Perverted by Numbers: Is the King Felix/Sabathia gap widening as quickly as Mark E Smith’s well dug trench mouth? Marchman inspects.
The other day I was trying to think of a way to argue that CC Sabathia is better than Felix Hernandez that didn’t involve pointing out that Sabathia has won more games, because it would be clever and thus virtuous.
Sabathia has pitched less, struck out fewer batters, walked more, given up more hits and home runs, pitched seven or more innings less frequently and pitched eight or more much less frequently. He’s faced easier competition, having pitched nine games against teams in contention right now against Hernandez’s 14, and worked in front of a better defense, as New York leads Seattle in turning batted balls into outs. He may have pitched in worse weather. I don’t know that he has. If it he has it would be a point for him.
...This leaves the second point, which would be that New York is world capital of politics, finance, the arts and baseball while Seattle is a miserable ######## full of sad, small people living unimportant lives, so that what happens in one place counts for more than what happens in the other. I don’t know if this is a position you can really argue someone out of. The best you can do is encourage them to really own it. (Incidentally, people who argue this point, no matter how many euphemisms they use, always sound to me like Mark E. Smith in ‘The Classical’: ‘HEY THERE ########!! HEY THERE ########!! There are twelve people in the world! The rest are paste!’)
All this so, the most interesting thing about the Cy Young race is probably just how little credit Hernandez is getting for even running up a winning record while being supported by a lineup that may as well comprise two winos drinking Sterno strained through stale bread, a Mongolian who learned how to hit by playing Bases Loaded, Alan Partridge, and five cardboard cutouts of Ski Melillo.
Reader Comments and Retorts
Go to end of page
Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.
But if you really wanted to make a non-wins-based argument for Sabathia, why wouldn't you start with something like ESPN's park factors?
"Ah-haaa!"
While there are certainly pitchers that you can argue are more deserving than Sabathia, I don't see how you can really make an argument for anybody other than Hernandez or Sabathia as being #1, if that makes any sense.
Basically, if your standard is best at preventing the other team from scoring runs, the winner's Hernandez in a landslide: according to BB-Ref, he leads the AL in every sabermetric counting measure, WAR (5.7, next best is 4.8), Adj. Pitching Runs, Adj. Pitching Wins, Base-Out Runs Saved, WPA (tied with Soria), WPA/LI, and Base-Out Wins Saved. He also leads in IP, K, and GS (tied w/ Sabathia).
If your standard is best at contributing to team wins, the winner's Sabathia in a landslide: he leads in wins (19, next best is 17), is 2nd in winning percentage, 2nd in IP, and tied for 1st in games started.
Any argument for a pitcher that gets them above Sabathia leaves them behind Hernandez, it seems to me.
I noticed that in a CarGo thread yesterday- the Humidor seems to no longer be in use, Coors is the Coors of 5 years ago
Actually, I'm not sure what is going on- Felix has an ERA of 2.30 and an ERA+ of 176
league ERA is 4.15
to get to an ERA+ of 176, Sean has to be using a park adjustment factor around 97 - which based upon ESPN's raw park factors is not even close....- it should be around 91 which would give him an ERA+ of 164
CC has an ERA of 3.13, to get to an ERA+ of 127 you would use a park factor of 96- which is lower than the one Sean is using for Safeco- that can't be right- raw home/road numbers say that New Yankee Stadium is a MASSIVE hitters park-the park adjustment factor should be around 108, and his ERA+ around 143 rather than 127.
164 is still better than 143
also Price's 146 is too high, it should be 133-135
also Ubaldo's 160 should be around 170
Josh Johnson's is about 5 too high
Halladay's is about right
Wainwright's is about 1 too high...
The NL ERA+ title is/should be a 4 way dead heat
I'm 99% sure that Sean uses the previous year's park factor in-season, so he's using 2009 park factors right now. If you look at the Yankees' 2010 team page, it shows a pitching park factor of 96, both 1-year and multi-year, which matches the Yankees' 2009 team page exactly. And, of course, the 1-year and multi-year are the same in this case because 2009 is the only previous season for New Yankee Stadium.
Eric Walker has demonstrated to me that any sample size smaller than 3 years does not increase the accuracy of adjustments (Harold Brooks has demonstrated to my satisfaction that seeming changes in single year park factors are something close to 80% noise)
But using multi-year park factors you can miss real change (as appears to be happening in Colorado). We know that things like game time temperature and wind (strength and/or direction) have fairly predictable impacts on run scoring. Multi-year park effects may miss unusual weather patterns for instance.
The way I like to do it is rank them using the multi-year park factors and the single year park factors and just scan the two lists. For most players there won't be much of a change.
so am I, I wonder what WAR is using as well since Sean gets that from another source.
Yes, personally I don't look at in season splits until, well now, September
It's pretty clear now that the 2010 park factors for certain parks- Safeco, Yankee Stadium, Coors, are going to be significantly different than 2009's- to the extent that certain players currently being discussed as MVP/CY Young candidates may have to be evaluated a little bit differently:
Better than they look by current ERA+/OPS+:
CC Sabathia
Ubaldo Jimenez
Longoria
Not as good as they look by current ERA+/OPS+:
King Felix
Robbie Cano
Carlos Gonzalez
David Price
Yes
You could argue that but you'd be wrong- the gap between them is just too much- .85 runs per 9, teh park factors just can't make that up
the real interesting one is Ubaldo- 2.76 ERA in Coors (acting like pre-Humidor Coors)
that actually should put him right there with Wainwright, Johnson and Halladay all around 2.30
Actually, the difference between Safeco (0.793) and Yankee Stadium (1.193) more than makes up for that (according to ESPN, anyway). But like I said, it's not like they pitch all of their games at home.
The park factor doesn't directly measure the run environment. If we wanted to look at the actual run environment for a given team-season, we could look at the actual number of runs scored in their games. That's the *real* run environment.
Why don't we do that? We know that the given team's offense and defense aren't average, and that the total runs is likely influenced a lot more by the team; we're really trying to find the run-scoring level of the environment, and not the actual run-scoring level.
So we use park factors; we essentially decide that the theoretical run-scoring environment is league average, offset by a park factor. To compute the park factor, we look at the distribution of runs between home games and road games. However, the park effect is only one part of that distribution; there are differences in schedule, in playing time, in weather, and just random chance and the like.
Whether we use single-season or multi-season park factors, either is just a way to estimate a run-scoring environment. The multi-seaons PF uses more input data, and therefore better captures signal rather than noise (even if the signal is a bit distorted by year-to-year changes).
You must be Registered and Logged In to post comments.
<< Back to main