PECOTA and the tribe
The PECOTA predictions from Baseball Prospectus are out. Here are my few takes on where the computer regressions may differ from conventional wisdom on key tribe members.
Hitters
PECOTA says that Jhonny Peralta is the 2nd best position player on the tribe in terms of WARP for 2008 and has the 2nd best career upside. It has always liked Jhonny partially because the skewed defensive metrics BP has. Still, Jhonny deserves much more love than he usually gets.
PECOTA loves Asdrubal Cabrera. For prospects it likes Carlos Rivero. Grady still has Barry Bonds as his #2 comparable.
PECOTA gives Hafner only 28 dongs. Ouch.
Predicts Andy Marte to be 50 OPS points less than Casey in 2008. Doesn't think much of Marte or Barfield (both with OBPs sub .310).
Pitchers
CC is in the top 5 pitchers, but gives CC a much higher attrition rate (14%) than the other top guys (Santana 3%, Peavy/Webb 8%). Something to think about when talking long term deals. Take this with a grain of salt as PECOTA has a hard time finding comparables for CC.
PECOTA loves Laffey. Basically says he is equal to Westbrook already in 2008.
PECOTA is not high on Carmona; it still thinks he is good, but expects a decent drop off. Only a top 30 pitcher in the majors, not top 10. Maybe it is the innings jump which it doesn't like (which would be reasonable). Laffey, Carmona, and Westbrook all have very high ground ball rates. Infield defense will be very important.
PECOTA hates on Sowers and Byrd. It doesn't like pitchers with what it considers no stuff.
Anyway, these are my takes on the numbers. It will be interesting how things shake out.
0 recs |
50
comments
Comments
Re: PECOTA and the tribe
PECOTA loves Asdrubal Cabrera.
and
PECOTA loves Laffey.
I'll choose to glance past the negative ones.
by Brick. on Feb 4, 2008 10:31 PM EST 0 recs
Re: PECOTA and the tribe
by CU Adam on Feb 4, 2008 11:06 PM EST 0 recs
Re: PECOTA and the tribe
by Brick. on
Feb 4, 2008 11:12 PM EST
up
0 recs
Re: PECOTA and the tribe
by fwembt on
Feb 5, 2008 12:33 AM EST
up
0 recs
Re: PECOTA and the tribe
Plus it looks like Fausto and Asdrubal.
by CU Adam on
Feb 5, 2008 2:08 AM EST
up
0 recs
Re: PECOTA and the tribe
http://cbs.sportsline.com/mlb/gamecenter/recap/MLB_20070517_MIN@CLE/rss
by rolub on
Feb 5, 2008 8:54 AM EST
up
0 recs
Re: PECOTA and the tribe
by Turkmenbashi on
Feb 5, 2008 9:03 AM EST
up
0 recs
Re: PECOTA and the tribe
...or maybe I was.
Boodles anyone?
by Harry Doyle on
Feb 6, 2008 4:35 PM EST
up
0 recs
Re: PECOTA and the tribe
I also wonder what his projection would have looked like had they extrapolated his post-May numbers out to a full season, rather than using his actual stats. Fausto's strikeout rate spiked after May, and he really doesn't look like the same pitcher if you look at post-May only. Specifically, he looks a lot less like a fluke.
Unless they've made changes, one minor weakness of PECOTA is that it doesn't consider in-season trends at all, which for a young player, or for a player injured early in the season, often would be significant.
by Jay on
Feb 5, 2008 10:38 AM EST
up
0 recs
Re: PECOTA and the tribe
by 7foot3 on
Feb 5, 2008 11:15 AM EST
up
0 recs
Re: PECOTA and the tribe
by Jay on
Feb 5, 2008 1:22 PM EST
up
0 recs
Re: PECOTA and the tribe
by 7foot3 on
Feb 5, 2008 1:28 PM EST
up
0 recs
Re: PECOTA and the tribe
by fleerdon on
Feb 5, 2008 11:43 AM EST
up
0 recs
Re: PECOTA and the tribe
by Thommy on
Feb 5, 2008 12:48 PM EST
up
0 recs
Re: PECOTA and the tribe
by Thommy on
Feb 5, 2008 12:58 PM EST
up
0 recs
Re: PECOTA and the tribe
by afh4 on
Feb 5, 2008 1:01 PM EST
up
0 recs
Re: PECOTA and the tribe
by Thommy on
Feb 5, 2008 1:53 PM EST
up
0 recs
Re: PECOTA and the tribe
The most basic kind deals with measuring raw production more accurately as it pertains to creating/preventing runs. Going from the Idiot Stats to something as basic as OPS.
The next deals with context -- correcting for differences in opportunity, park and league environments. EqA, Runs Created, etc.
Then we have the luck adjustments, going from ERA to RA to FIP, maybe looking at HR rates at the same time. Haven't really seen a compelling, consolidated stat that does this.
The next level would be more nuanced BIP adjustments, incorporating some degree of regression to personal and historical norms. In other words, correct the mistake of assuming that all pitchers/hitters regress to the same BIP or HR/F rates.
Back to Carmona, I think clearly he looks best considering the raw stats and traditional stats -- he nearly led the league in both ERA and Wins. He didn't look as good in FIP or xFIP, but the issue there is that they're "Level 3" stats, and Carmona really needs to be viewed under Level 4. (This isn't convenient homerism, either -- neither FIP nor xFIP gives any consideration to the fact that Carmona led the league by far in groundball rate, inducing about 150 more than your average #1 starter.)
What I've yet to see even theorized is a real synthesis of these, where you can look at a player's raw numbers, then neutralize it for context, then do a nuanced regression for luck. Basically, a BIP-regressed EqA or a context-neutral (and more nuanced) version of xFIP. I'm not sure why it hasn't been done, because the math doesn't seem that hard given that they already have some really good building blocks.
by Jay on
Feb 5, 2008 1:39 PM EST
up
0 recs
Re: PECOTA and the tribe
My guess is that this is a 2nd order effect in the sense that it is the combination of throwing hard and throwing heavy. PECOTA doesn't see the Ks so it assumes that they are not hard throwers, but that couldn't be further from the truth. This is something that any model will have difficulty with. PECOTA is a good starting base, but one must make adjustments over this base.
by oxforddave on
Feb 5, 2008 1:49 PM EST
up
0 recs
Re: PECOTA and the tribe
by Jay on
Feb 5, 2008 10:37 PM EST
up
0 recs
Re: PECOTA and the tribe
by Jay on
Feb 5, 2008 10:39 PM EST
up
0 recs
Re: PECOTA and the tribe
by CU Adam on
Feb 5, 2008 11:35 PM EST
up
0 recs
Re: PECOTA and the tribe
by gte619n on
Feb 6, 2008 8:00 AM EST
up
0 recs
Re: PECOTA and the tribe
by woodsmeister on
Feb 6, 2008 10:26 AM EST
up
0 recs
Re: PECOTA and the tribe
by 7foot3 on
Feb 6, 2008 12:17 AM EST
up
0 recs
Re: PECOTA and the tribe
I, on the other hand, agree with Jay and others that his Carmona's projectability is likely heavily variant, especially due to his improvement in K/9 during the second half of the season, and subjectively looking at his sinker as a pitch that not many pitchers have the ability to match in velocity and vertical movement. Also, if you look past FIP or DIPS, and talk about actual performance (ERA or RA) he is more dependent on his defense than most other pitchers (due to the GB% and k/9) and Asdrubal at 2B for the whole year, plus Marte overtaking at 3B would be an improvement on last years defense.
by hans on
Feb 6, 2008 2:19 AM EST
up
0 recs
Re: PECOTA and the tribe
APR: 3.12
MAY: 3.50
JUN: 6.62
JLY: 7.46
AUG: 6.00
SEP: 6.94
OCT: 5.14
PRE: 5.13
POS: 6.39
Also, there seems to be a bit of contreversy brewing over his middle name. Wikipedia says it starts with a 'J', while the RedSox-Yankees Man Love Association of Connecticut says it starts with a 'C'.
by gte619n on
Feb 6, 2008 8:05 AM EST
up
0 recs
Re: PECOTA and the tribe
by NickFantana on
Feb 6, 2008 9:15 AM EST
up
0 recs
Re: PECOTA and the tribe
by 7foot3 on
Feb 6, 2008 11:45 AM EST
up
0 recs
Re: PECOTA and the tribe
by hans on
Feb 6, 2008 7:48 PM EST
up
0 recs
Re: PECOTA and the tribe
Fausto's pre- and post- all star splits are not drastic enough, in my opnion, for us to say that it should matter. Yes, he struck out a few more guys post, but he also allowed disproportionately less hits (favorably so, or luckily so). I think his true talent level is somewhere between the two halves as a result.
We also have fantastic blinders on if we are mentioning his 2nd half performance and pretending like his 15 IP, 11 BB, 12 K, 7.20 ERA postseason didn't happen.
by Thommy on
Feb 6, 2008 10:25 AM EST
up
0 recs
Re: PECOTA and the tribe
by mauichuck on
Feb 6, 2008 10:31 AM EST
up
0 recs
Re: PECOTA and the tribe
Granted, it was the end of the season, and he had thrown more than he ever has before - I'm not knocking him. I'm just pointing out that we've kindly pretended that his horrible post-season didn't happen.
by Thommy on
Feb 6, 2008 10:44 AM EST
up
0 recs
Re: PECOTA and the tribe
by drerikbrady on
Feb 6, 2008 11:35 AM EST
up
0 recs
Re: PECOTA and the tribe
Thus, I think it's significant that his K rate more than double over the last four months what it was for the first two.
In the postseason, he got mugged by the Red Sox twice but only after totally dominating the Yankees -- and breaking all his own personal highs in IP. I will in fact totally ignore that.
by Jay on
Feb 6, 2008 11:44 AM EST
up
0 recs
Re: PECOTA and the tribe
by Brick. on
Feb 6, 2008 11:53 AM EST
up
0 recs
Re: PECOTA and the tribe
by Gradyforpresident on
Feb 6, 2008 12:10 PM EST
up
0 recs
Re: PECOTA and the tribe
by woodsmeister on
Feb 6, 2008 1:13 PM EST
up
0 recs
Re: PECOTA and the tribe
I'm otherwise with you, I'm sure we read much if not all the same stuff re: stuff, maturity, and what not. It is in my nature, however, to counterpoint very strong optimism (fausto) or pessimism (jhonny).
As for ignoring his post-season, I can within the context of the season, but can't (completely) disregard it because his year-year improvement in this area likely won't be complete. For that reason (in terms of discussing his projection for next year), I factor it in.
by Thommy on
Feb 6, 2008 12:30 PM EST
up
0 recs
Re: PECOTA and the tribe
http://www.hardballtimes.com/main/article/a-zone-of-their-own/
by 7foot3 on
Feb 6, 2008 11:51 AM EST
up
0 recs
Re: PECOTA and the tribe
I'm sure there were other games where Carmona faired just fine with a small zone ump behind the plate, but I find this interesting nonetheless.
by dvd1204 on
Feb 6, 2008 1:54 PM EST
up
0 recs
Re: PECOTA and the tribe
by mauichuck on
Feb 6, 2008 11:48 AM EST
up
0 recs
Re: PECOTA and the tribe
by supermarioelia on Feb 5, 2008 12:16 AM EST 0 recs
Re: PECOTA and the tribe
I don't even know what that means.
by Jay on
Feb 5, 2008 1:21 PM EST
up
0 recs
Re: PECOTA and the tribe
by fwembt on
Feb 5, 2008 1:49 PM EST
up
0 recs
Re: PECOTA and the tribe
I spent a bit going through the numbers, and here's what I got:
It cements Garko as a productive hitter in the middle of the lineup. Maybe it's just me, but I don't think we've appreciated his 2007 numbers enough.
It's got Dellucci as being a very productive hitter, although only over 156 ABs. His total value is low as a result, but I certainly hope his number of PAs will be 300 or so, moving him into the realm of being a very nice guy to have on your team. Chalk that up as a reason for optimism for 2008.
It reinforces my belief that Franklin is about the most interesting player on our offensive roster. I'm not sure it is going to be this year (2009 is more likely), but his ISO makes me think he's going to click before too long and be...I don't know...a lite Richie Sexson-type hitter?
It doesn't love Astro so much, either offensively or defensively, I think. His total value is a reflection of the poor quality and quantity of 2B, and him just not being too bad at anything. Mind you, this is only for next year, but the future is likely still bright.
FWIW, I'm in on the Laffey love. I think he is a LH Westbrook, but with the eventual upside of a higher strikeout pitcher. PECOTA isn't real strong on incorporating a single, previous season breakout for a pitcher. I think even keeping his head above water this year would lead to a pretty drastic change in his PECOTA next year.
It call for a yawner of a year for Choo, Kobayashi, and Michaels, the former two of which I think may be overly harsh.
by Thommy on Feb 5, 2008 12:27 AM EST 0 recs
Re: PECOTA and the tribe
I just don't get their defensive metrics. How can Peralta be an above average SS, and Asdrubal be an average 2B? They also have Miguel Cabrera as an above average 3B, and that is even worse than Peralta. They have to fix those metrics.
by oxforddave on
Feb 5, 2008 1:41 PM EST
up
0 recs
Re: PECOTA and the tribe
by mauichuck on Feb 5, 2008 1:00 AM EST 0 recs







