sitemeter:

August 7, 2009
  
You Are the Eventuality of an Anomaly, Omir

“In statistics, regression toward the mean refers to the phenomenon that a variable that is extreme on its first measurement will tend to be closer to the centre of the distribution on a later measurement. To avoid making wrong inferences, the possibility of regression toward the mean must be considered when designing experiments and interpreting experimental, survey, and other empirical data in the physical sciences, life sciences, behavioral science, and social sciences.” (Per Wikipedia)

All season long, many analytical Mets fans have been waiting for Omir Santos to regress to the mean. The reasoning for the latter is based on the fact that Omir’s empirical data (his minor league numbers) should represent his true ability at playing baseball.

Omir’s 9 year minor league career:

.258/.304/.348/.652, .090 ISO with only 120 BBs in 2440 PAs.

Omir’s 2008 season as a 27 year old in AAA

.269/.328/.323/.651, .054 ISO and a .331 BABIP

Given Omir’s minor league numbers, it would be a fair to infer that Omir is a poor backup catcher in AAA and probably has no future in the majors. However, with Schneider injured and Minaya’s poor backup plans, Omir was called up. In 26 April at bats, Omir slugged his way to a .269/.269/.500 line. Given his limited success and the lack of love for Castro, Omir was given 54 at-bats in May, producing a .278/.317/.463 line and a handful of clutch hits. To the analytical Mets fan, the writing was on the wall. Fans and management never fully embraced the oft-injured Castro and with Schneider coming off the DL, Omir’s .776 OPS best two months of his career start was going to replace Castro’s 806 OPS over the past three seasons. Since Omir was already 28, his hot start to the season seemed like a lucky streak, even Al Leiter and Figgy both hit a triple once, and he would inevitably regress to his true talent level. During the month of June, regression did seem to hit Omir as he proceeded to hit .246/.259/.333/.592 in 57 at bats. Then a funny thing happened, Omir bounced back in July with a .250/.321/.417/.737 and is 5 for 9 with a double in August.

Overall, Santo has been an above-average hitting catcher:

Average catcher – .251 EqA

Omir Santos – .260 EqA, 4.9BB%, 16.5K%, .155 ISO, .301 BABIP

For full disclosure, I was a huge fan of Ramon Castro; I don’t understand why some veterans are praised for “knowing their role as a backup,” but Ramon, who I’ve never actually seen quoted on the subject, is chastised for not wanting to be a starter. I also don’t understand why his defense was seen as poor even though the numbers don’t back up the claim and why his game calling was questioned. So, a part of me was frustrated when Omir was succeeding, though once Castro was traded, I became a fan because I care about the jersey more than the players. Omir is a flawed player whose power makes up for his low on base percentage; however, he hasn’t shown this kind of power in Triple-A or Double-A and, being 28 years old, shouldn’t have developed additional power. Santos batting line is not luck based on his BABIP but what about his power or his decreased strikeout rate. For help, let’s look once again at Pizza Cutter’s research on how many plate appearances are needed for a .50 r-squared to be reached.

Strikeout rate: 150 PA
LD%: 150 PA
Walk rate: 200 PA
GB%: 200 PA
GB/FB: 200 PA
FB%: 250 PA
Home run rate: 300 PA
HR/FB: 300 PA
BABIP: Doesn’t reach a 0.50 r-squared at 650 or below.

Santos currently has 210 plate appearances, thus a number of components that make up his hitting represent his true ability level this season. However, he still needs 90 PAs before his power can be declared for real. Still, Omir is having a very good year and it’s becoming harder to label it a fluke. It’s possible that Omir hasn’t regressed to the mean yet because his true ability level, or his mean, is much higher than in seasons prior, which would be rare for a 28 year old but, again, possible. Phil Nevin didn’t become a full time player until he was 29 and displayed much more power than his minor league numbers would indicate. For a more recent example, Ben Zobrist, who is also in his first full season as a starter and is 28, currently has an ISO of .272 which is way above his minor league .141 ISO. It is also possible that in Omir’s next 90 or so plate appearances, his power regresses or pitchers make additional adjustments. (Hint: Loves fastballs and cutters, but off speed stuff? Not so much, according to his pitch type values.) If Omir can continue to hit at his current pace combined with his average-ish defense—22.6% CS rate but only .209 PB+WP/Game, slightly better than Schneider—he would be an adequate backup catcher next year at a low price. I still have my doubts but it would be sweet to get to wear my shirt for another year.


12 Responses to “You Are the Eventuality of an Anomaly, Omir”

  1. Comment posted by Kneel Before Dep! on August 7, 2009 at 9:16 am (#1052604)

    good article.

    castro’s numbers with the white sox:

    .136/.255/.273/.528 OPS+: 37

  2. Comment posted by Danny on August 7, 2009 at 9:23 am (#1052608)

    I think my problem with Castro is that he didn’t put in the work to get himself into good enough shape to be a starter. He’s lazy. He has great talent and pissed it away.

  3. Comment posted by John on August 7, 2009 at 11:02 am (#1052769)

    I don’t get the 90 more pa’s to see if his power is real

    you mention the minor leauge stats but then only use 09 data. Are we to totally dismiss the fact he didn’t hit for power in the minors? That’s alot of pas to ignore

    90 more pa till we regress 50% to the mean. I might add this isn’t the MLB mean. It should be the mean of other catchers. The more info on omir the better we’d know what to regress too

    I guess what in saying Is if we knew nothing other than those MLB pa we regress 50 but we know so much more

  4. Gravatar
  5. Comment posted by Joe Sokolowski on August 7, 2009 at 11:13 am (#1052783)

    Hey John,
    That’s an excellent point. I was theorizing in my article that Omir has somehow developed beyond what his minor league numbers would indicate his ability was, I’m not familiar enough with his batting approach or physical shape to guess what has changed.

    As to your second point, I have not seen any research indicating that that statistics of players of different positions would become significant at different PAs but it’s an interesting idea.

    But yeah, I hear you, it’s always bad to ignore more data if it’s available. Really, what I was trying to say was that I was sure Omir would fail but he hasn’t and it’s either because he’s is having a lucky season (which becomes more unlikely as he gets more PAs) or his true talent has increased.

  6. Comment posted by John on August 7, 2009 at 11:52 am (#1052855)

    I think what happens is most research always use the mean as the mean of all MLB players….I would imagine tho if u are looking at a subset of players (catchers in this instance)…their mean is different from the overall mean…..and if thats the case I think it would mean it becoming significant at different PA’s……I’d have to imagine that the different would be little tho. I’m really not sure tho. Now I feel like asking someone that question lol

  7. Comment posted by John on August 7, 2009 at 11:59 am (#1052863)

    Also, it’s certainly possible Omir has developed beyond his minor league stats.

    When we are regressing, its an estimate of a players true talent level, however we all know that true talent level changes on a continual basis. If we are doing projections, not sure we’d wanna eliminate the minor league stats but probably weigh the more recent stats greater.

  8. Gravatar
  9. Comment posted by Joe Sokolowski on August 7, 2009 at 1:04 pm (#1052923)

    The difference in means would probably be small and might just be noise but it is an interesting questions.

    Usually when I do “projections” I use the simple weighted average of 5-3-1, and while the exercise is alright for established ML, it’s really flawed for younger players as it’s (obviously)impossible to project improvement using this method.

  10. Comment posted by littlefallsmets on August 7, 2009 at 1:09 pm (#1052926)

    He’s not as godawful as Schneider and so I can’t complain.

    I realize it’s a bit like telling a Yugoslavian that, hey, Tito wasn’t as bad as Hitler but there they were and here we are.

  11. Comment posted by John on August 7, 2009 at 2:40 pm (#1053083)

    Exactly.

    I think Marcels usually for young players just assume they are league average.

    I think in my projections with the mets (which totally went out the water with the injuries) I used 3 years and weighed everything evenly…..thats wrong…..I probably should weigh more recent stronger somehow. Not sure really what weights to use.

  12. Comment posted by John on August 7, 2009 at 2:40 pm (#1053085)

    water-window

    Cant type today

  13. Comment posted by istealllamas on August 7, 2009 at 6:03 pm (#1053450)

    Phil Nevin, we should note, was probably on steroids at the time.

  14. Comment posted by mrjimi on August 7, 2009 at 6:09 pm (#1053453)

    FYI- Santos, as a backup last year here in Norfolk, fashioned a 21 game hitting streak, 3rd longest in the International League in ’08. The pre-Matt Wieters catching situation was screwed up last year and Santos probably made more starts than projected…

rss/syndication: