy J. COL 7.56 6.51 Down Lohse K. MIN 6.65 4.25 Down Harper T. TB 6.27 7.36 Up Wood K. CHN 6.15 ---- Wilson P. CIN 7.25 7.77 Up Vazquez J. ARI 4.70 3.65 Down
The list worked pretty well; nine of the 11 pitchers who actually pitched in the last month did indeed improve. So here’s an updated list of pitchers most likely to improve in the second two-thirds of the season, based on June 4 stats:
Player Team IP ERA xFIP Diff Lilly T. TOR 45.0 7.60 4.91 2.69 Wilson P. CIN 46.3 7.77 5.11 2.66 Pineiro J. SEA 48.7 6.66 4.51 2.15 Lima J. KC 55.3 8.13 6.20 1.94 Padilla V. PHI 32.0 7.03 5.13 1.90 Glover G. MIL 48.0 6.19 4.34 1.85 Milton E. CIN 65.0 7.06 5.22 1.84 Wells D. BOS 52.3 5.85 4.05 1.80 Perez O. PIT 53.3 6.92 5.27 1.65 Kim B. COL 31.0 6.97 5.33 1.64 Cabrera D. BAL 58.3 5.40 3.92 1.48 Westbrook J. CLE 69.7 5.30 3.82 1.48 Ledezma W. DET 45.3 6.75 5.38 1.37 Dempster R. CHN 44.7 4.84 3.48 1.36 Weaver J. LAN 73.3 5.65 4.31 1.34
I mean, Ted Lilly and Paul Wilson have to turn their season around sometime, right?
Now for the other side of last month’s list: here are the pitchers whose ERAs I expected to increase the rest of the year, and a review of how they’ve done since May 10:
Previous Current Player Team ERA ERA Blanton J. OAK 2.67 6.66 Up Garland J. CHA 1.38 3.22 Up Chacon S. COL 3.27 3.83 Up Rogers K. TEX 2.11 1.65 Down Moehler B. FLA 2.19 2.59 Up Hampton M. ATL 2.47 1.83 Down Patterson J. WAS 1.60 2.85 Up Sabathia C. CLE 2.63 3.58 Up Contreras J. CHA 2.60 3.27 Up Seo J. NYN 2.00 ---- Santos V. MIL 2.88 2.73 Down Robertson N. DET 4.18 3.17 Down
Seven have indeed gotten worse, but four have actually improved. This list is a real surprise to me. Four months ago, Kenny Rogers had a 2.47 ERA, and I would have taken very heavy odds that his ERA would go up. But it’s actually gone down to a miniscule 1.65 ERA (and this doesn’t include his dominant game against Kansas City on Sunday, in which he lowered his ERA to 1.62). Mike Hampton’s ERA has actually dropped from 2.47 to 1.83. Crazy.
Rogers and Hampton lead the list of pitchers now most likely to see ERA increases the rest of this year. Sooner or later, it will happen. Here’s the current version of that list:
Player Team IP ERA xFIP Diff Rogers K. TEX 76.3 1.65 4.87 -3.22 Hampton M. ATL 59.0 1.83 4.64 -2.81 Ohka T. WAS 50.7 3.20 5.56 -2.36 Rusch G. CHN 55.0 1.96 4.20 -2.24 Santos V. MIL 62.7 2.73 4.73 -2.00 Chacon S. COL 51.7 3.83 5.83 -2.00 Robertson N. DET 59.7 3.17 5.08 -1.92 Obermueller MIL 36.0 3.00 4.91 -1.91 Clemens R. HOU 76.0 1.30 2.95 -1.65 Willis D. FLA 78.0 1.85 3.36 -1.52 Bedard E. BAL 61.0 2.07 3.46 -1.40 Moehler B. FLA 55.7 2.59 3.97 -1.39 Patterson J. WAS 47.3 2.85 4.14 -1.29 Marquis J. STL 69.0 3.39 4.67 -1.28 Chacin G. TOR 67.0 3.36 4.64 -1.28
In a way, these lists aren’t that impressive. If you take a bunch of pitchers with low ERAs and say they’re going to get worse, well, no kidding. So let’s construct one last leaderboard. Here’s a list of the pitchers whose ERA most exactly matches his expected FIP. In other words, these are the guys most likely to continue to pitch as well or as poorly as they have been pitching so far:
Player Team IP ERA xFIP Diff Martinez P. NYN 79.0 2.62 2.62 0.00 Backe B. HOU 71.3 4.67 4.67 -0.01 Lohse K. MIN 55.0 4.25 4.24 0.01 Lidle C. PHI 69.7 3.88 3.89 -0.02 Burnett A. FLA 68.0 2.91 2.88 0.03 Mussina M. NYA 73.0 4.32 4.36 -0.04 Webb B. ARI 76.0 3.20 3.15 0.05 Suppan J. STL 63.3 4.41 4.46 -0.05 May D. SD 32.7 4.96 4.89 0.06 Lackey J. LAA 67.3 4.01 4.08 -0.07 Pettitte A. HOU 70.0 3.47 3.55 -0.07 Mulder M. STL 74.7 3.62 3.69 -0.08 Glavine T. NYN 62.3 5.05 5.13 -0.08 Benson K. NYN 36.3 4.21 4.13 0.08 Park C. TEX 58.7 4.60 4.70 -0.10 Brazelton D. TB 42.0 6.43 6.33 0.10
Met fans will be happy to see Pedro at the top of this list, but sad to see Tom Glavine on it too. A few comeback players, such as Brandon Webb, Chan Ho Park and Mark Mulder are for real. And that’s good to see.
We’ll have the batter leaderboards on Thursday. Come on back then.