Monday, July 4, 2016

Advanced Player Rankings Using DICE

So far I have mostly been using a point system I learned 15 years ago from MLB.com's fantasy page to rank players. It helps measure the impact each player has had so far as a Firefly. But it doesn't really measure how well a player is doing. For pitchers, the point system gives a 30 point bonus for wins and saves and a 15 point penalty for losses. As most true baseball fans know, pitchers can't hit and therefore can't actually win and lose games by themselves so record stats aren't very good. But there is also controversy about whether ERA is a truly reflective stat on a pitcher as well. If fielders do poorly but don't make errors, a pitcher's ERA will suffer arbitrarily. That is why most advanced statisticians try to evaluate pitchers by only what they can control: walks, strikeouts, hit batters, and home runs (of course a small fraction of home runs are fieldable, and wild pitches probably should be included as well). The Defensive Independent Component ERA (DICE) projects what a pitcher's ERA would be using only those four stats in accumulated during his innings pitched. Unfortunately, this may have influenced the slowdown of baseball in recent years as pitchers at this level trying to impress the organization by making sure they are the one who beats the batter and not the fielders.

So let's now look at how the Fireflies' pitchers fair in DICE. These rankings are sorted by difference between the real ERA and DICE:

1. Nicco Blank (ERA: 10.22, DICE: 7.78, Difference: 2.44)
The huge difference between Blank's ERA and his DICE is because pitchers normally don't have an ERA over 10. Hopefully with more innings pitched both numbers will drop, particularly the real ERA.

2. Chase Ingram (ERA: 5.22, DICE: 3.48, Difference: 1.74)
One thing DICE does not measure is being susceptible to a few bad innings, which Ingram has had recently. Ingram being left in to allow 11 runs in less than four innings complete destroyed his real ERA, but not his DICE. Ultimately though the goal in baseball is to not give up a lot of runs however, so hopefully Ingram will recover from his recent outings. At the same time as long as Ingram doesn't make bad innings a habit he will put the Fireflies in position to win more often than not, which would mean he is more effective than his real ERA.

3. Joe Shaw (ERA: 4.72, DICE: 3.13, Difference: 1.59)
Shaw likewise has struggled with his consistency and has had some bad games recently. But when he is on he is capable of overpowering batters which is what DICE likes to see. DICE however does not take note of giving up runs in bunches as is also the case with Ingram.

4. Thomas McIlraith (ERA: 6.23, DICE: 4.99, Difference: 1.24)
McIlraith had a few bad outings most likely due to an underlying injury that has kept him out since. He can be overpowering (like when he through 6 innings of a combined no-hitter) but became wild as he tried to pitch through an apparent injury. Hopefully when he comes back he will be given a tight pitch count so he can recover.

5. Alex Palsha (ERA: 2.67, DICE: 1.69, Difference: 0.98)
Based on real ERA Palsha is in the middle of the pack among bullpen pitchers but DICE has him as the best, which may be why he is the team closer. Like most of the starters with a weak ERA, Palsha has been hurt in his real ERA by a couple meltdowns, most notably on Opening Night in Charleston. But ever since then he has been a dominant pitcher which is reflected in his DICE.

6. Tyler Badamo (ERA: 3.92, DICE: 3.40, Difference: 0.52)
Surprised a bit that a low strikeout guy is looked upon favorably by DICE, but there is still something to be said about having good control. Badamo doesn't beat the batters, nor does he beat himself. It's possible however that Badamo's stats could suffer if he gets promoted to a level where players have more muscle to hit home runs off him.

7. Christian Montgomery (ERA: 3.34, DICE: 3.17, Difference: 0.17)
Montgomery has a few bad innings, particularly one this year where he walked a couple guys and then gave up a home run when the pressure was on to throw strikes. That is reflected a bit though in his DICE as well as his ERA. But generally DICE still rewards power pitchers more than ERA does.

8. Andrew Church (ERA: 0.71, DICE: 0.63, Difference: 0.08)
In two starts Church had an impressive ERA. But his DICE was even more impressive due to 15 strikeouts in just over 12 innings pitched with no walks. That is why he did not stay very long here at all. Now the real question is why didn't the Mets have him pitch here in April? Are they trying to keep good prospects out of Columbia?

9. Kevin Canelon (ERA: 3.27, DICE: 3.40, Difference: -0.13)
Now we have the first Firefly pitcher whose DICE is higher than his ERA, meaning based on fielding-independent stats Canelon is slightly lucky to have an ERA as low as his. That's because he is slightly more overpowering than Badamo but has also been more home run prone when he gets in trouble (he leads the Fireflies in HRs allowed with 9). But he also has the most innings pitched with Conlon promoted, Ingram being forced into early exits, and Badamo now hurt so his ERA and DICE are probably both representative of how he is doing given the narrow difference.

10. Johnny Magliozzi (ERA: 4.32, DICE: 4.59, Difference: -0.27)
I'm actually slightly surprised DICE doesn't dislike him more, given that he is not particularly overpowering and leads the bullpen with 6 home runs allowed. But his K/BB ratio is 3.71 which is pretty solid.

11. Craig Missigman (ERA: 3.07, DICE: 3.68, Difference: -0.61)
By contrast Missigman's K/BB ratio is 1.73 which is not particularly good. Missigman when he has good control is solid as a steady and smooth pitcher, but can get into trouble by walking guys.

12. P.J. Conlon (ERA: 1.84, DICE: 2.64, Difference: -0.80)
DICE still regards Conlon as the best of those with three or more starts but not by as much as his regular ERA does. That's probably why it took him half a season to be promoted. Yet there is something to be said about a pitcher who is consistent which DICE doesn't reward much.

13. Tyler Bashlor (ERA: 1.47, DICE: 2.75, Difference: -1.28)
It's hard to believe Bashlor leads the bullpen in ERA. and DICE has a hard time believing it as well. He has 4.4 walks per 9 innings and a WHIP of 1.09, which is decent but not great like his real ERA. Bashlor however has a high strikeout rate which I thought DICE would love more, and there is something to be said for being avoiding the big hit (or perhaps it's luck). What DICE does not notice about his walks is that unlike some pitchers his walks don't snowball on him into completely falling apart (as is the case for most of the rest of the bullpen).

14. Seth Davis (ERA: 2.27, DICE: 3.61, Difference: -1.34)
Davis has been effective this year despite not being a prototypical pro pitcher. That's something most scouts don't like and neither does DICE. He still has value though as a consistent left-handed pitcher which every bullpen needs.

15. Gaby Almonte (ERA: 4.29, DICE: 5.86, Difference: -1.57)
Almonte is not overpowering and fell victim a bit to the longball in his last outing. It's hard to make use of his numbers yet though with just 4 starts.

16. Witt Haggard (ERA: 3.55, DICE: 5.29, Difference: -1.74)
He's getting better, but he still hasn't thrown enough innings for DICE to excuse his sloppy start in Charleston before going on the DL. On a side note, it's odd to see a reliever get a decision in half his outings. That's not normally what you want though as his record is 1-4.

17. Bryce Beeler (ERA: 0.00, DICE: 1.89, Difference: -1.89)
He was dominant in his first start without being overpowering. Obviously he needs a lot more outings than just one before these numbers become valid.

Tomorrow I'll be using DICE to look at the batters. Yes, I'll be using a pitching stat to look at hitting!

No comments:

Post a Comment