Fred Lewis, BABIP exception to the rule?

by FairweatherFan on Jul 2, 2008 9:23 AM PDT

"Conventional" wisdom states that any given player can be expected to maintain a BABIP of approximately .12 + their line drive percentage.

Now, that is sort of some funny math - but for a guy with a 20% LD% that means you would expect a BABIP of .320.

Fred Lewis currently has a BABIP of .366, yet a LD% of only 18%. The sabermetricians amongst us would insist that he will regress to the mean, and that his BABIP is due to fall into the ~ .300 range. It seems however that he has settled into a pretty consistent level of production at the plate.

I can see two reasons why a player like Fred Lewis may present an exception to this "rule":

1.) His speed allows him to both run out infield hits as well as hurry infielders in making plays, thus leading to errors and mistakes resulting in his reaching safely. (While errors don't directly increase BABIP, hurried plays that are scored as hits do).

2.) His propensity for driving ground balls up the middle. (Hit em where they ain't)

In both of these cases, he is increasing the chances of his ground balls going for hits. I do beleive that it is possible for individual players to deviate from the established sabermetric rules due to propensities or talents like the two I listed.

So, if I can get the poll to work properly, please vote and discuss.

Oh yeah, 75 75 75 75.

Poll

Will Fred Lewis maintain a BABIP well above his "predicted" value?

This FanPost is reader-generated, and it does not necessarily reflect the views of McCovey Chronicles. If the author uses filler to achieve the minimum word requirement, a moderator may edit the FanPost for his or her own amusement.

3 recs | Comment 94 comments

Story-email Email Printer Print

Comments

I think that Fred is pretty much doing what he'll do

And I am fine fine fine with that

¿Julio is tourist in San Francisco? Harper's Bizarre!

by hairball on Jul 2, 2008 9:31 AM PDT 0 recs

I actually saw him smile yesterday. At the end of the game. You almost never see his teeth.. Would you be surprised if he had a grill under there? =)

by GiNgiNxbOi on Jul 2, 2008 9:44 AM PDT up 0 recs

I like the 75 filler when you’re already past 75 words. That’s a classy touch.

Trent Kline: Decentish. Also, my website is called ChatterBalks Dot Com and on it I make jokes about things.

by groug on Jul 2, 2008 9:46 AM PDT 0 recs

I think we’re jumping the gun quite a bit to assume after half a season that Fred is consistently capable of a higher BABIP than you would expect from his LD%. I love him and I don’t want it to happen, but significant regression is still a distinct possibility.

Matt Cain: He'll save children, but not the Dodger children.

by jponry on Jul 2, 2008 9:57 AM PDT 0 recs

And even if he doesn’t regress, he’d hardly be the first player to outperform his xBABIP for a season.

Neglectful father of David Quinowski

by marcello on Jul 2, 2008 10:18 AM PDT up 0 recs

Speed and BABIP

The answer is no, Fred Lewis isn’t the exception to the rule. Although, at First Inning they have him with a LD% of 21%, which puts him much closer to his expected BABIP.

Neglectful father of David Quinowski

by marcello on Jul 2, 2008 10:17 AM PDT 0 recs

My #'s came from fangraphs.

That is quite a discrepancy. I wonder why…

Eugeniooooooo!!!!

by FairweatherFan on Jul 2, 2008 10:30 AM PDT up 0 recs

Also

I think it has less to do with speed and more to do with location. Hard ground balls up the middle are much more likely to go for hits than those hit to either side of the infield, and Lewis hits a lot of those.

Eugeniooooooo!!!!

by FairweatherFan on Jul 2, 2008 10:31 AM PDT up 0 recs

Yeah, but where’s the proof that he hits an unusual number of ground balls up the middle? Or that that is a repeatable skill?

Neglectful father of David Quinowski

by marcello on Jul 2, 2008 10:46 AM PDT up 0 recs

Well, hit tracker could provide that info - how avaliable is it ?

As far as it being a repeatable skill – I think so. As much as some players are dead pull hitters and others go opposite field, it is reasonable to believe that Lewis tends to square the ball up very perpendicular and shoot it strait up the middle more often than the norm.

Eugeniooooooo!!!!

by FairweatherFan on Jul 2, 2008 1:47 PM PDT up 0 recs

Confirmation

Www.mlb.com’s hit chart on Freddie at AT&T confirms that the plurality of Freddie’s outfield hits come in center field, with left field coming in a close second and right field a distant third.

One thing about hitting in a particular area though. Eventually teams will play you there and take away hits.

by sharksrog on Jul 2, 2008 5:00 PM PDT up 0 recs

Not proof

but the mlb.com stats for players provide a function to view all balls hit by the batter in each ballpark he plays in by result. So you can select Lewis and then outs at AT&T to see where all his outs are made. Then click to get all his singles.

Not scientific, but then you view all the balls in play for outs and singles and see whether it seems like he gets a lot of hits up the middle or not.

Adoptive parental unit of Kevin "Most Spectacular Pitcher" Pucetas.

"I'm a Giant now... I like watching the ball get up there" - Wendell Fairley
"I'm really proud to be on this team." - Nate Schierholtz

by obsessivegiantscompulsive on Jul 2, 2008 1:48 PM PDT up 0 recs

Thanks, Martin

I’ve always wanted to get a closer look at Fred’s balls.

"I been waitin' a long time for this! I been waitin' since the f**kin' amateurs!" --WILL "THE THRILL" CLARK

by Josh from Hollywood on Jul 2, 2008 2:21 PM PDT up 0 recs

I looked at that too. His singles seem to be pretty spread out, but there is a bit denser clump in center field. Of course, I have no idea what most players are like in that regard.

Less arm, more talk. Raisingcain is a GAMER.
Adopted Giant: Henry Sosa

by raisingcain on Jul 2, 2008 2:22 PM PDT up 0 recs

I will point out that it doesn't take very many groundballs going for hits instead of outs to bump BABIP significantly.

Point being, even a minor propensity to hit the ball up the middle more often + a few infield hits legged out could result in a significant increase in BABIP over xBABIP

In Fred’s 315 PA’s so far this year, if ~ 10 of his GB went for hits instead of outs as compared to “average” it would explain the abnormal BABIP.

He has 7 Infield hits this year. How many of those would have been outs if he were “average” speed? How many more ground balls up the middle has he hit as compared to “average” ground ball distribution?

Probably only takes 3-4 ground balls up the middle + the Infield hit’s to explain his current BABIP.

Eugeniooooooo!!!!

by FairweatherFan on Jul 2, 2008 2:38 PM PDT up 0 recs

Okay, I'm Bored

So Fred has hit 26 singles at home this season. Of those 26, 14 have been to center field, 4 to right field, and 8 to left field.

Without really having any sort of average to compare that to, I’m not sure what that means. However, he does hit the ball up the middle more often than other places.

Eugeniooooooo!!!!

by FairweatherFan on Jul 2, 2008 2:51 PM PDT up 0 recs

EDIT:

He does get more hits up the middle than other places. I couldn’t say from that data if he hits it up the middle more often however.

I can also say he has grounded out to the right side of the infield far more times than anywhere else. 24 times in fact vs 15 total everywhere else.

So, if you discount XBH (which certainly aren’t ground balls), Fred has hit it to the left side 17 times (hits + outs), the middle 15 times (Only one out!), and to the right side 28 times(hits + outs).

So, certainly his BABIP on hits that go up the middle of the field is very high (.93) while his BABIP on hits that go to the right of 2nd base is very low (.14) and pretty okay to the left of 2b (.470).

He also certainly pulls the ball most often, opposite field second most, and up the middle last.

However, how does this compare to “average” distribution? No idea.

Eugeniooooooo!!!!

by FairweatherFan on Jul 2, 2008 3:00 PM PDT up 0 recs

xbh

Who says they can’t be grounders? Doubles, or even triples could easily be hard grounders down one of the lines.

Nice research, btw.

Less arm, more talk. Raisingcain is a GAMER.
Adopted Giant: Henry Sosa

by raisingcain on Jul 3, 2008 2:34 PM PDT up 0 recs

Yeah, the discrepancy is curious. I have no clue why it’s so different.

Neglectful father of David Quinowski

by marcello on Jul 2, 2008 10:42 AM PDT up 0 recs

Uhh

the author himself concludes:

It may be true that some players are able to consistently exceed their expected BABIP because of their speed; however, this study suggests otherwise.

He did not conclude no. He used 12 guys. He didn’t clarify what he means by “speed”.

Also, he only looked at exceeding expected BABIP. He didn’t try to look at players who might underperform their expected BABIP.

You can’t conclude definitely “no”, based on such a small sample.

ZIPS: Milledge: 466 HR, 485 2B, 2282 hits, 278-379-524

by rfloh on Jul 2, 2008 11:04 AM PDT up 0 recs

Well, you ignored the, ”...this study suggests otherwise” part.

Anyways, I linked that study in response to point 1 in the Fanpost just to show that there isn’t much of a discernible link between speed and increased BABIP.

He did not conclude no.

I concluded no, I didn’t say the author of that link concluded no.

He used 12 guys.

Over 49 seasons.

He didn’t clarify what he means by "speed".

What? He probably meant guys who are fast, but that’s just a shot in the dark. Anyways, he wrote this to clarify:

these players are among the league leaders in stolen bases and, anecdotally, are some of the fastest players in the entire game.

Neglectful father of David Quinowski

by marcello on Jul 2, 2008 11:21 AM PDT up 0 recs

Well ok, among leaders in SBs, and anecdotally fast

is a “clarification”. It’s a pretty damn poor clarification.

Yes, 12 guys, 49 seasons. Compared to the MLB total of how many seasons? How many full time starters are there in MLB in a single season? How many over the 5 season period that he used? Compared to that, 49 seasons is a RIDONKULOUSLY small sample.

I know you concluded no. Based on what? A study with a ridonkulously small sample in which the author himself doesn’t draw the conclusion that you do.

There isn’t much of a discernible link between speed and BABIP based on a study with a ridonkulously small sample.

ZIPS: Milledge: 466 HR, 485 2B, 2282 hits, 278-379-524

by rfloh on Jul 2, 2008 2:19 PM PDT up 0 recs

There isn’t much of a discernible link between speed and BABIP based on a study with a ridonkulously small sample.

or “of 12 players picked for whatever reason, no discernible link between speed and outperforming xBABIP was found. Thus, readers can choose to extrapolate those findings to the other thousands of players who have planed MLB in the last 50 years at their own risk.”

I just did a study a few seconds ago and concluded from 12 players (the Giants position players) that the average production from a Major League Offense is approximately 4.04 runs per game.

Eugeniooooooo!!!!

by FairweatherFan on Jul 2, 2008 2:29 PM PDT up 0 recs

Wow!

That was anecdotally fast!

Your 2011 SF Giants: the 2008 Augusta Greenjackets!

by Lyle on Jul 2, 2008 4:22 PM PDT up 0 recs

Comparing sample size...

to the total population isn’t really necessary. For example, if you’re looking at a large sample (say, 1000) with a small standard deviation, the sample size will be big enough to make a judgement about the entire population, whether the population consists of a million or 83,942,847,238,974,892,374,892,374. In taking statistics and finding p values, the size of the entire population is not taken into account as it does not influence the certainty of the findings of a study based on a sample.

The short: It doesn’t matter that 49 is small compared to the entire population, although I’m not certain whether you can draw a definitive conclusion based on these findings.

I only have a signature because I recognize everyone else by their sigs, not their usernames..

by lmaozedong on Jul 2, 2008 2:40 PM PDT up 0 recs

So what is the standard Dev in relation to this subject?

IMO, it’s really hard to say. There certainly are extreme outliers (Eugenio Velez Vs. Bengie Molina).

Eugeniooooooo!!!!

by FairweatherFan on Jul 2, 2008 2:42 PM PDT up 0 recs

Sure, comparing sample size

to total population isn’t really necessary.

But, in this case, he doesn’t freaking have a large sample. He has 49 seasons, not 1000, picked according to criteria that he doesn’t state clearly and specifically.

Furthermore, he just lumped everyone together. Since he’s using such a small sample, only 12 players, why not show the BABIP and expected BABIP’s for each of these players? Maybe some guys might show a trend of outperforming their expected BABIP’s over their careers. A guy like Ichiro Suzuki, for example: career BABIP of 357. career LD rate of 21.1%. “Expected” BABIP of 331.

ZIPS: Milledge: 466 HR, 485 2B, 2282 hits, 278-379-524

by rfloh on Jul 2, 2008 3:17 PM PDT up 0 recs

I love it

I just love it when guys can’t disprove something, so they try to discredit with things such as small sample size. Small sample size IS a problem. But just because a sample size is small doesn’t automatically mean its conclusion is wrong.

My guess is that when his career is over, Freddie will have an above-average BABIP. My guess is that it will still be 40 to 50 points lower than it has been so far this season. .366 just doesn’t seem likely to be sustainable. Tony Gwynn’s BABIP wasn’t that high, and he had speed AND bat control. Freddie has only the former.

by sharksrog on Jul 2, 2008 5:06 PM PDT up 0 recs

Tony Gwynn had speed?

Your 2008 Athletics: It's Nothing Personal.

by PaulThomas on Jul 2, 2008 5:42 PM PDT up 0 recs

He got it from Neifi Perez, like everybody else.

My Dave Righetti is better than your Dave Righetti.

by howtheyscored on Jul 2, 2008 7:02 PM PDT up 0 recs

I thought Mark Sweeney was the dealer.

Trent Kline: Decentish. Also, my website is called ChatterBalks Dot Com and on it I make jokes about things.

by groug on Jul 2, 2008 10:22 PM PDT up 0 recs

Yeah, and where do you think he got it from? Huh?

This goes higher up than you realize, man.

My Dave Righetti is better than your Dave Righetti.

by howtheyscored on Jul 3, 2008 1:20 AM PDT up 0 recs

You must not work in a research field.

There are no acceptable conclusions that can be drawn from that study (really a cursory glance.) That in and of itself doesn’t mean that the conclusions it comes to are wrong, but the study itself is worthless.

by Viliphied on Jul 2, 2008 7:16 PM PDT up 0 recs

As I said

As I said, such a study can’t prove that something is WRONG. No one said it proved anything right.

But it DID provide information from which a broader study might be formed.

by sharksrog on Jul 3, 2008 12:54 AM PDT up 0 recs

No it didn't.

by Viliphied on Jul 3, 2008 8:34 AM PDT up 0 recs

Come on now

Come on now. If a sample is made that is too small, it DOES provide information from which a broader study might be formed. All it would take would be adding to the sample.

Let’s look at it this way. You go to the lab and pee in the cup. They tell you your sample isn’t big enough. What do you do? Do you start all over again, or do you add to what you already have?

by sharksrog on Jul 4, 2008 1:23 AM PDT up 0 recs

I love it

I just love it when guys can’t prove anything, so they try to make ridonkulous claims based on even more ridonkulously small amounts of data.

Just because someone did a “study”, doesn’t mean that the conclusion of that “study” is correct.

ZIPS: Milledge: 466 HR, 485 2B, 2282 hits, 278-379-524

by rfloh on Jul 2, 2008 10:59 PM PDT up 0 recs

No, it does not

Not, it does not. Nor does it mean that the conclusion is incorrect.

by sharksrog on Jul 3, 2008 12:55 AM PDT up 0 recs

Right,

so we’re right back where we were before the study was conducted.

Ergo, the study was worthless.

QED.

by Viliphied on Jul 3, 2008 8:29 AM PDT up 0 recs

haha, snap.

Eugeniooooooo!!!!

by FairweatherFan on Jul 3, 2008 8:45 AM PDT up 0 recs

No one said differently

No one said we were much ahead of where we were before the sample was taken. We have some ideas, but they may or may not be correct.

I merely argued with your comment that the study was worthless. If it spurs further study, it wasn’t worthless.

Maybe we’re just talking semantics. But you are the one making the questionable statements. I am merely trying to refute them in an effort to identify the truth of the situation.

by sharksrog on Jul 4, 2008 1:26 AM PDT up 0 recs

Great point

Regression is indeed a distinct possibility for Freddie Lewis, whose BABIP the last time I looked was slightly above his BABIP in the minor leagues. I would guess there are players whose BABIP didn’t regress significantly in the majors from its high minor league level, but I would suspect they are very few and far between.

As for those who cite Freddie’s 21% line drive rate, do you really think that is likely to hold up? I would be very surprised if Freddie’s line drive rate was that high in the minors.

By the way, www.fangraphs.com shows Freddie’s line drive percentage at 18.0%, which would actually be about a percent below league average. He has consistently been a ground ball hitter in the majors though, which with his good speed should be a positive. His 6.3% infield hit percentage is four times his 1.5% infield hit percentage of last season.

Freddie is going to strike out about once every four at bats, and he isn’t going to hit a lot of home runs. That indicates he will likely have to post an above-average BABIP merely in order to keep his batting average above .250.

I would guess that Freddie might be a .310-.320 BABIP guy, maybe even slightly higher, but I doubt he truly is the .366 BABIP guy he has been thus far this season.

If Fast Freddie doesn’t cut down on his strikeouts considerably, I don’t think he will hit for a very good average. If Fred didn’t get a single hit the next 45 times he puts the ball into play, he would still have an average BABIP. And if he hit no home runs and assuming he somehow didn’t strike out even once during that time, his batting average would drop to just .241. If he continued to hit home runs and strike out at his normal pace, his average would fall off to just .232.

I’m not suggesting Freddie is only a .232 or even a .241 hitter (although that isn’t totally impossible). I just think that he is truly much closer to being a .250 hitter than he is to being the .285 hitter he has been thus far this season.

Freddie has a .290 career batting average in 449 at bats now. That would seem to be enough bats to indicate that batting close to .300 is a possibility for Fred. But then, hitting .241 or even .232 might be possible, as well. My guess is that Freddie is the .250-.260 hitter I have been expecting him to be all along.

I’m hoping he truly is the .290 guy he has been in his brief major league career. But I fear that won’t be the case. Not unless he significantly cuts down on his strikeout rate as Nate Schierholtz has done.

Because Freddie walks a lot more often than Nate and has far greater speed, he wouldn’t need to make as much of an improvement in that regard as Nate has made. But half the improvement Nate has made would likely entrench Freddie as the near-.300 hitter he has shown to be thus far in his career. Without such an improvement - or a significant increase in home runs - I don’t think he his likely to maintain his fairly lofty career batting average (which is actually HIGHER than he averaged in the minors).

by sharksrog on Jul 2, 2008 10:49 AM PDT 0 recs

I'm with you, but more optimistic

Freddie has shown that he can learn lessons at the major league level, on the bases, and in the field. If that applies to his other tools, then I think he could become not just a serviceable outfielder, but a premier leadoff guy.

¿Julio is tourist in San Francisco? Harper's Bizarre!

by hairball on Jul 2, 2008 11:24 AM PDT up 0 recs

I'm all for it!

I’m all for Freddie’s becoming a premier leadoff man, but I think the chances of his doing so are perhaps one in ten, and almost certainly no better than one in five. He just flat-out strikes out too much. How many premier leadoff men do you know who strike out once every four at bats?

by sharksrog on Jul 2, 2008 5:08 PM PDT up 0 recs

data, data, data, and not a drop to drink

I don’t know what the data source is for either Fangraphs or Firstinning, but I use the latter as a rule, so let’s take their data. Going back to 2006, in 1200 PA Lewis has an LD% of only 16% (due mainly to his 2006 in Fresno), over which period he has a .348 BABIP across all levels. That would make his 2008 BABIP/LD% anything but an outlier in terms of his recent career: if anything, he is slightly underperforming his history.

His history, of course, is most likely noise. Even 1200 PAs isn’t all that many in the grand scheme. Then his BABIP probably is on the high side—but still probably not tons. A 21% LD% gives you a .340 BABIP, and he has a .366. That makes him somewhat better than a .250 hitter. Moreover, if his history isn’t just noise.. well, I’m optimistic he can continue to deliver a wOBA in the .350s.

by wcw on Jul 2, 2008 12:51 PM PDT up 0 recs

I'm confused.

I don’t know that much about all this saberish stuff, but isn’t the fact that he is outperforming his career LD% a bad thing, if we’re assuming he will regress back to the mean for his minor league career? The fact that his 2008 BABIP/LD% has gone down is a product of his possibly flukey LD%, is it not?

I only have a signature because I recognize everyone else by their sigs, not their usernames..

by lmaozedong on Jul 2, 2008 2:44 PM PDT up 0 recs

In general one would think that if he is outperforming his career LD%, then he will likely regress.

However, I’m not sure how much of a “career” anything a guy with < 1000 ML PA’s has. Factor into that the fact that he is still considered a “young” player by the organization and could very realistically be benefiting from better hitting instruction at the ML level as opposed to his “career” in the minor leagues.

One of the things that, IMO, sabermetrics really doesn’t process well is that young players can and do get better. Pitchers harness their control and learn new pitches and batters learn pitch recognition and improve their swing/stance. (Look at all the saber predictions for Sanchez’s performance this season as an example of this. Both Marcel and ZiPS predicted him to have a 4.7+ ERA and < 4 wins at this point in the season. Marcel thought he would only have 54 k’s thus far. Something obviously clicked in Sanchez this year that his previous performance was unable to accurately predict.)

Just because a guy hasn’t been able to hit curve balls his whole MiL career (and as a result suffers with a low OPS) doesn’t mean that if at some point he learns how to read and hit a curve ball his OPS is condemned to regress to the previous value.

Players, especially younger players, do learn and get better. Fred Lewis and John Bowker are, IMO, two great examples of this.

Eugeniooooooo!!!!

by FairweatherFan on Jul 2, 2008 3:12 PM PDT up 0 recs

Err,

The concept of age, and age to level is a very important part of “sabermetrics”.

All the proprietary projections systems, ZiPS, PECOTA, use an ageing curve of some sort. Only MARCEL doesn’t. That is because Marcel is expressly NOT designed to this. All Marcel does is average out a player’s MLB record, with weights assigned to each season. Marcel is deliberately simplistic.

Yes, younger players do learn and get better. That is a concept that many sabr analysts and writers are well aware of. Why do you think so many Sabr writers are so high on Lastings Milledge, for example? Why else do you think that ZiPS and PECOTA are so high on Milledge? His young age is a big reason.

ZIPS: Milledge: 466 HR, 485 2B, 2282 hits, 278-379-524

by rfloh on Jul 2, 2008 3:24 PM PDT up 0 recs

I realize that age is a factor

But I am not aware of any SABR system that accounts for the possibility of a step function in a players abilities. How could they ?

Eugeniooooooo!!!!

by FairweatherFan on Jul 2, 2008 3:32 PM PDT up 0 recs

And because a statement is worthless without an example

Kevin Youkilis:

After hitting 16 and 13 HR in his two previous seasons (and OPS’ing .800 and .846) Kevin is on pace this year to hit 24 HR’s with an OPS of ~ .900. Meanwhile, after walking 77 and 91 times the past two seasons he is now on pace to walk only 54 times this year.

Is there a SABR explanation for this? No. His ZiPS projection for this year says 12 Hr, .832 OPS and 71 Walks.

Why the disparity? Because he has apparently changed his swing/approach at the plate to hit for more power this year. It has made him a better player, and has thrown a wrench into the predictive abilities of SABR.

These things happen. They are not the norm, but that is why SABR is normally right – not always right. If there is a good reason why someone may continue to defy their SABR projection (ie not regress to the supposed “mean”) then it is reasonable to conclude that they may.

In other words, regressing to the mean only applies when the “mean” is still valid. Step changes in a players abilities/approach/role can often invalidate the previous mean.

Eugeniooooooo!!!!

by FairweatherFan on Jul 2, 2008 3:43 PM PDT up 0 recs

but actually Youkillis’ season means nothing of the sort.
What is the SD for HR, BB, OPS over a season? What if he goes back to his .820ish OPS self next year? Did he “significantly change his approach back?”

Every year things like this happen – and would happen EVEN IF THERE WERE NO step functions in baseline talent.

That’s is not to say that such step functions don’t exist – or even that they are undetectable (they probably are) but 1/2 a season of Youkillis doesn’t mean squat.

FIRE BRIAN SABEAN

by zenbitz on Jul 2, 2008 3:52 PM PDT up 0 recs

91 then 71 then 54 walks/season is one hell of a standard deviation

http://www.fangraphs.com/blogs/index.php/the-greek-god-of-swinging-away/

Eugeniooooooo!!!!

by FairweatherFan on Jul 2, 2008 4:02 PM PDT up 0 recs

No one can disagree that young players improve in unpredictable ways

However, in the specific case of Lewis, the improvement in performance (ignoring the LD%/BABIP issue) is explained at least as well by strict platooning as an improvement in their core abilities. MiLB splits took down their 2007 numbers, so I don’t know if Lewis was platooned in Fresno, but I suspect not.

VAE PVTO DEVS FIO

by Bhaakon on Jul 2, 2008 5:21 PM PDT up 0 recs

Well, yes

projection systems can’t account for a change in approach, a new pitch. So what? Can a scouting approach account for that? Let’s say a scout watched Youkilis for the whole of 2007. Could he have predicted Youkilis’ breakout?

As for Youkilis, have you watched him this year? A lot? Do you have scouting evidence that he has changed his swing / approach?

Are you aware that Youkilis’ first half OPS over his career is 888, whereas his OPS in the 2nd half is 734? Youkilis looks like he has broken out every year. And then he slumps in the 2nd half. Every year. I’d wait till the end of the year before I loudly proclaim that the sabr systems have failed with regards to Youkilis.

Also, let’s say that Youkilis did change his swing and approach. Yes, no sabr projection system can project that. And no scout can project that either. Not unless they have access to his coaches or Youkilis himself. So what is your point?

ZIPS: Milledge: 466 HR, 485 2B, 2282 hits, 278-379-524

by rfloh on Jul 2, 2008 11:10 PM PDT up 0 recs

My point is that

If someone is outperforming their SABR projection, it is entirely possible that there is a credible reason why (Change in approach, etc) and that their previous “mean” is invalid.

Hence, regression is not 100% guaranteed.

I based my comments on Youk on the article I linked.

Eugeniooooooo!!!!

by FairweatherFan on Jul 3, 2008 8:47 AM PDT up 0 recs

Yeah, okay. I agree

No one claims that these projection systems are anything close to being perfect. No one claims that regression to the mean is 100% guaranteed.

ZIPS: Milledge: 466 HR, 485 2B, 2282 hits, 278-379-524

by rfloh on Jul 3, 2008 9:46 AM PDT up 0 recs

However

Many imply that regression is garaunteed.

“he is out performing his projection, he will regress” is basically what you hear about every player who has a breakout year.

Baseball is a funny sport in which, if you always predict failure, you will be right a huge majority of the time. Saying “it won’t last” or “he can’t keep this up” doesn’t take any real insight or understanding of the sport. It is just generally true – and as such a “safe” position to take.

Correctly identifying when a player is having a breakout season on the other hand usually requires some understanding of what is going on combined with a little inside knowledge of WHY that player may be experience more success than they have in the past.

Fred Lewis this year is, according to his own statements, focusing more on going opposite field with the bat. That, IMO, is going to result in more balls hit to the opposite field and up the middle than otherwise.

Again, IMO, this may result in his outperforming his xBABIP.

Eugeniooooooo!!!!

by FairweatherFan on Jul 3, 2008 9:52 AM PDT up 0 recs

Let's see

Let’s see how it turns out—both this year and over Freddie’s career.

by sharksrog on Jul 4, 2008 1:27 AM PDT up 0 recs

Jonathan Sanchez

Jonathan Sanchez has a lot in his minor league history to indicate he had the potential to become a dominant pitcher. Freddie Lewis has almost nothing in his minor league history to indicate he would approach becoming a top hitter.

Freddie’s best skill in the minors was drawing walks, which for a potential leadoff man isn’t to be sneezed at. But it also indicates that his OBP will be more impressive than his BA, SLG and OPS.

by sharksrog on Jul 2, 2008 5:12 PM PDT up 0 recs

apples, oranges..

..Lewis is not a dominant hitter. He is a good hitter - if he plays CF. He is a very, very pedestrian hitter in LF. The current discussion is not whether he is a top hitter or (what he is) an average-to-slightly-above-average hitter - it is whether he is an average-to-slightly-above-average hitter or Yet Another Craptastic Giants Hacker.

I argue, not. You?

by wcw on Jul 2, 2008 6:20 PM PDT up 0 recs

Hmm,

League average EQA for LF is 270 in 2008. Lewis’ EQA is 285. To date, Baseball Prospectus have him at 6 runs above (the average for the) position.

That hardly seems very very pedestrian. If he can keep up his current performance, he would be an above average hitter, in LF. In CF,

ZIPS: Milledge: 466 HR, 485 2B, 2282 hits, 278-379-524

by rfloh on Jul 2, 2008 11:15 PM PDT up 0 recs

leage avg eqa in LF is only 270?

Jeepers.

Me, I like wOBA. Lewis’s has been in the .350 range across ‘07 and ‘08. That puts him square in the middle of the pack among your normal LF performance, hence ‘very, very pedestrian’. Not bad - commonplace. In CF, .350 is good. Not leage-leading - last year’s leaders were in the .380 range—but quite good.

by wcw on Jul 3, 2008 2:07 PM PDT up 0 recs