clock menu more-arrow no yes mobile

Filed under:

Improving the win

Instead of just complaining about how useless the pitching win is, can it be improved? Cubs broadcaster Len Kasper has a way.

If he didn't wear a Cubs uniform, you would all know who this is
If he didn't wear a Cubs uniform, you would all know who this is
Matt Kartozian-US PRESSWIRE

I listen to 670 The Score in Chicago for my sports information, and last Thursday (May 8th), Cubs TV play-by-play announcer Len Kasper was on the McNeil and Spiegel Show. Host Matt Spiegel and guest host Laurence Holmes were discussing the fate of Jeff Samardzija so far this year, who to date has eight starts, seven quality starts, a 1.45 ERA . . . and zero wins. Len went on to discuss one way to revamp the win-loss record:

It took me a couple tweets to get it right, but in essence Len is suggesting recalibrating wins and losses to better reflect pitcher performance. Len went on to write about this in a column in Sunday's Daily Herald.

The link in my first tweet goes to a page at Baseball-Reference.com that tabulates and defines cheap wins and tough losses for starting pitchers:

Cheap wins--Wins in starts with < 6 IP or more than 3 ER. Or wins in non-quality starts.

Tough losses--losses in quality starts

Part of my discussion with Len involved Cubs pitcher Travis Wood, who has the following numbers so far in his career:

29 wins, 39 losses, 62 quality starts (out of 101), 2 cheap win, 14 tough losses and 34 no-decisions

Last year was typical for Wood -- 32 games started, 24 quality starts, 3.11 ERA and a 9-12 record to show for it. Using Len's transformation, Wood's career numbers would change from 28-39 to 27-25. Reviewed using traditional measures, Wood appears average at best, but changing the view toward one based on what Wood can control depicts a more effective pitcher, one with a winning percent far exceeding his team's -- .519 vs. the Cubs .386.

This data is available going back to 1938, and this table shows pitchers most helped by this change since 2000, with explanation to follow (minimum 100 games started):

Pitcher From To GS W L W% ND Wc Lt Wa La Wa% W%d
Josh Johnson 2005 2013 160 57 43 .570 60 9 20 48 23 .676 .106
Jordan Zimmermann 2009 2014 120 45 36 .556 39 7 16 38 20 .655 .100
Travis Wood 2010 2014 100 29 39 .426 34 2 14 27 25 .519 .093
Clayton Kershaw 2008 2014 185 79 46 .632 60 7 19 72 27 .727 .095
Jon Niese 2008 2014 125 45 42 .517 38 8 18 37 24 .607 .089
C.J. Wilson 2005 2014 147 65 39 .625 43 10 17 55 22 .714 .089
Ben Sheets 2001 2012 250 94 96 .495 60 15 39 79 57 .581 .086
Brandon Webb 2003 2009 198 87 62 .584 49 8 23 79 39 .669 .086
David Price 2008 2014 155 74 42 .638 39 10 17 64 25 .719 .081
Cole Hamels 2006 2014 248 99 76 .566 73 15 30 84 46 .646 .080
Anibal Sanchez 2006 2014 178 62 61 .504 55 8 22 54 39 .581 .077
Adam Wainwright 2007 2014 193 103 58 .640 32 15 23 88 35 .715 .076
Jake Peavy 2002 2014 312 132 99 .571 81 21 38 111 61 .645 .074
Roy Oswalt 2001 2013 341 159 99 .616 83 19 36 140 63 .690 .073
James Shields 2006 2014 259 103 85 .548 71 13 30 90 55 .621 .073
R.A. Dickey 2003 2014 181 70 66 .515 45 6 21 64 45 .587 .072
Jered Weaver 2006 2014 239 117 62 .654 60 17 24 100 38 .725 .071
Madison Bumgarner 2009 2014 123 53 42 .558 28 9 16 44 26 .629 .071
Matt Cain 2005 2014 271 93 91 .505 87 17 35 76 56 .576 .070
Felix Hernandez 2005 2014 277 113 87 .565 77 14 30 99 57 .635 .070

ND=no decision

Wc=cheap win

Lt=tough loss

Wa=adjusted wins, subtracting the cheap wins from the actual, leaving wins that were also a quality start. La likewise removed the tough losses, with Wa% the adjusted winning percent for the pitcher. The table is sorted by win percent difference (W%d) of the two methods, showing Josh Johnson as the pitcher most penalized by playing for teams that gave him no run support, followed closely by Jordan Zimmermann, Clayton Kershaw and . . . Travis Wood.

Some time ago I saw a tweet from MLB Historian John Thorn regarding the win in early baseball history -- my memory was hazy, and he was kind enough to set me straight:

Almost one hundred years ago the man who was the first commissioner of the American League saw the win as inferior to other more informative measures. He was foresighted enough to see this even when pitchers were completing a fair number of games started -- look how that percentage has changed over time:

Complete_game_2

It's been almost one hundred years since pitchers completed even half of their games and almost forty since completing twenty-five percent -- Old Hoss Radbourn left the building a long, long time ago.

Baseball front offices left the win behind as a meaningful measuring stick of pitcher effectiveness years ago and understand the failure to score runs shouldn't be blamed on the pitcher. This message is gradually spreading, but it's still not uncommon to hear people complain how the win is still relevant. I'm sure Jeff Samardzija is bothered to not have any wins, but he knows it's not due to any true shortcomings on his part, and his agent (Mark Rodgers) is certainly aware of how Cubs losses in Samardzija's starts stem from the fact the Cubs have given him the worst run support among starting pitchers in all of baseball.

It's a tough battle, but with broadcasters like Len Kasper and his partner Jim Deshaies who don't just call the game but attempt to explain and enhance the viewers knowledge, it's one that will be won. I admire what Len did in not just complaining about an archaic stat but making it better and more informative as well.

As I was writing this post, I received a tweet from one of my favorite people to follows on Twitter, HighHeatStats, who suggested I look at starter effectiveness in a different way. I gathered the data, and to say my mind has been opened is an understatement. I'll write about it next Monday.

Data from Baseball-Reference.com. Any mistakes in amalgamating it are the author's. Special thanks to Len Kasper for taking the time to discuss this with me.

Scott Lindholm lives in Davenport, IA and has been waiting 'til next year for the better part of his life. Follow him on Twitter @ScottLindholm.