I Believe In EYE, Not In CT% (part 1)

This one definitely needs its own thread, because this would be a tool for the carry toolbelt if true.

Dr. K, Taro, G-Money and umpteen cyber-grandmasters piled on to agree that high K's in the minors are, if not a guarantee of failure, then certainly a giant red flag.

Excellent saber discussion.

=== Dr K Says ===

Hi Doc,

It's been a few years since I first ran across the idea -- maybe at baseball prospectus -- but contact rate is a huge issue for minor leaguers. The pitch stalker archetype generally has a substantial increase in strikeout rate when going from the minors to the majors.

Just a few examples:

Adam Dunn: PA/K=3.8 in MLB, 5.5 in MiLB
Jim Thome: PA/K=4.1 in MLB, 6.5 in MiLB
Mark Reynolds: PA/K=3.0 in MLB, 4.3 in MiLB
Richie Sexson: PA/K=4.3 in MLB, 5.4 in MiLB
Rob Deer: PA/K=3.2 in MLB, 5.8 in MiLB
Pete Incaviglia: PA/K= 3.7 in MLB, 5.4 in MiLB

The one important exception I found was Ryan Howard.

Ryan Howard: PA/K=3.6 in MLB, 3.7 in MiLB

I don't know if you paid much attention to the cyber-snarks over Ryan Howard's prospect status, but most of the fights went:

Pro-Howard: "Wow, look at that power!"
Anti-Howard: "Wow, look at the moron that doesn't realize strikeout rates like that in the minors guarantee failure."

Ah, the good old days. This is why I think Halman is still likely doomed, but let's hope not.

=== Need to See the Studies On This One ===

Am not a real dr. like drs. Naka-san and Gaffney, but then neither was Young Frankenstein...

Thanks for the input guys.

Am definitely not YET going to begin writing off every prospect who strikes out a lot in the minors. :- ) The concept "anybody who fans 150 times per 550 AB's in the minors is a writeoff" definitely does not jibe with my memory of the almanacs.

But the input is interesting and am going to look around for the studies. Good stuff. Hopefully the real Dr's will counter the below strongly enough so that dr. D gets to install a few light bulbs....

=== Quibbles and Bits Dept. ===

1. Of course strikeouts will increase in the majors. They increase as you move from Kansas City to New York, for that matter. :- )

2. Mike Wilson's MLE reflects an AVG adjustment. In Tacoma he's hitting .292/.405/.605. His MLE drops his AVG from .292 to .262. Are we recommending a special MLE for high-K players because of a "critical mass" factor?

3. There is certainly a K rate in the majors beyond which point you are defying gravity. Mark Reynolds fans 220 times a year, but you don't want to bet on this kind of player. I do sympathize with the idea that a batter fanning 160 times per 550 AB's in AAA, is going to have too many holes for MLB.

4. Are we saying that a player who had a K rate of 150/year in 2009 cannot ever succeed? Or is it okay if he improves it? For example, if Greg Halman fans 170/550 AB's this year but then fans 110/550 next year, is he then a prospect? That's the kind of thing I'd like to examine the studies for.

5. I believe in EYE absolutely, though not (yet) in K% in isolation. (Y'know what I mean.) BaseballHQ alone has published so many convincing studies on EYE that we've got an axiom here, not a theory.

6. I wonder whether, if you went back and adjusted the studies to isolate the EYE principle, to what extent it was really the EYE law of gravity that the CT% studies were detecting.

............

In other words, I wonder if the guys who were "predicted" for failure based on high K rates in MiLB weren't actually a group of players with terrible EYEs -- and if the high-K players with high BB's did fine.

Kelly, Dr. N and Pirata will be very familiar with this experimental-design syndrome of looking for A, thinking you're getting it reflected directly, when actually it's related-variable B that you are seeing reflected indirectly.

Not saying I'm sure that's what's going on. Definitely suspect it, though.

Part 2

Printer-friendly version

Add new comment

Filed Under

Sports

Mike Wilson

Greg Halman

Replies

Mini-study

I too lazy to do a real study of my own, so I did a cursory investigation with google. Regarding prior studies, there is no perceived or presumed contract rate cut-off and I did not see any evidence for calibrating the strikeout rate with the walk rate. All that the studies appear to emphasize is that strikeout rate goes up, on average, as players advance from A-ball to MLB. If a player bucks the general trend, lower strikeout rate in AAA than A+ for instance, I think the logical conclusion would be to view that event as noteworthy and deserving of attention.

I did go back and inspect my previous outline of a study that you quote above with better formatting (thank you) with an eye toward 'eye'. Let me start by saying that the list above consists of MLB 'successes'; players that received sufficient at-bats in MLB to get sufficient at-bats to end up on the list of players with the most single season strikeouts. Note that most players are recent for two reasons: (1) strikeout rates have been higher for the last decade than ever before and (2) minor league strikeout records are not complete for older players, such as Mike Schmidt and Reggie Jackson.

I'll spare everyone a list of numbers, but an inspection of the players already presented in your post roughly preserve the eye ratio they had in the minors in the majors. Stated specifically, Dunn and Thome had excellent eye ratios in the minors, while Sexson, Reynolds, and Howard had acceptable eye ratios. As a general rule, the eye-ratio worsened with promotion to the majors, with the major league BB/K ratio about 75% of the minor league ratio. Sexson and Howard bucked this trend and sustained the same eye ratio in the bigs as in the minors.

In short, I don't see the importance of contact rate as a proxy for the importance of eye ratio. To say that eye ratio is more important than contact ratio is analogue to saying that OPS is more important than slugging percentage. Eye ratio reflects roughly two skills -- pitch recognition and ability to square up a ball -- so I would argue that eye ratio indirectly reflects the importance of contact rate by mixing it with walk rate.

If you want to use only one number, surely choose eye ratio, but don't expect it to possess as much useful information as independent contact rates and walk rates. For instance, Vladimir Guerrero and Adam Dunn had similar eye ratios and OPS in the minors, despite their extremely different approaches at the plate:

Guerrero Dunn

OPS 0.99 0.95

Eye 0.83 0.85

In instead you attempted to present a more thorough approach to the numbers:

Guerrero Dunn

BB% 8.0 15.5

K% 9.9 18.2

ISO 0.240 0.221

BA 0.345 0.304

These stats paint an entirely different picture of the two batters consistent with their approaches.

+++++++++++++++++++++++++++

To the more general question of how 'true' the contact rate trend is, I would argue that no statistical trends in baseball are sufficiently strong to warrant strong attachment, be it age arch, eye ratio, or BABIP. I would put more stock in an intelligent trained eye at Tacoma telling me whether Mike Wilson have figured something out, than I would on average trends in the minors. I suppose the value of statistics is that it should indicate to Jack Z and the player development crew that they should pay Mike some extra attention.

Blocked or tackled?

Great post and reply!

The thing that I see as impossible to "know" is whether a skill has been "tackled" - a REAL (and sustainable) change has happened -- or whether a player is just riding a particularly long 'streak', and will revert to career norms at any moment.

I've seen enough guys put up unheard of numbers for an ENTIRE season - and then go back to same-ole' same-ole' that it's hard for me to get overly excited about ANY prospect, (no matter how gaudy their "spiked" numbers read).

Clement ... Adam Jones ... Wlad ...

Guys cruising along with solid, (but unspectacular) numbers suddenly put on a 1000 OPS show for a bit, and they're suddenly viewed as "we HAVE to get him to the bigs immediately" prospects.

Wilson's eye has been 1:2 for 8 years. In 2010, it's 3:4. Major improvement in Eye. But, his actual walk rate hasn't really budged, (100 points of patience and a walk every other game is same as always - going back to 2003). So, he really just cut down on his Ks. The lack of change in walk rate calls into question the idea that he changed his "approach". (Maybe he did - maybe he didn't).

Production-wise, he jumped from a career 210 ISO to a 2010 290 ISO. In 2008, he showed a 270 ISO with ZERO change in his eye ratio. Basically, he's doing in AAA in 2010 exactly what he did in AA in 2008 ... except his eye ratio is better. Me? I think the power increase is "real". It was there in '08. I think the "eye ratio" for 2010 is illusory -- a good year coupled with a slightly weaker K-rate for the league (7.1 in SOUL in '08 - 6.9 this year in PCL).

Question: If you think the eye is illusory - and he's "really" still a 1:2 eye -- .5 BB/9 and 1-K/9 guy - do you still want him? (How many 300 ISO hitters do you have to choose from?)

Or, compare him to Matt Mangini. He of the .745 career OPS, who is posting a .914 in AAA. Same 1:3 eye ratio, (but lower K-rate than Wilson). But, this year, Mangini ups his ISO from a career 150 to 240.

Over time, I have reached the conclusion that whatever our individual "hot button" stats happen to be ... there is a tendency to perceive *MY* key stat as non-illusory. (Guilty as charged). It's easy to look at Mangini, see a high BABIP, and dismiss him as a 'true' emerging player. But, what does going from a 150 to 240 ISO mean while your LD% is dropping at the same time? For his career, he's had trouble with LHP. THIS year, his OPS split has gone from 130 to 20 -- and he's slugging 559 instead of 355 against LHP.

Learning the NEW SKILL of hitting lefties is *HUGE*, (if it is real). But, it's a small sample ... only 79 ABs. Heck, I got no idea whether this is real of illusory improvement. I'm more prone to get excited about a 24-year old spike than I am about a 27-year-old spiking. (And, of course, a 20-year old spiking would be even more thrilling).

But, even 20 year olds get "hot". Not every statistical change is "real" change. That's one reason you HAVE to have quality scouts and coaches telling you ... "same old player ... just seeing the ball really well right now." or "same old guy ... just in a bit of a slump ... his tools are still fine."

Mike Wilson, (to me), is the Mariner's Brooks Conrad. Nice tool for the toolbox, but a BIIIG reach if you're hoping for him to turn into a key starter in the future.

Of course, with Smoak and Branyan and Bradley available to DH, I think Wilson's hopes have been dashed. Saunders is CLEARLY the LF choice for "prospecting" at the moment, and Wilson hasn't played anywhere but OF ... so, it's LF or DH for him. (I still cringe at the idea that so many people think it perfectly reasonable to consider a player for a FULL TIME position at 1B who hasn't played an inning there before). (Full time as opposed to rare/occasional spot start to deal with minor injuries, etc.). Part of me wonders -- what if Joe Buffoon at first inadvertently takes out Felix with a blind side tackle -- will the idea that there are some REAL important aspects of playing 1B you want your MLB 1B to have before planning to play him there 5 days a week? But, I digress.

Not saying Mangini is an answer to anything ... but for me, he is certainly a question ... which becomes VERY important should Lopez get moved.