Saturday, September 15, 2007

More on plate discipline

Earlier this season, Pizza Cutter published an interesting study (which I summarized here) that attempted to more carefully assess batter patience and discipline. Pizza's study looked at player at-bats and categorized each swing or non-swing into categories: correct decisions (swinging and making contact, taking a ball), incorrect decisions (swinging and missing, hitting a ball foul with less than two strikes, or taking a strike). His findings indicated, among other things, that Adam Dunn made a lot of mistakes, most specifically that he seemed to take too many pitches. He also found that someone like Scott Hatteberg tended to make very few mistakes overall.

The primary limitation of Pizza Cutter's study was that he was unable to assess where the pitches actually were in the strike zone. Players like Vladimir Guerrero had very high "correct decision" rates, even though scouting data indicates that he swings at a lot of bad balls...and makes contact. The result was some uncertainty about what these values really tell us about a player. For example, despite Adam Dunn's high "incorrect decision" rate, he is still a prolifically productive offensive ballplayer, and his extreme patience does result in a lot of walks, which is a big part of his offensive game. So perhaps his approach is still a good one?

Yesterday, Dan Fox published a study (subscription req'd) that in many ways fills in some of the gaps in Pizza Cutter's methodology using the PITCHf/x data from MLB Gameday. With these data, he can identify strikes and balls independent of hitters' decisions on how to respond to them. He can then assess hitters' success rates in responding to different sorts of pitches. He has yet to release a spreadsheet containing all of these data--I'm hopeful he will at some point, but those BPro folks tend to like to keep things under wraps--but here's what I was able to pull out about particular Reds hitters from his article.

Legend (approx mean values in parentheses):
Square (0.87) - % of pitches in strike zone that a hitter swung at and made contact with.
Fish (0.325) - % of pitches outside of strike zone that hitter swung at.
BadBall (0.73) - % of pitches out of strike zone that a hitter swung at and made contact with.
Eye (0.25) - % of pitches in strike zone taken for strikes (excludes 3-0 counts).
* Data values in italics were among the most extreme in Fox's dataset. Blue values are "better" than average, red values are "worse" than average.
* Order is as they appear in Fox's article. After the first six, I'm eyeballing the numbers from Fox's graphs.
NamePitchesSquareFishBadBallEye
Keppinger, J
426
0.982
0.216
0.860
0.361
Phillips, B
698
0.899
0.446
0.705
0.215
Hatteberg, S
519
0.911
0.229
0.866
0.245
Conine, J
298
0.888
0.229
0.872
0.280
Hamilton, J.
435
0.858
0.349
0.711
0.102
Hopper, N.
---

0.365

0.189
Encarnacion, E.
---

0.375

0.212
Ross, D.


0.305

0.302
Gonzalez, A.


0.392

0.305
Dunn, A.


0.290

0.202
Griffey, K.


0.315

0.256

Fox presented the following categories of hitters, based on his Fish and Eye stats. I've identified Reds that fall into each category:
  • Disciplined: Doesn't swing at "bad" pitches, swings at "good" pitches in the strike zone.
    • Scott Hatteberg - completely consistent with Pizza Cutter's data.
    • Adam Dunn - Needless to say, this is inconsistent with Pizza Cutter's data. His data would put Dunn into the next category. I can't really speculate on the cause, except that Fox's data are all '07 data, while PC's data were for '06, when Dunn clearly had an off year. Maybe we're seeing the change in his approach? I'd love to know what Dunn's Square rating was...because there's got to be some reason that he strikes out so much.
  • Patient: Doesn't swing at "bad" pitches, but takes too many "good" pitches for strikes.
    • Jeff Keppinger - Keppinger's fish rating was among the lowest in baseball, indicating that he almost never goes after a bad pitch. And his square rating was also among the highest in baseball, indicating that even though he may take too many pitches, if he swings he will nearly always make contact.
    • Jeff Conine - I probably shouldn't include this ex-Red, but I thought it interesting that he combined excellent ability to hit bad balls with a tendency to avoid swinging at them. In other words, he's intelligently selective with regards to his bad ball hitting.
    • David Ross - One of the things that really helped Ross's numbers last year, in addition to his fine power, was his ability to take a walk. This indicates that he has continued to try to be patient this season. Unfortunately, his walk rate is down. Perhaps taking all those strikes is part of why he's struggled? Another guy who I'd love to see Square data on.
    • Ken Griffey - Griffey's almost dead-on average for "eye," so he barely makes this category. In general, he seems to be a well-disciplined hitter.
  • Aggressive: Swings at bad pitches, but also swings at good pitches in the strike zone.
    • Brandon Phillips - Surprise, surprise. Phillips is nothing if not an aggressive hitter.
    • Norris Hopper - Again, surprise, surprise. As a confidence builder, it's good to see my own observations of hitter tendencies reflected in these data.
    • Josh Hamilton - This one did surprise me. Hamilton always seems to me to be pretty patient, and his walk totals generally support that. But he may be a bit of a victim of the dichotomous way I categorized folks here--in truth he's fairly average for most values...except Eye. Hamilton has the lowest eye score in Fox's dataset, indicating that he will swing at almost anything in the strike zone. Can't fault a guy for being aggressive on balls in the zone. Joe Sheehan's work certainly indicates that this will tend to result in higher BABIP than swinging at balls out of the zone, as does John Walsh's recent study.
    • Edwin Encarnacion - Eddie showed good on-base ability last season, though in truth his OBP might have been slightly "inflated" by a tendency to get hit by pitches. This year, and especially in August, he has been more aggressive. It shows in these data.
  • Pitchers' Friends: Swings at too many balls outside of zone, takes too many pitches in the zone.
    • Alex Gonzalez - Gonzalez has never been much of an on-base threat, instead deriving most of his offense (when there's any present at all) from his power, which is decent for a middle infielder. This might show part of the reason for that -- a tendency to chase bad balls, and a tendency to take a lot of strikes.
Ted Williams pic ripped from John Walsh's article, and I think was extracted from Williams' book The Science of Hitting..?
Dunn photo by AP/Tom Gannum, while the Phillips photo is by AP/John Bazemore.

5 comments:

  1. I don't have a BP subscription, but I'd be interested to see Utley's numbers. Mostly just to reaffirm my own general observation that the difference between batters with and without plate discipline is the difference between Utley and Phillips.

    ReplyDelete
  2. I would hesitate to agree with the statement that Keppinger takes too many pitches in the strike zone. If a batter has an extremely high 'square' rate and an extremely low 'fish' rate, it only makes sense that he would also have a high 'eye' rate. In other words, if in your mind you know that you are both likely to make good contact when you decide to swing, and unlikely to swing at a ball outside the strike zone, this will allow you to take alot of 'marginal' strikes early in the count. The various rates make complete sense in the case of Jeff Keppinger. It is a very interesting article.

    Thanks,
    Dave

    ReplyDelete
  3. Bradley,

    Utley's interesting in that he actually falls into the Alex Gonzalez category according to Fox's study. I can only get numbers from the graphs on him, but Utley has a 31% Eye rating (takes above average number of pitches in strike zone), while he has a slightly above average (essentially average) fish rate. But my comment to Dave below.

    Dave, I agree, it's probably too early in this analysis work to place value judgments on these tendencies. A pitch in the strike zone isn't necessarily a good pitch to hit. It's just that pitches in the strike zone are better pitches to hit, on average, than pitches out of the strike zone. The next step in this analysis would probably be to work in pitch type...a hard-breaking slider on the inside corner might be one that you should let go, while a fastball down the middle is one you should hammer.

    And then, of course, there are different sorts of batters, who have different hot zones...

    Nevertheless, I find this work fascinating. As many have said, the advent of this gameday work really does allow us to blend stat analysis and scouting to a degree that we've never seen before. It won't be long until we can create charts like that in the Ted Williams picture for every hitter in the big leagues based on hard data with large sample sizes. Exciting stuff!
    -j

    ReplyDelete
  4. I was looking at the Ted Williams' hitting chart and it struck me that it is quite possible that there are areas outside of the strike zone that would result in a higher batting average than certain pitches that are inside the strike zone. That seems odd. E.G. If you look at the lower outside corner of the strike zone those averages all tend to be .275 or lower. But if you look at the belt-high inside pitches, those pitches all tend to be well over .300. So, I would assume if a ball was pitched just a little further inside - which would make it outside of the strike zone - that the average would be at .300 when hitting that certain pitch. In which case a batter would be wiser to swing at a potential called ball than a potential called strike. I hope this made sense. So while, technically, you may go fishing outside the strike zone, it is actually a better pitch to swing at than those low and away.

    Thanks,
    Dave

    ReplyDelete
  5. Hi Dave,

    Makes perfect sense, and I bet there are cases when this is true.

    In general, though, it does seem to be the case that it's better to swing at pitches in the strike zone than pitches out of the strike zone.

    Here's a recent article by John Walsh on that issue.

    And Joe "NotFromBPro" Sheehan broke down BABIP in different regions in and out of the strike zone for righties and lefties here.

    Doesn't mean, of course, that there won't be exceptions. Different hitters have different swings, which means that they will vary in how successful they are at hitting different pitches. Vlad Guerrero, for example, does just fine swinging at all kinds of pitches out of the strike zone. :) -j

    ReplyDelete