The False Dichotomy is one of the classic errors of logic. It bares its fangs relentlessly in politics, domestic life and sports. "Do you support our troops or are you opposed to the war?" is the Mac Daddy of all False Dichotomies because not only is the dichotomy false, the premise is also. But that's another discussion.
The quintessential False Dichotomy in baseball is this: Stats or scouts? Below is an example of how badly that dichotomy is begging to be put out of its misery.
Man on first, none out, bottom of the seventh inning of a tie game. Should you bunt? The answer, transparently, is that it depends. New baseball research tells us that, on average, trading an out for a base is a bad swap. Over the history of baseball, you can expect to score 28% fewer runs from a man-on-second, one-out scenario than from a man-on-first, no out situation.
That suggests that, generally, many teams bunt way too often. Recently, some franchises have recognized this and altered their strategic approach. The saber-savvy A's and Red Sox both bunt pretty rarely.
But few situations are average. If your pitcher is up, a bunt is probably a good idea. If a .300 OBP batter is at the plate and a .400 OBP guy waits on deck, a bunt might also make sense. If the two opposing hurlers are aces shutting down the offenses, you might be playing for one run. New research shows that on average, there's virtually no difference in the frequency of scoring a single run between the man-on-second, one-out versus man-on-first, no out circumstances. In that case, a seamhead would shrug his shoulders, but a manager could tip the balance one way or the other with his soft skills.
Imagine that the manager knows his on-deck hitter, a .400 OBP guy, is ailing. Or struggles against side-arming lefties, like the fellow now on the mound. Or is going through a rough patch at home that's distracting him. The skipper might decide in that case to swing away with the .300 OBP guy in that case, odds be damned.
Managers and their staffs bring massive amounts of useful information to every game, or at least we hope they do. Simply playing the odds would eliminate the manager's value-add. That's as backward as ignoring the statistical research. The two have to be synthesized for the best decision-making. Smart franchises use numerical analysis to create frameworks of decision-making, but leave specific situational decisions to the men in uniform.
Besides, if a team always plays the odds, it's totally predictable. Sometimes, a card player has to bluff just to keep his opponents off-balance. A good field staff understands that they are managing their players for the best results over 162 games, which isn't always the same as managing for the best results in any single game. Harkening to the discussion above, the manager might want to provide his players some practice bunting in real-game situations so that they're more proficient by September, even if doing so slightly reduces his odds of winning any one game in May.
The point is that scouting and stats aren't opposite sides of a coin; they're complimentary heuristic tools. Sometimes they point in different directions, but an unbiased user can weight the information they provide appropriately and come to a well-supported conclusion.
One last thing about decision-making: one outcome does not vindicate or repudiate a decision. Sometimes the bonehead decision turns out well, or the smart move backfires. Even when the odds favor our choice, there is a non-zero chance that an ill wind will blow. The reverse is true too as evinced by the blind squirrel and the broken clock.
As the saying goes, the battle doesn't always go to the strong, nor the race to the swift. But that's the way to bet."
b
No comments:
Post a Comment