Stark Responds about Andruw

Well, not to me but to Scott Boras on David O’Brien’s AJC blog. Boras’s initial comments were posted by DOB on Tuesday.

But was he exactly the same player over the last few years that we perceived him to be? No. And Scott can manipulate his own numbers and “indexes” all he wants. But he can’t explain away those 100 balls a year that Andruw used to catch that he wasn’t catching anymore – until, by some remarkable stroke of fate, he got himself back in A-1 shape this year in a contract year (and now is magically catching them again). Do the math. If the guy was down 100 putouts a season, that’s four balls a week he used to catch that he wasn’t catching anymore.

I said in the book that I was surprised to see those numbers myself. But I didn’t make them up or manipulate them. They’re real. And Scott’s trashing of Zone Rating is purely his way of discrediting research he doesn’t agree with.

I only looked at Zone Rating because my initial inclination, as I wrote in the book, was NOT to believe the raw numbers. I wanted to factor out variables like whether the Braves’ staff had more ground-ball pitchers than it used to, etc. The defensive stat that does that best, in my opinion, is Zone Rating.

I’ve asked plenty of sabermatricians about Zone Rating. And they sure characterize it differently than Scott does. It doesn’t assign wider zones to players like Andruw because he’s so good. All centerfielders are assigned the same zone. So how does it penalize players with more range?

Andruw’s Zone Rating dipped in exactly the way his other numbers dipped. He used to lead the league. Last year, he finished at the bottom of the league. Any attempt to explain that away is an attempt to make the conclusion differ from the facts – which was the opposite of the way I went about it.

First, let’s talk about those 100 balls a year that Andruw is no longer getting to. Well, that is a bit of an exaggeration. From Andruw’s “The Guy” period from 1998–2002 (as defined by Stark in his book), the difference from 1998–2002 to 2003–2006 is 62 raw putouts, not 100. From his best year to his worst year the difference is 128, but it’s a little unfair to compare peaks and valleys, especially because some of this is a function of playing time.

Andruw averaged 449 put-outs per 162 games played from 1998–2002. From 2003–2006, he averaged 397 put-outs per 162 games played. Comparing his best years to his worst years in raw put-outs is deceiving. After you control for games played, the difference is down to 52 putouts a year. If you normalize it by innings played in CF, and assume he played as many innings as he did in his best year (1999) the difference is down to 43. In fact if you include his 2007 campaign, Andruw is on a pace to put out 494 batters in 1447 innings (his 1999 playing time), which is about equal to his career high. The difference from “The Guy” to the present averages is only 26 put-outs.

Year    Games   Innings PO     PO     PO
           		(Raw)  (162G) (1447 Inn)
1998    159    	1373    413    421    435
1999    162    	1447    493    493    493
2000    161    	1430    439    442    444
2001    161    	1435    461    464    465
2002    154    	1357    404    425    431
2003    155    	1329    390    408    425
2004    154    	1347    389    409    418
2005    159    	1366    365    372    387
2006    153    	1317    378    400    415
2007    63	557    	190    489    494
Mean    1998-2002       442 	449    454
Mean    2003-2006       381	397    411
Mean    2003-2007        	416    428

Now, this is not an insubstantial difference, but it is much smaller that what Stark claims it to be. I don’t think Stark is trying to manipulate the numbers, I just think he’s not looking at the big picture.

Now, let’s move on to Zone Rating, which is a rate statistic. The problem is that the traditional Zone Rating reported by ESPN is deeply flawed. Players are at a disadvantage if they are making plays outside of their zone, because an out-of-zone play is treated as a part of the player’s zone for that play. The out-of-zone put-out goes into both the numerator and denominator of ZR. If a player is making shifts from a standard zone, this can actually hurt a player, and he is not awarded extra credit for making an out-of-zone play (see here for a full explanation). I don’t know of any stat-head who considers zone rating useful, nor would I care if anyone did. Zone Rating is flawed in its design for punishing players who make plays outside of their zone.

While Stark responds specifically to Boras, he has yet to respond to what the superior defensive metrics indicate about Andruw: he is still one of the best in the game. If you want to find a metric that treats all fielders the same, why not go with the Plus/Minus system created by the man to founded Stats, Inc. and invented Zone Rating? It’s a system that employs several video scouts to plot the speed, trajectory, and location of all balls hit to all fielders over several seasons. If you are worried about the influence of the pitching staff, this is the metric that you need to use.

I would like to add that I like Jayson Stark, and I think he is a good writer. I just disagree with him on this. It’s not like he’s an idiot for using ZR. In fact, I applaud Stark for objectively evaluating Andruw. If ZR didn’t measure Jones as declining, Jones wouldn’t have made the list. He has no reason to bash Andruw just for the sake of doing so. It’s not widely known that ZR has serious problems and that a superior defensive metric exists. And it just so happens that it makes a big difference in Andruw’s case.

8 Responses “Stark Responds about Andruw”

  1. rob says:

    jc — nice retort to stark. i’ve heard him do several radio shows and was surprised with his explanations as it seems he only did *some* homework, not all the homework required.

  2. tangotiger says:

    I agree with JC that the problems with ZR is STATS’s implementation (extra counting of plays outside the zone in the denominator).

    MGL’s UZR has Andruw as roughly average from 2003-2007, though he rated fantastic in his career prior to that.

  3. Chris says:

    I agree that the zone rating may not be the most reliable stat to say someone is overrated, but this isn’t the only reason that Jones could be considered for this and not the only thing showing him declining. His homeruns and rbis can be viewed as impressive, but the amount his strikes out is also detrimental to his team and he’s becoming more and more like an Adam Dunn type player (I mean come on 65 strikeouts in 66 games this season is a joke). Now you can say well he’s hitting homeruns when he tries to swing for the fences, but that apparently isn’t working very well because he’s got a .215 average this season and hasn’t hit near .300 since his 2000 campaign. This also happened to be the last year he stole 20 bases. Now if none of these seems like downfalls to his game you can prove me wrong or tell give me an explanation, but there is no doubt in my mind that he is not the same player that he used to be.

  4. kc says:

    Great work JC. I know watching Andruw day-to-day that Stark’s claim is not true.

    Chris, the argument here is not on Andruw’s offense.

  5. Brent says:

    Here’s why you shouldn’t worry about Andruw. First of all, strikeouts are no more detrimental than other outs, for the most part. Would you rather him ground into a double play than strike out? Let’s not get too worked up over something that doesn’t hurt the team that much more than any other out.

    Second, Andruw’s BABIP is only .250 this year, down from .270 last year. He’s had a little bad luck in that the hits just haven’t dropped. Batting Average on Balls In Play has a tendency to work itself out, so odds are his batting average will rise as a result.

    Third, it’s easy to get worked up over his 2000 season due to that pretty .303 average, but he’s not that different a ballplayer these days. In 2000, he had 78 extra base hits. The last two seasons he’s had 78 and 70, respectively. So a few doubles are turning into homers. No biggie. But where have the hits gone? They’re turning into walks, as Andruw’s plate discipline, believe it or not, has gotten much better. This year he’s walking in 12.8% of his plate appearances, up from 12.3% last year and 9.5% the previous year. So while it’s fun to have a nice batting average, what we need from Andruw is a solid on-base percentage and a good slugging percentage. You can hit .250 and help your team if you don’t rely on singles to create your value. Once Andruw’s BABIP gets back to normal, his OBP should be up around .350 again.

    Fourth – As for his base stealing capabilities, that’s just a natural thing for a guy with Andruw’s size and skill set. Andruw got big a little earlier than most, but plenty of players stop stealing bases once they are veterans. Sammy Sosa and Chipper Jones, for example. It just happens when you get bigger and stronger. Also, I don’t think it hurt us that much. Andruw was never really a top-notch base stealer for us.

    So really, I don’t think they’re downfalls at all. Most important of all, I don’t think two months is a big enough sample size for us to start forgetting all the great strides he’s made over the last two seasons.

  6. Jones still seems to be a good center fielder to me. Most published stats put him around average over the last few years.

    The thing that makes Jones overated is that he was one of the best center fielders ever (certainly the best I have ever seen), and then he had that 50 HR season, so most people add together that player (the younger version who caught everything and the older delux power hitter who hit 92 HRs over the last two seasons) and think he’s a superstar. But he’s a good centerfielder and a low average slugger who gets respect walks. He’s not a super-star.

    Did MGL ever add fielding priority to UZL?

  7. Gibson Rules says:

    Does anyone find it a little troubling that Stark needed a sabermatrician to figure out zone rating when its flaws are readily understandable with 5-grade math?

  8. Richard says:

    Well, the controversy must be getting to him — Andruw, I mean, not Jayson Stark. He’s under the Mendoza line now, with an obp of .299 and slugging .381. He may well be in a serious decline phase.

    On the defensive side, I looked up the Braves’ DERs for the past 10 years. Starting in 1998: .700, .700, .695, .692, .702, .712, .699, .688, .693, .686, .693 (this year). Four of the five best ones are in Andruw’s first five years. I don’t know enough to isolate each position, but those numbers do not weigh in Andruw’s favor. I wouldn’t want my team signing him to a significant contract.