A few weeks ago I did a post on which stats correlate with the mvp winners over the past 
few years. I found that batting average correlates better than some of the more advanced statistics likely because of the uneducated voters. Many of the people who vote for the MVP vote based on the traditional stats for better or worse. The problem is that these people are also the ones that vote for the Cy Young. I remember how happy the sabermetricians were with last year’s King Felix Cy Young winning because he only won 13 games. I want to do the same thing I did with the MVP and apply it to pitchers (because I never give them enough love). Anyway, I’m going to look at 2010, 2005, 2000, and 1995 to see:
- If the voting has changed over the years
- The stats that best correlate with the winners over the past few years
First up, ERA. We think that generally ERA is the best indicator of how well a pitcher is pitching. What matters more than the runs a pitcher allows? I contend nothing. However, the voting may indicate that dominance is more important. ERA had correlation numbers of .165, -.116, .62, and 0. Its fairly safe to say that ERA’s correlation numbers aren’t the best indicator for how well ERA and Cy Young standing went together. Of the top 4 Cy Young finishers every one finished in the top 10 in ERA that year.
Innings Pitched can be a good indicator of success. If you pitch a lot of innings you are effective in those innings, durable, and your manager chooses to let you continue to pitch. It makes sense that the best pitchers in a given year would pitch the most innings. The average correlation for IP is .474. Not bad.
I also looked at strikeouts. Average of .354 but, there was less correlation than that. Within the top 5 Cy Young finishers, we had only 8 of the 20 finish in the top 5 in strikeouts.
Next we have WAR. Wins above replacement is supposed to combine all factors to discover the true worth (in wins) of a player. In 2005, the top 5 Cy young Winners finished 1, 2, 3, 4, 6 respectively in WAR. The average correlation for WAR was .606.
Here are the average correlation numbers... in chart form
Wins are thought to be the most outdated of statistics in baseball. Too many things can influence the number of wins a player gets (bullpen blows, or runs scored while the pitcher is in the game). Its really a dumb statistic but, voters have put emphasis on it over the years. Average correlation for Wins: .536. But, there are some instances when a pitcher gets Cy Young votes for little more than his win total. Ramon Martinez in 1995 finished 3rd in wins and 2nd in IP but, 18th in ERA, 15th in SO’s, 29th in WAR, and 30th in FIP.
Fielding Independent Pitching average correlation of .464. Little else can be said of the trend, but FIP is an important statistic for those sabermetric fans out there.
I also wanted to look at the change in voting over the years. So I generated a chart
Here you can see the trends of statistics. I ran into a problem transferring the key over from Excel but, blue is ERA, yellow is IP, red is Strikeouts, purple is WAR, grey is wins, and he lighter blue is FIP.
What I notice is the trend in voting away from innings pitched. I also notice that the grey line (wins) seems to be going up. Cy Young voters have started to like winning pitchers more in recent years! Congratulations to King Felix but, this data suggest that the voters are putting an increased emphasis on Win totals.
I also wanted to point out that teammates were nearly all in the top 5 in Cy Young votes. Clemens, Oswalt, and Pettitte received votes in 2005 while they were Astros. Tom Glavine, and Greg Maddux both received votes in 2000 and 1995 and a Cardinal’s pitcher finished in the top 5 every year. So, if you’re trying to win the Cy Young, it looks like you should win the WAR race and have a teammate who’s a really good pitcher…. OR pitch for the St Louis Cardinals.
Stat of the Day: The Shawshank Redemption is cinematic excellence.
-Sean Morash
