Friday, November 14, 2014

Clayton Kershaw was not valuable in 83% of games.

Clayton Kershaw was voted 2014 National Conference Most Valuable Player (MVP) by the writers, who seem to be trying to absolve themselves for stiffing Mike Trout in each of the two previous seasons when they voted Miguel Cabrera American Conference MVP.  Yesterday Trout finally received an MVP award.  But the writers were determined to display a newfound awareness of modern analytics, even if it flew in the face of common sense.  Maybe, especially because it made so little sense.  That would make the writers seem even more clever.  Argh!

1Clayton KershawLAD355.018.085%8.027635110314.
2Giancarlo StantonMIA298.08.071%6.514553989155371051394.288.395.555.950
3Andrew McCutchenPIT271.04.065%6.41465488917225831884.314.410.542.952
4Jonathan LucroyMIL167.00.040%6.7153585731761369466.301.373.465.837
5Anthony RendonWSN155.00.037%6.515361311117621831758.287.351.473.824
6Buster PoseySFG152.00.036%5.2147547721702289047.311.364.490.854
Notice the Wins Above Replacement (players) WAR for the top three.  But not reflected in that for a pitcher is the pitcher's batting.  Their OPS (On Base Plus Slugging averages):
Kershaw .442
Stanton .950
McCutchen .952

Nor are fielding or base running considered for a pitcher.

Case against Kershaw for MVP: 198 innings.  Monday, November 10, 2014

If a pitcher throws in 200 innings he appears in about 14% of his team's innings. A modern iron man throwing about 300 innings would be in about 20%. You may think that the pitcher is doing much more than players at other positions. If the pitcher is striking out most of the batters that he faces, I might agree. However, those other players are retiring the batters who hit the ball. Why should the pitcher get credit for retiring all those batters who do not strike out?

4,374 (1,458 innings * 3 outs) are retired in 162 nine inning games. Let's ignore minor anomalies like the bottom of the ninth when the home team leads and extra innings. If a pitcher strikes out 200 batters he is retiring 4.5%. 300 strike outs: 6.85%...

Kershaw was 21-3, 239 SO.  For those who want to kill the win consider Roy Face 1959: 18-1, 93 innings, all in relief...

Kershaw gets a break in that the other two finalists also missed time in 2014 after being hit by pitch:
- Giancarlo Stanton: 145 games, 638 plate appearances
- Andrew McCutchen: 146 games, 648 plate appearances
- Clayton Kershaw: 27 games, 198 innings pitched.

What the heck?  You can play WAR games all you want but it doesn't make sense.

MVP pitchers: more on the case against Kershaw.  Wednesday, November 12, 2014

Pitchers are part time players.  Relievers are part time pitchers.  Pitchers, especially relievers, should not be MVP...

Kershaw pitched in only 27 of his team's 162 games: 16.66%.  The Dodgers played 135 games without Kershaw.  So, how valuable could Kershaw have possibly been when he did nothing in 83% of the games?  NOTHING...

MVP pitchers through 2013: ...

most innings 336 Denny McLain 1968
fewest innings 251 Justin Verlander 2011 ...

The fewest innings by a starter (251) through 2013 is 27% more than the 198 Kershaw had in 2014.

I've also noticed that the "kill the win" movement stopped moving for this vote.  Kershaw won 21 and lost 3.  The fact that his Dodger team had a much lower winning percentage in non-Kershaw games is used as the primary revelation as to his value.  But isn't that true about any pitcher with a gaudy won-loss record?  The pitcher MVP in the Cy Young award era, i.e., since 1956:

Justin Verlander 2011 24-5
Roger Clemens 1986 24-4
Vida Blue 1971 24-8
Bob Gibson 1968 22-9
Denny McLain 1968 31-6
Sandy Koufax 1963 25-5
Don Newcombe 1956 27-7

(And don't forget about non award winner Roy Face 1959: 18-1, 93 innings, all in relief.)

Hey, what happened to Gibson?  Nine losses with that puny 1.12 ERA?  His Cardinals were shutout a few times when he pitched.  That could have happened to Kershaw, too, but it didn't.  If the Dodgers were shut out in some of Kershaw's games and his won-loss record was less gaudy, would Kershaw have been less valuable?  Or would that simply have influenced the writers?

Notice that Newcombe won as many games as Kershaw appeared in and McLain won four more.

But didn't Kershaw have a season for the ages?  Kershaw had very good numbers in averages, including ERA.  But don't averages tend be better the less you play?  Could Kershaw have sustained that had he not missed the entire month of April?  And would the writers have voted for Kershaw had he missed September instead of April?  Stanton missed September.  Suppose Stanton had missed April instead?  Would Stanton have been voted MVP.  In other words, if you reverse the months missed for Kershaw and Stanton, would that have changed the outcome?  If it's even a possibility, then the actual result is bogus on that basis alone.

Kershaw pitched and won the Dodgers first game of 2014:

Saturday, March 22, 2014, 4:00 am, Sydney Cricket Grounds, Australia
Attendance: 38,266, Time of Game: 2:49
Dodgers 3, Diamondbacks 1

Kershaw pitched 6.66 innings and was helped by three relief pitchers.  However, Kershaw did not pitch again until May 6 in Washington, DC: Kershaw won again: 7 innings, no runs, again helped by three relief pitchers.  In between the Dodgers went 17-15.

Paul Maholm appears to be a pitcher who started some games in place of Kershaw.

12442Mar 23LAD@ARIW,7-59-9990.100001000.001432100000100001000000.440.0110.279b 1-- 1 out a49b 1-- 2 out a4
22457Apr 5LADSFGL,2-7GS-5L(0-1)124.175521209.642180471039941028000191001001.25-0.330-3.101t start tie5t --- 1 out d4
324610Apr 9LADDETL,6-79-931.010000007.9431593121100000003000100.160.0050.429t --- 0 out d39t 3 out d3
424712Apr 12LAD@ARIW,8-58-821.011100018.1051792031000000004000000.130.003-0.558b --- 0 out a68b 3 out a5
524815Apr 16LAD@SFGL,1-2GS-636.051132004.9725925313713730057000220002011.140.1541.521b start tie6b 3 out tie
624920Apr 21LADPHIL,0-7GS-5L(0-2)45.085432105.6027107582238147003200024200001.74-0.243-2.881t start tie5t 3 out d5
725025Apr 26LADCOLW,6-3GS-7W(1-2)47.062213004.742584531361294005900123100200.890.1080.971t start tie7t 3 out a3
825131May 3LAD@MIAW,9-7GS-666.093323204.70271026020412106004300025200100.77-0.033-0.271b start tie6b 3 out a4
Maholm pitched OK is some games, not so well in others.  The Dodgers were 2-3 in Maholm's starts during Kershaw's absence. The point is that an actual pitcher replaced Kershaw, not a mythical replacement player.  Suppose Maholm had been hot for that month and the Dodgers had won all five of his starts?  Then the Dodger record would have been 20-12 instead of 17-15.  Would that have made Kershaw less valuable.

Finally, let's cruise back to the 1972 MVP vote:

1Johnny BenchCIN263.011.078%8.614753887145401256100.270.379.541.920
2Billy WilliamsCHC211.05.063%6.11505749519137122362.333.398.6061.005
3Willie StargellPIT201.02.060%3.91384957414533112165.293.373.558.930
4Joe MorganCIN197.05.059%9.3149552122161167358115.292.417.435.851
5Steve CarltonPHI124.01.037%12.5411176231805.

Steve Carlton got all 24 first place votes for the Cy Young award but only one for the MVP despite having much higher WAR, a concept that was decades away from creation.  Carlton was all the rage because his Phillies had won only 59 games and he had won 27 of them.  But even that is skewed.  Look at the other Phillies starting pitchers:

1SPSteve Carlton*272710.7301.97414103080346.125784761787831012813511822.010.9936.
2SPKen Reynolds*25215.1184.2633231200154.114976731760687101652853.891.3548.
3SPBill Champion24414.2225.0930221200132.215580751154854107584713.911.57510.
4SPWoodie Fryman*32410.2864.3623174321119.213164581539969222518833.901.4219.
5RPWayne Twitchell2459.3574.06491517111139.213872636566112214602892.601.3898.
6RPBucky Brandon3177.5003.4542617002104.110649409468676134541053.731.4579.
7RPBarry Lersch2746.4003.043684310100.28637348337483134161193.551.1827.
8RPDick Selma2829.1825.5646102510398.29167611373958506459655.311.6628.
Did the fact that the other starters were not very good make Carlton more valuable?  Here were the leading 2014 Dodger starters:

1SPZack Greinke30178.6802.7132320000202.119069611943320720128211292.971.1528.
2SPClayton Kershaw*26213.8751.7727270620198.1139423993102392277491971.810.8576.30.41.410.87.71
3SPDan Haren331311.5424.0232320000186.01831018327367145318776874.091.1778.
4SPHyun-jin Ryu*27147.6673.3826260000152.0152605782921393026311032.621.1919.
5SPJosh Beckett3466.5002.8820200110115.2964137173921075014751214.331.1677.
Zack Greinke (17-8) and Hyun-jin Ryu (14-7) had good records.  How did that impact the evaluation of Kershaw?  They don't even play in the same games.  Paul Maholm started only 8 games, 5 during the absence of Kershaw; he also relieved in 22 games.  Maholm for the season: 1-5, 4.84 ERA, 70 innings.  Did Paul Maholm help Clayton Kershaw win the MVP award?

No comments: