That's What Cheesehead - 38.30 - Old School Clashes with New School

[7teen]

There will always be debate on what stats matter most or are most important. One person may value one stat over another while the other person values the exact opposite. What stats matter, or at least reflect good performance, is becoming a major challenge for Wolves manager Millard Davis as he tries to decide which pitchers to give innings to and which pitchers may need a cut on his outings. As a case study, check out these three pitchers based on different Old School versus New School Stats.

ERA:
PItcher A: 4.73
Pitcher B: 5.40
PItcher C: 6.49

Clearly here everything indicates that Pitcher A is the better of the 3. Obviously you can't just isolate this one stat in itself, especially without other defining factors. One of those would be Innings Pitched.

IP:
Pitcher A: 59.0
PItcher B: 71.2
Pitcher C: 68.0

So PItcher A has the better ERA but hasn't quite thrown as much. Obviously anything could happen over the next 9-12 innings it would take pitcher A to catch the other two. Does his ERA balloon up or does it decrease even more. Everything in Davis' mind up to this point suggests he should perhaps increase Pitcher A's usage to see.

WHIP:
Pitcher A: 1.36
Pitcher B: 1.60
Pitcher C: 1.93

Another older school stat, but again, it indicates that Pitcher A is the better of the trio. He's allowing less base runners over his less time pitched but the stat indicates he's the one who should be used more.

BB/K
Pitcher A: 18/52
Pitcher B: 32/27
Pitcher C: 21/32

Perhaps a stat that transcends time, the number of walks to strikeouts typically indicates how effective the pitcher is in himself to get outs or surrender base runners without worrying about the defense. Pitchers B and C, despite walking more guys in more innings pitched, have also struck out fewer batters. So up to this point, Pitcher A has more Ks, a lower ERA, and a lower WHIP.

If you're an older school guy you have to be thinking A deserves more of a shot, right? When you combine in the fact he has 5 wins on the year and the other two guys have 4 and 1 respectively, it screams use A more often. But then we start to look at the new age stuff.....

WAR:
Pitcher A: -0.8
Pitcher B: 0.3
Pitcher C: 0.9

Now, let's not go crazy here. None of these are particularly good for this point of the season. I think it was obvious anyway had you gotten this far on a Wolves pitching TN anyway. But it's sort of mind boggling to look at those first series of stats and then find out that Pitcher C is worth more wins than Pitchers A and B. The new age stats of FIP suggest that pitchers B and C are performing better than their numbers really are which ties into the overall WAR numbers. What has hurt Pitcher A, I suppose, is the fact he's allowed 20 homeruns over 59 innings pitched, meaning he's on record setting pace if his innings were increased. Meanwhile Pitchers B and C are ground ball guys not getting a lot of defensive support from the Wolves swiss cheese infield defense.

But then it comes back full circle. If you're Madison, who do you use more often. Pitcher A who has shown to be effective old school (Juan Trujillo) or Go with Pitchers B (Mario Gonzalez ) and C (Enrique Gomez) who haven't gotten it done by an old school metric but new school thinking suggests their better!?

[RonCo]

These are always the fun questions.

And then there's always the idea that big picture thinking makes the best bets, but anecdotes make the best TNs.

[Ted]

I've never really totally been sold on FIP in OOTP. I think the reason for that is that the pitching model is not entirely realistic. A pitcher like Trujillo doesn't exist in real life. (And his results won't occur in a large sample.) Real pitchers don't have elite swing and miss stuff, pinpoint control, and serve up dingers at a ridiculous rate. So the reason FIP doesn't work well for him all the time is that an unusual number of his HR allowed are solo shots.

For most guys, with more reasonable ratings, FIP works just fine. But video game players like Trujillo break the model.

The more common place FIP doesn't work in well in OOTP is low stuff, high mov, high control pitchers. While there are guys in real baseball who can limit hard contact without striking out a lot of players, the degree to which this is possible in OOTP isn't very realistic.

This all happens because of the movement (which is really HR prevention) is separate from STUFF in OOTP which isn't how it actually works. HR prevention isn't an independent pitcher skill, that is independent of actual pitches and command. So while OOTP continues to have this model, FIP will always be limited/unreliable to a small degree in pitcher analysis. (And as such, pitcher WAR will less reliable as well.)

[RonCo]

It's times like these that like to remember that Ferguson Jenkins led the league in HR allowed seven times and made the Hall of Fame.

[RonCo]

FIP is like all other generalized sabermetric stats...it's generalized. So it generally works, but, yeah, when you get out on the edges of things, you can warp the stat.