THE BOOK--Playing The Percentages In Baseball

Another brilliantly written and edited article with some brilliant insight from the scouts:

The proverbial five tools for position players—hitting for average, hitting for power, defense, arm and speed—are covered throughout the survey, in one way or another.

Only one player really scored high in all of the above: Ichiro.

He has averaged 9.6 HR/year in MLB.

My bad…

...and he could hit 50 home runs if he wanted to...

And I could be a Pulitzer-winning writer, if only I wanted to.

This survey makes me question the whole “scouting vs analysis” thing. Maybe we’re being too kind to the scouts, to want to give them equal footing. Maybe they’re only worthwhile at the lower levels of play. Once you get to AA, maybe the stats, properly adjusted, is the only way to go.

Honestly, I think folks like Tango and Theo are far too kind to scouts, in terms of seeing the ideal analysis as a blend of both.  Sabermetrics never got a free pass with me… if someone told me DIPS ERA predicted next year’s ERA better than this year’s ERA, I damn well wanted to see that proved before I believed it. 

If someone tells me that some college kid’s “inverted W” mechanics are a bad sign for his long-term arm health or strikeout rate… why would I just accept that and use it as part of my analysis?  Because some guy with a radar gun who’s been scouting for years says so?

Quantitative analysis has, to me at least, proved its value.  Qualitative analysis hasn’t… and of course, the irony is that I want to see some numbers on qualitative accuracy before I let them join the conversation with those methods that have already “proved themselves” to me.  One of the more fascinating aspects of living in this current era is that all sorts of professions, ideas, and dogma(s?), are being held accountable.  They’re being questioned, as opposed to blindly accepted.  Sabermetrics and “freakonomics” are just really fun applications of this.

Ditto Mike (whoever that is) and DS.

This is something I’ve never really understood.  Why is it so hard to believe that Ichiro actually might possess the ability to turn on a fastball and pull it over the wall with some regularity, but has figured out that he’s more valuable as a .340/.380/.430 guy than as a .280/.320/.450 guy?

50 homers a year is obviously hyperbole, but I’ve never seen anything remotely like a convincing argument that says he doesn’t have the skillset to pop 25-30 a year if he chose to utilize those skills differently.  Now, those extra home runs would almost certainly come with a significant cost due to the change in approach, but what quantitative evidence is there that his performances reflect a lack of ability rather than a cognitive choice?

Until you figure out how to quantify motive and intention, you’re going to be missing a fairly significant piece of the puzzle.  You might not want to throw scouting out the window until you’ve figured that one out.

I’m sure lots of players “could” hit more HR’s if they changed their approach.  What is the point?  It is ridiculous to say that so-and-so really has power because he COULD hit more HR’s.  And there is a heck of a lot of difference between 25-30 and 50 HR’s.  If the scout meant 25-30 why not say 25-30 or even 40?  Saying 50 just gives someone no credibility.  Not to mention the fact that NO 5’11”, 170 lb. players hit 40 HR’s.  It is not possible.  If you want to bring up a guy who COULD hit 40 HR, bring up some big strong guy (I’m sure Ichiro is strong for his size).  But to call Ichiro a “power hitter in disguise” or to say that he could even hit 35 HR per year (as true talent) is just ridiculous.  I don’t know that he “can’t”, but as I said, to just hypothesize that if someone changed their entire approach, they could hit a lot more HR, is just ridiculous speculation.  The scout that said he could hit 50 HR if he wanted to is just parroting what he has heard other scouts say.  If he knew nothing about Ichiro, never heard of the guy, and saw some 5’10”, 170 lb. speedster with an unconventional swing, he would have to be an idiot to say, “Yeah that guy could hit 50 HR if he wanted to.” It makes no sense.  I would venture an educated guess that it is physically impossible for someone his size and body type to hit 50 HR (without an enormous random fluctuation).

So we have two points of view which could not be more antithetical.  One is that it is virtually impossible for a 5’10” 170 speed demon to hit 50 HR a year and the other is that Ichiro could be a 50 HR guy if he wanted to.  You couldn’t have more polar opposite points of view.  Which is more likely an accurate statement to any reasonable, intelligent person who knows just a little about baseball, let alone a professional scout who has been in the business almost all of his life?

I’m sorry, there is no middle ground here.  To say that Ichiro COULD hit 50 (or even 40 HR), if jhe “wanted” to, whether you want to call it hyperbole or not, is just preposterous on any level.

And my comments and thoughts on the matter have NOTHING whatsoever to do with “throwing scouts out the window.”

The bottom line is that scouts plus stats = the best information.

The other bottom line is that sabermetricians are 100% into finding the optimal blend of information to do whatever it is you want to do.  It just so happens that they don’t have the skill set to come up with some of that information.

Scouts on the other hand, while possessing lots of useful skills and providing lots of useful information, are X% full of crap.  That does NOT mean that they are not valuable to an organization.  Of course they are.  It just means that they are X% (put your own number in for X) full of crap.

I make mistakes now and then of course, and some of the things I believe will eventually be proven wrong, but I challenge anyone to come with something (specific) that I am full of crap on.  We can take almost any random scout and there will be 100 things that they are full of crap on!

How’s that for (not) being PC?

I dunno.  How about Joe Morgan?  27 HR in 472 AB in 1976 (34 per 600), and in a 5-year span he averaged at a rate of 27 HR per 600 AB.  In today’s HR dollars, that’s gotta be around 35.  So, it can be done.

***

There are alot of “thin” players, like Luis Gonzalez, who had high totals, but of course they are not short.  The “wingspan” might be the more important thing. (But Morgan I seem to remember had a fairly compact swing.)

Junior was a pretty thin guy in Seattle, right. Eric Davis too.  Wilie Mays.  Heck, Hank Aaron was not a big guy in his younger days.

Bill Hall got 35 (though you can call that a fluke at this point if you like).

***

In terms of true talent, since I almost never forecast someone with over 40 HR, clearly, the person claiming that someone could hit 50 HR is not thinking of it in terms of true talent, but just sampling.

According to baseball reference:

Willie Mays 5’11 180
Hank Aaron 6’0 180
Mel Ott 5’9 170

But even if he could decide to hit 25-30 homers, I don’t see the point of giving Ichiro credit for power that he has not shown, and more importantly has no intention to show.  I think Ichiro has responded to the question something like “Yeah, I could, but I’d only hit .250” So if you want to credit theoretical Ichiro with that power, you can’t also rate him as the best batting average hitter in the game.

Not that I get a lot of detailed scouting data to look over, but here’s my feeling on the matter.

There are quite a few things scouts are very good at, and at the moment we don’t have a lot of good statistical proxies for. A lot of it has to do with inputs and processes. You need that sort of information if you’re trying to figure out why the results you’re seeing are occuring.

But they’re very poor at aggregating that information and coming up with a player’s overall value. It’s a matter of putting a relative price onto their observations - some things carry more weight than others, and their weights are all wrong.

Stepping even further into my own wildly unsubstantiated hunches - I think scouts tend to overrate “bad ballplayer” skills - things that you absolutely have to do right to stick around in baseball if you lack the physical gifts to compensate for those deficiencies. And why not? You get to be a scout by being a “bad ballplayer” - not having the raw, physical gifts to succeed. And so you spent most of your career learning the “fundamentals” and applying them perfectly; that’s the set of knowlege that you bring to the table.

But they’re very poor at aggregating that information and coming up with a player’s overall value. It’s a matter of putting a relative price onto their observations - some things carry more weight than others, and their weights are all wrong.

This is the exact philosophy behind the Fans’ Scouting Report.  An observer won’t necessarily give you the right weight, be it for fielding, hitting, or pitching.  He knows what he knows.  That’s his value.

The merging of the data, to model reality, that’s what the rest of us can do.

If a scout says that Ichiro’s batting skills and approach is such that he can be either a great power hitter or a great average hitter (but not both), that’s what he is.

Ichiro himself, through many years of experience, will determine the optimal approach to hitting.

The same can be said of Vlad.  A scout can say that he has a tremendous eye at the plate.  Clearly, there is just about no one in baseball that is as successful as Vlad on pitches outside the strikezone.  That if Pujols adopted Vlad’s approach, Pujols might be a worse hitter than Vlad.  The same scout will say that Vlad needs to be very aggressive.  Again, if Vlad approached the game like Pujols, he might be a very much worse hitter than Pujols, even if he ends up with 100 walks.

Whether it’s Vlad himself that knows his optimal approach to hitting, or it’s guys like us (see Clay Davenport’s excellent article in BP08, the second half of that article) that can try to model such a thing, who knows.  I’ll lean toward the player as to knowing mostly what he’s doing, especially if it’s a great player.

But say Juan Uribe or Pedro Feliz?  Maybe those guys’ personal scouting is much worse than what we can offer.

The main point is to separate components from the overall model, and being good at component-analysis doesn’t make you an expert in modeling (and vice-versa).

Vlad might be good at both.  And Feliz might only be good at one.

Hi MGL - I’ve been reading here for a long time, and probably have commented on this blog more than anyone but you and tango (admittedly, my rate of insightful comments may not be as high as dave, david, guy, jinaz, sky, chone, pizza, rally, studes, etc., but I try) But I’ve been around. Nice to meet you

Luis Gonzalez is NOT a skinny dude, he has biceps the size of my head.  Not that biceps are useful in hitting home runs (I don’t think), but just sayin’, skinny face != skinny dude http://mlb.mlb.com/mlb/photo/photogallery/ws_top20_large/20.jpg

I’d bet Ichiro could hit 25 HR per year if he went with a Griffey-esque uppercut.  Baseball-reference has him at 5’9”, 160, which I find ludicrous, as that would probably make him about the 15th shortest dude in baseball.  And personally, if I was his coach and saw him hitting all these home runs in BP, I’d be pretty intent on getting him to try it in games.  Regarding the wingspan theory, Griffey is listed as 6’3”, 205, which makes me think he was maybe 190 or so as a rookie?  Not a hulking late-90’s ballplayer, but not a small dude by any means.

There was that guy who posited that steroids don’t help you build leg muscles, but that it’s lower body muscles that drive HRs.  If that’s the case, Griffey ca. 1992 might provide some backing to his latter claim: http://seattletimes.nwsource.com/ABPub/2007/06/19/2003753725.jpg
Baby got back.

Mike, you’ve got a ways to go until you catch up to Guy and Rally:
http://www.insidethebook.com/ee/index.php/site/comment_leaders/

***

Joe Morgan seems to still hold up as the extreme example that Ichiro could follow.

Jimmy Wynn is another good example.  Lots of HR for someone from the 60s.  I didn’t realize how small Miguel Tejada was, but there you go, another small guy with big power.

Small guys with power are BULKY!  There is NO way that someone with Ichiro’s body type CAN ever (again, other than a fluctuation) be a 40 HR guy.  I don’t mind when someone exaggerates a bit.  But to say Ichiro and 50 HR’s per season in the same sentence is just plain wrong.  And again, I think that the only reason Ichiro is brought up when people speak of players who COULD hit a lot more HR’s is because they have heard that before.  I see nothing about him, other than he is a great athlete, and he occasionally pulls a ball a long way - something that anyone can do, BTW, that suggests that he could hit even 30 HR a year (I suppose anyone but a Pierre or Eckstein type could hit 25 if they changed their approach).  Why would anyone say that Ichiro could hit 50 HR a year unless they just like to hear themselves talk and want someone to quote them?  And why would a professional scout say that when he knows that he is going to be quoted in the newspaper, if it is not true?  How about he gives his name to the newspaper guy if he is so confident with his statement, so that the team he works for can see how smart he is?

Anyway, bottom line, as Rally said, and I tried to say, who cares what someone CAN do, not that we know whether it’s true or not, and even if it is, maybe it is true for almost ANY player?  The only thing we can and should use to characterize a player is what he does (if it is a young player and we are speculating on his future, that is another story of course).  Can we say that Dunn hits for power and average - I mean he COULD hit for average if he tried, since he has good strike zone judgment and is not impatient?

And we certainly can’t say that Ichiro has 5 tools because he hits for average AND power (well, he COULD hit for power), when HE himself admits that if he hit for power, he’d probably hit .250!

As I said, X (a pretty big number)% of what comes out of many, maybe most, scout’s minds and mouths is garbage.  Again, that does NOT mean that the other 1-X% is not real valuable.

Same is true for players, managers, and especially true for commentators and sports writers.

The other day, I was listening to Rollins on XM radio.  He appears to be one of the more eloquent and intelligent players around, at least the former.  The interviewer asked him who the best all-around player in baseball is.  I expected him to say A-Rod, Pujols, Utley, Jeter, or Sizemore, maybe even Tulo, or even Ortiz if he wasn’t going to include defense in his mind. He said Ichiro and Polanco.  Nice to see Polanco get some deserving love, but huh?

----"Clearly, there is just about no one in baseball that is as successful as Vlad on pitches out of the strike zone”
_____________________________
True, but so what. The avg for that is about a 400 OPS, I believe. Vlad’s typical OPS on bad balls is about 590, and that doesn’t include the impact when he swings and misses on bad balls. (Actually, if you swing at a bad pitch, you’re probably better off missing, unless there are 2 strikes).

A 590 OPS is below repl level. Since Vlad puts lots of bad pitches into play, in order to achieve his excellent overall OPS, he must be absolutely destroying the ball on pitches in the strike zone.

B James points out that Vlad is a super extreme outlier on the % of pitches thrown to him which are out of the zone. They simply don’t throw him strikes unless they have to.

So from that, it would seem that Vlad is not maximizing his ability to hit the ball with authority, because he is not ‘forcing’ the pitcher to throw him good pitches to hit. But...James makes the point (I’m paraphrasing/expanding) that his readiness to swing enables him to bypass the hesitation another hitter might have regarding whether to swing or take, and that this advantage compensates for all of the bad pitch swinging.

This may very well be true in Vlad’s case, but it is not inevitable, at all, that batters who are disciplined on pitches out of the strike zone, also must be passive on pitches within the strike zone (IOW, take lots of strikes). Examples of this, from memory, from 2007, are Ortiz, Ordonez, Chipper, and others. These guys take the balls, and swing hard at the strikes.

So, even if a batter has aggressive tendencies, it is entirely possible for him to learn to distinguish balls from strikes, and to let-er-rip only on strikes.

If Vlad had that skill, he might be the best hitter ever.

A hitter gets smarter as he gets older.  His physical skills also deteriorate as he gets older.  Vlad will eventually adapt to being a more passive hitter.

If he follows the standard aging curve for a power hitter, then I’d say that how he played in his 20s is his optimal self.

If on the other hand he ages more gracefully for a power hitter, it may be that he was suboptimal in his 20s.

But, you have to believe that a guy that is already this good can’t probably get better, only worse, if he changes his hitting approach.  Same deal with Soriano.

I suspect that most mature players at the major league level have an approach that is near optimal for their skill set. If you have a player like Vlad who has incredible contact ability but (I assume) poor pitch recognition skills then the optimal approach is to take your hacks. Being patient would reduce his overall effectiveness as he would be not swinging at balls he could paste and striking out more by going deep into the count.
Cust, the bizarro world Vlad, is just the opposite with poor contact ability yet good pitch recognition. He can be maddening to watch as he will strike out looking again and again, but he knows not to swing unless the pitch is where he wants it and what he expected.
I think Vlad is the superior talent of the two, but Cust seems to be able to hold his own in spite of his extreme approach. Vlad more than holds his own with his extremeness.

Scouting based evaluations of specific tools are not valuing anything.  There seems to be something of a disconnect here because, in essence, the only thing MGL cares about a thing is what it’s value is.  It’s basic pragmatism, and while there’s nothing wrong with that, it’s also pointless to call something worthless or wrong that is not trying to be pragmatic. 

Sometimes, an observation is just that.  Ichiro has the skills necessary to hit 25 or 30 home runs a year - that’s an observation, not a valuation.  You might think that observation is totally useless, and that’s fine, but a lot of people find it extremely interesting that one player possesses such a wide variety of skills. 

The observation is inherently interesting to some people.  That it doesn’t fit into a preordained box of approved usefulness based on it’s pragmatic value doesn’t disqualify it from being true.

David/20:
Yes, observation means understanding that Ichiro COULD do something different. And I also realize the value of a hyperbole. However there are a few issues going on:
1) You can’t really call Ichiro a 5-tool player. Even if you assume that by hitting with a lot less contact, he could hit many more homeruns, he’s still a 4-tool player. Either contact or power. He’s not, say, Albert Pujols, who can hit .330 with 40 HR’s any season. So regardless of the fact that you may believe that Ichiro COULD hit 30 HR’s, you have to observe that he can’t do so without sacrificing other (more important, in his case) skills. You can’t call him both a power and a contact guy, because he’s either one or the other (assuming he could fully be a “power guy").
2) Ichiro ranked highly in every department, including power. This means that, even if just from an observatory point of view, scouts think he’s one of the potential elite sluggers in the Majors. Aside from the fact that I personally believe this to be silly, this is done because they think that if he “altered his approach” he could improve his HR’s. Well sure, but why is this being done to him only? If Adam Dunn choked on the bat and became more aggressive he could probably hit .300 but he wouldn’t walk or hit HR’s and would become less useful, as #15 said. Dustin Pedroia has a big uppercut swing, so maybe we can say he’s got power as a tool, likewise. And I’m sure he might hit 20 out of the park if he settled for hitting .240, but who would care?

Moreover, if other sluggers themselves became even more extreme in their “power” approach (and accepted hitting .200 or so), they would still have an edge over Ichiro. All in all, do scouts really think that Ichiro has more power (even if just potentially so) in his bat than Howard, Ortiz, A-Rod, Pujols, M. Cabrera, C. Pena, Vlad, Fielder, Wright, Hanley? I just mentioned 10 random “power” names.

The point I’m trying to make is that I’m sure Ichiro could hit many more HR’s, but he would do so while sacrificing more important skills, as you also noted. Moreover, unless scouts are trying to show “something” (such as the fact that they see what mortals don’t see), it is pointless to make such an observation about Ichiro but not about every other Major League player alike, as if he’s the only one sacrificing one tool to optimize his overall skills, while this is probably true for most players.

From http://www.signonsandiego.com/sports/padres/20080330-9999-1bb3depodest.html

DePodesta said his appreciation for scouting wasn’t conveyed by the book. He often praises the scout-driven Braves, who won 14 consecutive playoff berths starting in 1991.

“They deserve a best-selling book,” he said. Part of his success in Oakland, he said, owed to heeding scouting reports such as those by A’s scouts who recommended Kevin Youkilis, memorably dubbed by DePodesta as the “Greek God of Walks.” Youkilis became a poster player for “Moneyball.” The A’s were unable to trade for him, and he went on to become a Red Sox star who helped Boston win the 2007 World Series.

“Our scouts liked him a lot,” said DePodesta, who prefers to scout amateur players without knowing their statistics.

The article also reveals (to me, anyway) that DePo had a recurring non-speaking part on the great show Homicide: Life on the Streets, a Tango favorite. He would have been on in Season 3 or 4, I think.

Pretty cool.

Theo Epstein’s sibling, I think sister, was also a writer on the show, around the 4th season I think.

Via imdb, it looks like Anya Epstein was Fontana’s asst. season 3, a staff writer season 4, story editor season 5, producer season 6, and consulting producer in the final season, and wrote about 20 teleplays.