Friday, April 10, 2009
Hit Rate ObserverPosted by John Burnson at 4:00am
Welcome to Hit Rate Observer! Each week, HRO will try to find unexplored avenues for success in fantasy. I thank Dave, David, and Derek for the opportunity.
A perennial quest in fantasy baseball is to boil players down to one number—some easily understood figure that permits quick comparisons between players. The idea has drawbacks, but it also can serve to focus the mind.
The modern sabermetric toolkit contains a wealth of candidates for our “one number”—OPS, Runs Created, BaseRuns, wOBA. Colin Wyers covered a lot of them in his column yesterday.
I’m thinking of something simpler, though. If you are in the market for one number, a case can be made for good old lineup position. Here you have a number (and single-digit at that!) that is strongly suggestive not just of a player’s overall offense but also of the split in his character between being a baserunner (OBP) and driving home other baserunners (SLG). For fantasy leaguers, lineup position also instantly insinuates the player’s contributions not only in Runs and RBI but also in HR, BA, and SB. Again, all on a 1-9 scale.
However, there’s a second aspect to lineup position of importance to fantasy players, which is that players who are higher in the order get more chances to show off their bats. Here is the average number of plate appearances per game, ordered by the position in the starting lineup (all data in this article are for 2008):
(We include data for both the AL and NL but excluding all pitchers. Also, we are considering only players who started in the lineup; consequently, the downward trend reflects not only the slighter PA from batting lower in the order but also the greater potential for lesser hitters to be pulled for a pinch-hitter later in the game. Both are hazards of the hinterlands.)
The graph is pretty linear from 1 and 7, and then it falls off more sharply. All in all, a starting hitter in the #1 slot averages 750 appearances in 162 games, whereas a starter in the #9 slot garners only 535 PA. Apart from their difference in make-up (which is, to be sure, large), a #1 hitter has a profound edge in plate trips.
There is a weakness in this analysis, however. In the above graph, we looked at opportunities for the majors overall. But teams don’t generate the same number of opportunities. If we truly want our “one number” to be useful, we’ll have to do better.
Consider these two AL teams in 2008:
Texas scored 210 more Runs than did Kansas City and carried 34 more points of OBP. It stands to reason, then, that the Rangers got more work from their starting nine—and they did, about 1.75 more PA per game. The starting #1 hitter on Texas logged 8% more appearances (57 more PA) than his counterpart on KC. For a typical #1 batter last year, that equates to 8 more Runs, 2.25 more SB, and 51 more at-bats of a .276 BA. (If you’re wondering why Texas got more PA from the starting #5 slot than from the starting #4, it’s because their #4 batter was more often replaced—20 more times, in fact.)
The upshot is that lineup position is relative. What we really want is effective lineup position (ELP)—a number that captures not only the batter’s position within the order but also his team’s position on the offensive spectrum.
We can derive Effective Lineup Position if, instead of using a batter’s lineup position to estimate his PA/G, we use his PA/G to estimate his lineup position. We can consult our first graph to build a suitable table:
Wrap your head around that: an effective lineup position of 0. The #0 spot represents a batter who, through a combination of breathless team play and personal indispensability, gets so many plate appearances that it’s as if he's batting ahead of the typical lead-off man!
Does such an animal exist? Sure. We’ve already mentioned Texas’s #1 hitter (in truth, he was halfway to a lineup spot of -1). The Rangers actually had competition in this department from the Mets, whose #1 hitter had an identical 4.79 PA/G. The lead-off men for Boston, Cleveland, and Chicago (NL) also batted like #0.
Here the top 10 hitters in PA/G last year, among batters with sizable PA:
Note that Dustin Pedroia ranks ahead of teammate Jacoby Ellsbury even though Pedroia tended to bat #2. The reason is that Pedroia never batted lower than #4, whereas Ellsbury, on those occasions when he didn’t bat #1, never batted above #6. Drafting Pedroia—a #2 hitter—gave you more chances than from most other teams’ lead-off men.
And the upheaval isn’t limited to the top of the order. Tampa Bay’s starting #9 hitter last year averaged 3.81 PA/G; with the same rate, he could have slipped unnoticed into Seattle’s #7 spot.
Why talk of ELP rather than PA/G? Two reasons: First, ELP turns an obscure rate (PA/G) into a familiar one—most people wouldn’t know that 4.51 PA/G and 4.39 PA/G were sharply different, much less where those numbers fell in the order. Second, lineup position introduces more color to the picture; even with a fictitious spot like #0, we know the sort of skills that the hitter brings to the park (good OBP, good speed, little power).
This year, Heater is predicting big things from the offenses of Cleveland, Colorado, both Chicago teams, and (of course) Texas. Pursuing the lead-off man (or even the #2 hitter) on one of those clubs could spell the difference in a tight fantasy contest.
Lineup position—both straight and effective—will appear in Heater Magazine this year. You can subscribe to Heater, as well as Dave Studeman's Batted Ball Report, at this link.
Compiled by THT Staff.