M's 3, Tigers 2 - Batting Orders
..................
=== Previously on L.A. Law ===
In the late 1980's, Bill James demonstrated that batting orders don't matter. In a Strat-O-Matic sense, anyway. Later, others followed this. The pioneering idea was never altered much.
If the Texas Rangers were to move Nelson Cruz to leadoff and drop Ian Kinsler to #6, they'd lose 20 R in the leadoff slot and lose 20 RBI in the #6 slot. But! Guess what else ... you guessed it. The Rangers would gain 20 RBI in the leadoff slot and gain 20 runs scored in the #6 slot.
James, in one of his inventive studies, created a "worst possible" lineup including Josh Hamilton* hitting 9th and stuff like that, and simulated 1000's of seasons, finding that the difference between an optimal lineup and the worst one was about 5%. He concluded, "If the difference between the worst and best lineups is 5%, what do you think the difference between two reasonable lineups is? That's right. It's nothing."
Sabertistas have refined and super-refined this paradigm, finding microns worth of difference, putting the best hitter #2 vs #3 and so forth. Dr. D has never been interested. Butterflies like that get lost in the maestrom of real-world chaos.
.
=== Evil Spock Alternate Reality Dept. ===
There is a completely different paradigm available, one that annoys pure sabermetricians because they can't capture its complexity with 8-variable formulas. That is the paradigm that has been used by John McGraw, Casey Stengel, Walter Alston, Earl Weaver, Tony Larussa, Lou Piniella and Eric Wedge. This is the paradigm that says that batters sometimes approach their at-bats differently depending on where they are in the lineup. As one local sergeant put it, if we 'net rats don't think batting orders matter "they don't know what the hey you're talking about. Tell them I said that."
As we've said many times, sabermetricians are lost at sea, as to what causes a player to have an UP or DWN year. Neither are most of us interested in attacking the problem, because we have as much chance of getting on base as Brendan Ryan does tomorrow against Justin Verlander. Next subject, then, lads ...
James himself is, typically, more frank about his limitations than most, and much less hostile to the suggestion that real things occur in baseball that he will not be able to position himself as teacher on:
Hey Bill - You've said that "we do not have near-perfect measurements of baseball players. To assume that we do is foolish." Couldn't agree more. How about the issue of "Why An Active Player Had an UP or DWN Season." Beyond the ILLUSION of an up year - e.g. park effects, lucky balls in play that we can measure - any thoughts on what might ACTUALLY cause a hitter to perform above or below his norms in a given season?
Do you consider that attitude / state-of-mind (divorce, death in family, getting sober, big teammate hates him, new coach, contract, etc) is one of the many variables driving a player's UP or DWN year? If so, can you *conceive* of any way to measure or *predict* state-of-mind, or is this simply a good place for the sabermetrician's humility to enter in?
Asked by: jemanji
Answered: 4/11/2012
There is no bad place for humility except in the bedroom. What seems obvious is that there are tens of thousands of things that can cause up/down fluctuations in performance; therefore, these fluctuations are not ulitmately predictable at an 80-90% level EVER.
On the other hand, we can predict fluctuations in performance up and down with SOME accuracy--more than 51%--NOW, and obviously more knowledge of more factors could be used to improve the projections. Knowledge of how hard the player has been working, how stable his life is, that sort of thing. .. .it's potentially useful, but limited also by respect for the player's privacy.
A tantalizing thought, that one about predicting fluctuations with more than 51% accuracy. The Red Sox apparently deployed this predictive ability in signing David Ortiz to his last, saber-snuffed, contract.
Back on track, those "tens of thousands of things" could obviously include a player's spot in the lineup. Most players and ex-players believe that this in fact the case. Let's stipulate that something off-the-field could affect performance, such as a player having a horrible fight with his wife. If something off-the-field can affect performance -- and Soviet sports psychology took this as first principles -- then why wouldn't something ON the field, something that occurred 5 minutes ago, or which is occurring now, affect performance?
Chone Figgins spoke hopefully of his problemos being due to his spot in the lineup. There's a big fat helpin' of nothin'. I mean, if you put ME at ninth or first, the order's not going to matter. But how about for a good hitter who is taking the wrong approach?
.
.