Derek Jeter Revisited

How is Jeter’s defense this year?

On Opening Day 2003, Derek Jeter suffered a shoulder separation in a collision with Toronto catcher Ken Huckaby. Jeter missed the next 36 games, as the shortstop position was divided between rookie Erick Almonte and veteran backup Enrique Wilson. On May 13, Jeter returned to the lineup against Anaheim, and has played in every game but one since then. Some analysts have noted that Jeter, again, has a relatively low total of chances handled compared to his backups:

Because these SS have played behind the same group of pitchers, and (in theory) should not be significantly affected by variations in opportunity, the speculation yet again is that Jeter?s defense is hurting the Yankees, especially since the Bronx Bombers have not performed particularly well defensively since Jeter?s return:

Without Jeter (through 12-May-2003): 26-11, 3.86 RA/game

With Jeter (since 13-May-2003): 14-17, 5.06 RA/game

So yet again, the question arises: Why are Jeter?s successful chances so low compared to his backups? And what effect (if any) is this having on the Yankee?s overall defensive picture?

MLB.com maintains play-by-play data files for each game on its Web site in the basic event format used by Retrosheet. I have been able to convert those files into the Retrosheet event file format and have all of the Yankee games in my database. While the MLB data isn?t as complete as I would like ? detailed ball location information is not available ? nonetheless there is enough information there for a deeper look at the Yankee shortstops.

Before looking at the PBP data, I want to note that the basic zone ratings suggests that the advantage in the raw stats is likely due to extra opportunities for the backups:

None of those ZRs are particularly good ? the average SS ZR is in the .850 range - but Jeter?s is the best of the bad. So if the other SS are making more plays, but handling a lower percentage of the balls that are close to them, that would suggest that they are also seeing more total opportunities than Jeter.

The biggest problem with zone rating, in my opinion, is that it narrows the field too much. I don?t think ZR tells the entire story, and therefore I went to the play-by-play files. In looking at the play-by-play data, since I didn?t have specific ball location data for every play, I counted the following balls as opportunities for the shortstops:

• Ground balls actually fielded by the shortstop
• Ground ball singles fielded by the left fielder
• All ground balls fielded by the center fielder

Ground ball extra base hits to left are usually down the line, so I excluded those. I went over my scoresheets for the past couple of years, and on balance the SS has more chances to get to ground balls to center than does the 2B and therefore I decided not to exclude any of those.

Looking at the numbers this way yields the following results for the Yankee SS:

Jeter did, in fact, have fewer opportunities to make plays than did his backups. He also made fewer of them, on a percentage basis. The difference is on the order of 6 plays at Almonte?s rate and 9 plays at Wilson?s rate ? somewhere around an extra hit every 3-4 games.

But Jeter?s overall lower rate of making plays is only a part of the story. The rest of the story is that the Yankee defense, as a whole, wasn?t significantly worse at making plays with Jeter in the lineup than with Jeter out of the lineup, even given Jeter?s markedly lower rate of converting balls into outs:

Whatever Jeter?s failure to make plays was costing the Yankees, they were making up with improved defense elsewhere. In the series of articles I wrote last fall and winter, I suggested that the Yankee defense was positioned in such a way as to reduce Jeter?s opportunities. It seems as though that is still the case; the Yankees don?t appear to have been hurt in terms of the team?s ability to convert balls into outs even if one accepts that Jeter?s backups are doing a better job of getting to balls in the SS area, and that almost has to be because the other fielders are positioned to get to more balls.

Although I certainly don?t think one can draw a definitive conclusion from half a season?s worth of data, and there are likely biases in the data for which I haven?t tried to account (quality of opposition, for one), I think the evidence so far suggests that Jeter?s defense has slipped a notch. It might be, as one Yankee fan on a discussion list that I frequent has suggested, that Jeter has become tentative in the field in an effort to prevent reinjury to the shoulder. Or it might really be that he?s starting to slide to the point to which his defensive performance will indeed match the reputation for terrible fielding that he has achieved among the statistical analytical community. Regardless of the reason, though, it looks as though the rest of the Yankee defense has adjusted for the moment to counter whatever problems Jeter is having.

So why are the Yankees allowing more RPG with Jeter at SS? Well, the table below might offer a clue:

“TTO” is shorthand for Three True Outcomes, which is used by statistical analysts to refer to walks, strikeouts, and home runs ?the events that the defense does not directly affect in any way. I also include HBP with walks ? thus the categories are K, WH, and HR.

Note that in the games prior to Jeter?s return, Yankee pitchers walked fewer hitters, struck out more hitters, and allowed fewer HRs. I would suggest that it is those pitcher-specific events that are at the root of the Yankee problems since Jeter came back.