Thursday, August 12, 2004

Jack Wilson: An Obsessive's Perspective

Honest Wagner, his readers and I have been having an interesting discussion regarding my recent post about Jack Wilson. They’ve advanced some intriguing ideas about why Jack Wilson might continue to have a high batting average in the future, such as that his speed allows him to reach base more often on hits than most players. His quickness/speed, along with a recent increase in power that renders him finally able to hit the ball out of the infield, are the main factors in Wilson’s recent success, or so goes the HWag crew hypothesis.

One problem with the discussions we had was that they were too general to come to a solid conclusion; not enough effort was made on either side (and understandably so, since what I’m about to present took a long time to research) to parse out the various causes of Wilson’s success this year. I’d like to do that here.

First, the numbers:
2003: .256/.303/.353/.656. AB: 558. HR: 9. XBH: 33. BB: 36
2004: .319/.341/.490/.831. AB: 457. HR: 9. XBH: 49. BB: 14

I don’t think anyone doubts that Wilson’s power spike represents real improvement; his numbers are way up in extra base hits, and it’s clear from watching him that he’s hitting the ball with a lot more authority than he used to. He’ll likely retain most or all of that power going forward.

What I find weird – and sorry to those of you who’ve read this from me before – is the big improvement in batting average, and the drop in walk rate. An improved walk rate might indicate a better approach at the plate, which would explain the increase in batting average, but Wilson’s walk rate has actually gotten worse.

There are players who put up extremely high batting averages and extremely tiny walk rates. But most of them tend to be slap hitters with lots of speed, like Alex Sanchez. Wilson can’t be called a slap hitter anymore. And while HWags and others at his site claim he’s a quick player, I don’t know why that quickness wouldn’t have translated to a high – or even average – BA in the past. And he’s never stolen many bases.

I’m skeptical. But let’s see if they’re right. Let’s look at some more numbers:
2003: K: 74. GB: 207. FB: 179. GB/FB: 1.16
2004: K: 48. GB: 176. FB: 142. GB/FB: 1.24

These GB/FB ratios, which come from ESPN, don’t seem to include line drives since, according to MLB.com, Wilson has had 139 flyball outs this year (compared to 142 flyballs).

Wilson has hit groundballs and flyballs in fairly similar proportions to his number of at-bats both years. So it doesn’t appear that he’s tried to slap the ball more this year to get more infield hits. And since speed isn’t a skill that suddenly appears in players who didn’t have it before, it seems unlikely that his increase in batting average has much to do with his speed.

More numbers:
2001 groundball outs: 117. Total Ground Balls/Ground Ball Outs: 1.21
2002 groundball outs: 167. GB/GBO: 1.18
2003 groundball outs: 164. GB/GBO: 1.26
2004 groundball outs: 126. GB/GBO: 1.40

These numbers appear to me to be significant. Many more of Wilson’s groundballs are becoming hits this year than in the previous three. Perhaps Wilson deserves some credit for this – perhaps he’s improved at hitting seeing-eye singles. But this seems unlikely, since it would be strange for a guy who can’t even control the strike zone well enough to draw a walk to suddenly excel at hitting the ball between fielders. Perhaps he’s hitting his grounders harder than last year. Or perhaps his speed has improved, although again, I think that’s unlikely.

I think it is more likely that this sudden improvement in GB/GBO is simply luck. In any case, if Wilson had the same GB/GBO ratio in 2004 that he had in 2003, he would have about 14 fewer hits.

Let’s subtract these 14 hits from his 2004 totals and calculate his batting average again:

AB: 457. H: 146-14=132. BA: .289

Wow. That’s a big difference. I don’t really know if GB/GBO represents a skill or not; I strongly suspect that it is not. I calculated the GB/GBO’s of several players and could not find a meaningful relationship between GB/GBO’s by the same player in different years. Speedy players do tend to have better ones, but other than that I didn’t find any other trend that suggests a good GB/GBO is a skill. I even checked some players who had big power spikes to see if their GB/GBO improved after the spike. It did in some cases (Javy Lopez ’02-’03, for example) and didn’t in others (Alfonso Soriano ’01-’02). I haven’t done a scientific study, but I don’t think the improvement in Wilson’s GB/GBO reflects much actual improvement in Wilson. It could be the subject of a good study, though; if anyone is aware of any research on this subject, please let me know.

Still, a .289 batting average for Wilson would still be a big improvement over the Wilson of old. What explains the rest of the improvement in batting average?

2001 AB/K: 5.57
2002 AB/K: 7.12
2003 AB/K: 7.54
2004 AB/K: 9.62

As Wilson has matured, he has struck out a lot less. Offensive strikeouts don’t tell us anything about how good a player has been, but they may be able to tell us something about why a player was good (or not). Wilson’s putting the ball in play more than he used to, and therefore giving himself more chances to get hits. There may be some element of luck here, but since players seem to strike out or not at fairly stable rates, it’s likely that Wilson has simply improved in this area.

Now let’s take that .289 average (the one we got from adjusting for GB/GBO) and adjust it down by pretending that Wilson struck out this year at the same rate he did last year.

AB: 457. K: 61 (AB/K: about 7.54)

If he’d struck out at his 2003 rate, he would have put 14 fewer balls in play. He probably would have gotten hits on about five of those (Last year, Wilson’s BA on balls in play was .295; this year it’s .358). So:

AB: 457. H: 132-5=127. BA:.278

So Wilson has improved his batting average about 11 points by striking out less; I think these gains will stick.

When a player homers, his opponents have no chance of getting him out. Wilson has been hitting more homers last year than he did in the past.

2003 AB/HR: 62
2004 AB/HR: 51

Now let’s adjust 2004 for his 2003 homer rate.

AB: 457 HR: 7.37

So if Wilson had continued to hit homers at his 2003 rate, he would have hit seven or eight rather than nine. Let’s say he got one extra hit this year that he wouldn’t have gotten last year, since there’s some chance of one of his homers dropping in for a double in 2003.

AB: 457. H: 127-1=126. BA: 276

So after adjusting for these explanations, Wilson’s 2004 BA would have been .276. That’s still 20 points higher than his 2003 BA, but it’s much closer than we were before.

How might we explain these other twenty points? I’m not sure, but I bet they’re mostly the result of Wilson hitting the ball a whole lot harder and getting more line drives. So the only part of Wilson’s 2004 batting average I don’t think is the result of real improvement is his GB/GBO. After we adjust for that, we still have a batting average of .289. That, I think, is very likely to be close to Wilson’s ‘true level’ going forward.

So let’s create an OPS for Wilson using the new batting average along with his current walk rate and isolated power, which I think more or less represent his ‘real’ level of ability:

.289/.309/.457

There’s a lot of potential error here, especially since there are twenty points of batting average I don’t really know how to explain. And batting average is notoriously hard to predict from year to year anyway. But I think the above numbers are the best I can do. (Please note, also, that after all that analysis, those numbers are pretty close to HWags' offhand .290/.310/.430 prediction a couple days ago.)

Wilson’s walk rate will likely continue to be atrocious, but the average should still be quite good and his power should be good too. Any further dropoff in his batting average could be offset as his walk rate returns to his usual levels (which are still pathetic, but better than they’ve been this year). And it’s possible he could continue to get better. I can’t believe I’m saying this, but I think most of his improvement this year is going to stick. 2004 will probably turn out to be his career year, but he should be a productive player from now on.

I must also add that it would be a good idea to supplement this projection with knowledge of the career paths of players similar to Wilson. An up-to-the-minute PECOTA projection, for example, would be helpful. Perhaps players who did what Wilson is doing in 2004 tended to flame out the next year as pitchers started to exploit their tendencies to swing at junk, for example. But given all the information to which I have access, the prognosis looks... better than I thought, at least.

I advance this not as and end to debate/discussion, but as a beginning.

5 Comments:

Blogger Rowdy said...

Looks good, Charlie. We'll see what happens over the next few years.

A lot of the projection business is alchemy and snake-oil business. Nate Silver (or was it Sheehan?) has said that himself. It's not that they have tremendous confidence in PECOTA over there, for example, so much as they know that exuding tremendous confidence sells more annual memberships.

Not all rookie seasons are similar experiences. When Jack broke in with the Pirates, they gave him far more PT than any other team in the league would have given him. He didn't force his way into the big leagues the way David Wright did. Read the latest stuff on Jose Castillo and the way his head is swimming this year. He doesn't even speak English.

I just don't think Jack Wilson's stats from before this year mean as much as this year's stats. It's a general rule of thumb, when making projections, that last year counts twice what the year before counts, etc. With young guys and guys who came up the way Jack came up, I think the latest stats are the most telling.

My guess is he finishes strong, that sinks in, and everyone forgets he used to suck. The same thing will happen with Adrian Beltre.

Anyway we'll see what happens with Jack and figure out who was right in two years, OK? Until then there's not much else we can do.

1:05 AM  
Blogger rpND said...

Nice analysis, and nice blog in general. You and I are similar thinkers.

Since you seem to have done a lot of legwork on this topic, I was wondering if you had the numbers regarding Jack's average-on-balls-in-play this year compared to his previous seasons. I've noticed that Jack's pitchers per plate appearance this year is the lowest of his career (not by a ton, like 3.56 vs. an average of 3.66 - I think), and his walk rate is also down as you mentioned. Jack is clearly being more aggressive at the plate this year, putting more balls in play and getting more hits. I would be interested to see his avg/bip, since if it is unusually high it would suggest a big luck factor to this season, and if it is not it might suggest solid improvement.

12:14 PM  
Blogger chill74 said...

I can understand the numbers that you have analyzed but the fact that if you say putting the ball in play more is luck, I can't buy it. You make your own destiny, Jack now has confidence in himself. He understands that he doesn't have to have the perfectly placed Gwynn hit, but he can hit it and make the other team make the play. There are many factors to reaching a base safely and and the odds are higher if you push the issue. An example, the run he scored against the Giants, if he stayed at third, we may have lost the game, but he said make a good throw, catch and tag first....well A.J. didn't get the tag and you know the story. The same for the 10 triples, he could have 5 more doubles now, but he has the confidence to make a play. He has turned over a new level of confidence. He can look at his past years of failure and use that to his advantage. Good luck to Jack.

12:59 PM  
Blogger Charlie said...

Chill74: I was trying to say that it's very likely that the part of his improvement that comes from putting more balls in play (i.e., striking out less) is NOT luck. I think that part of his improvement will stick.

rpND: Thanks. According to my calculations, Wilson's 2003 BABIP was .295; this year, before I made the calculations a day or two ago, it was .358. He IS striking out less and putting the ball in play more, but a lot of his improvement has to do with BABIP. I think he'll hold some of those gains too, though, because he's hitting the ball harder. Still, his BABIP will probably drop a bit going forward because his GB/GBO will regress to the mean. Boy was that last sentence dorky.

Rowdy: I agree that this year's stats matter more, but then again my final guess does look a lot more like his '04 stats than any of his previous three. Yes, let's see what happens.

2:16 PM  
Blogger rpND said...

It's an interesting season for Jack, when you think of how it began. After he won his arbitration case, I remember him responding to criticism by saying he did need to work on his plate discipline, draw more walks, and get on base more to be more useful to this team. Ironically, he accomplished his goals by taking the exact opposite approach, and showing even less patience.

3:05 PM  

Post a Comment

<< Home

FREE hit counter and Internet traffic statistics from freestats.com
Listed on Blogwise Weblog Commenting and Trackback by HaloScan.com