Evaluating pitchers using Game Score

Marlins fans, four words for ya -- wait 'til next year. It should be worth it. - Denis Poroy

Bill James developed the Game Score to measure pitcher effectiveness. Can it be aggregated to evaluate pitching careers?

My last post discussed using Win Probability Added (WPA) as a method to measure pitcher effectiveness, arguing it was a quasi-counting stat that in the aggregate could be used to evaluate a pitcher's career. The creator of the stat, Tom Tango, was kind enough to point out the shortcomings in this approach, and I strongly suggest reading the three links he referenced. I'm a firm believer in allowing creators to define their creations, so disregard my Monday post.

In particular, Tom's second post alluded to what had been my initial thought, which was to use Game Score and see how that worked as a counting stat. With this jumping-off point, I continue my odyssey in search of a better way to judge a pitcher's career other than wins.

Game Score was developed by Bill James, and the beauty is in its simplicity. This is the definition:

  1. Start with 50 points.
  2. Add one point for each out recorded, so three points for every complete inning pitched.
  3. Add two points for each inning completed after the fourth.
  4. Add one point for each strikeout.
  5. Subtract two points for each hit allowed.
  6. Subtract four points for each earned run allowed.
  7. Subtract two points for each unearned run allowed.
  8. Subtract one point for each walk.

There is no advanced math involved, no arcane statistics, and includes the bread and butter of what defines effective pitching -- outs, strikeouts, walks, and hits. This table shows the top 10 pitchers since 1914 in aggregate Game Score:

Pitcher From To GS GSc GScAvg BF
Nolan Ryan 1966 1993 773 46262 59.8 22310
Don Sutton 1966 1988 756 42887 56.7 21500
Roger Clemens 1984 2007 707 41731 59.0 20227
Greg Maddux 1986 2008 740 41511 56.1 20388
Steve Carlton 1965 1988 709 40790 57.5 21457
Gaylord Perry 1962 1983 690 39680 57.5 21140
Phil Niekro 1965 1987 716 39319 54.9 21585
Tom Seaver 1967 1986 647 39016 60.3 19337
Bert Blyleven 1970 1992 685 38944 56.9 20429
Warren Spahn 1942 1965 665 37845 56.9 20960

Using Nolan Ryan as an example, he made 773 starts (GS) and had an aggregate Game Score (GSc) of 46,262, or an average of 59.8 per start, facing a total of 22,310 batters (BF). Game Score, in this instance, is a counting stat, and pitchers like Ryan, Don Sutton and Gaylord Perry benefit from their respective long careers. In one sense, this recognizes pitchers like this with long careers and "inferior" records, acknowledging solid pitching on bad teams.

What happens when viewing the chart in terms of Game Score average to see who was the most effective during his career? These are the top 10 among pitchers with at least 400 starts since 1914:

Pitcher From To GS GSc GScAvg BF
Pedro Martinez 1992 2009 410 25093 61.2 10970
Bob Gibson 1959 1975 482 29398 61.0 15669
Tom Seaver 1967 1986 647 39016 60.3 19337
Randy Johnson 1988 2009 603 36122 59.9 16937
Nolan Ryan 1966 1993 773 46262 59.8 22310
Juan Marichal 1960 1975 457 27044 59.2 14118
Roger Clemens 1984 2007 707 41731 59.0 20227
Jim Palmer 1965 1984 521 30480 58.5 15714
Curt Schilling 1988 2007 436 25489 58.5 12510
Don Drysdale 1956 1969 465 27137 58.4 13700

These are arguably the best pitchers in recent memory with an interesting mix of long career vs. short-and-good. It clearly recognizes more recent pitchers due to the rise of strikeouts, but older pitchers begin to show up not much further down the list. The entire list of pitchers with at least 100 starts since 1914 can be viewed here -- feel free to play around with it and reach your own conclusions.

So far this reaches back to the dusty recesses of baseball history -- what about today? Can Game Score identify those pitchers whose careers are off to hot starts? This chart shows active pitchers with at least 30 career starts (data through Monday, May 19th):

Pitcher From To GS W L  
ND 
GSc GScAvg BF
Jose Fernandez 2013 2014 36 16 8 12 2281 63.4 886
Clayton Kershaw 2008 2014 186 79 47 60 11318 60.8 4822
Yu Darvish 2012 2014 69 32 20 17 4176 60.5 1878
Chris Sale 2012 2014 63 31 22 10 3803 60.4 1736
Hisashi Iwakuma 2012 2014 52 24 10 18 3099 59.6 1338
Stephen Strasburg 2010 2014 85 32 22 31 4974 58.5 2008
Adam Wainwright 2007 2014 194 103 58 33 11176 57.6 5341
Felix Hernandez 2005 2014 279 115 87 77 16053 57.5 7789
Cole Hamels 2006 2014 249 100 76 73 14267 57.3 6629
David Price 2008 2014 157 75 43 39 8978 57.2 4222

This includes a pitcher's record in games started -- any decisions in relief are not included. There's a wide range of pitchers, from the young-and-injured like Jose Fernandez, recent imports like Yu Darvish and Hisashi Iwakuma, and those who are safely in mid-career and poised to be among the best of their generation in Adam Wainwright and Felix Hernandez. I strongly suggest going to the referenced worksheet and seeing who just missed the top 10.

Slowly but surely, the win is losing its effectiveness as a manner of determining pitcher value for the simple reason that wins are decreasing. The current career leaders in wins are Tim Hudson (209) and CC Sabathia (208), and a review of past pitchers with similar totals generates a resounding "meh." This is a topic I'll return to when discussing Pedro Martinez and what should be his obvious Hall of Fame credentials despite "only" winning 219 games, but it's time to find other methods to evaluate a pitcher's career.

The Game Score is very high on the list, and any number of methods could be construed to make it easier to understand and integrate -- award wins and losses on whether a pitcher crossed a certain threshold, tweak the constants to better reflect current benchmarks, gradually interject it into the conversation as opposed to the quality start, etc. None of the pitchers' inclusion in any of the tables should elicit stunned disbelief, since they are widely considered to be among the best of their respective eras. Sometimes dominant performance is overshadowed by a pedestrian win-loss record.

These are the best and worst Game Score games since 2000:

Player Date Tm Opp Rslt App,Dec IP H R ER BB SO HR Pit GSc
Matt Cain 6/13/2012 SFG HOU W 10-0 SHO9 ,W 9.0 0 0 0 0 14 0 125 101
A.J. Burnett 5/2/2012 PIT STL L 3-12 GS-3 ,L 2.67 12 12 12 1 2 2 72 -13

For the record, the best Game Score of all time was 105 in Kerry Wood's 20-strikeout game on May 6th, 1998 (in his fifth career start). There is some danger in using a stat designed as a game descriptor and broadening it to measure a career, but there's also value in comparing this number with his contemporaries. Using aggregate Game Score can help introduce another method with which to evaluate a pitcher's career.

As mentioned earlier, Tom Tango introduced four different Game Score methods, and my next post will focus not on evaluating them but seeing how pitcher careers fare depending on the method used. The referenced spreadsheets give a hint of how pitchers performed, so feel free to get a head start and see what the different methods show.

All data from Baseball-Reference.com. Any errors in amalgamating or augmenting the data are the author's.

Scott Lindholm lives in Davenport, IA. Follow him on Twitter @ScottLindholm.

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Beyond the Box Score

You must be a member of Beyond the Box Score to participate.

We have our own Community Guidelines at Beyond the Box Score. You should read them.

Join Beyond the Box Score

You must be a member of Beyond the Box Score to participate.

We have our own Community Guidelines at Beyond the Box Score. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9351_tracker