Game Theory 101: The Chaos Gangsta

Gary Cohen made an interesting point last night, and it got me thinking.  For all the things Jerry Manuel is, one thing he is not is predictable.  This was noted by Gary in a positive light.  In a sense, this point is very relevant.  Game decisions in almost any competitive situation should have a degree of unpredictability, of apparent chaos to them.  This is why a human can beat a computer at chess.  A computer is always going to make the best computable move, whatever move seems to yield the best mathematical outcome. A person who also understands the outcomes, in this case, has an automatic advantage, because he can also accurately predict what moves the computer will make.  The probabilities to the human are no longer probabilities, they are causalities, and can thus be countered at every turn, even if its by a move that may not seem mathematically ideal.  When your opponent is a human, this level of presumption goes out the window.  You cannot predict a person's decisions with this degree of accuracy, and a person can even go so far as to try and consciously deceive you.  Thus, chaos and deception become a fundamental and dynamic element of game theory. But there's a problem with trying to project this justification for unpredictability on Jerry.  Deception is only relevant when your opponent's reaction is important.  When it comes to decisions that have little to no effect on your opponent's decisions, it doesn't particularly matter how obvious or subtle you are in your reasoning.  It just matters that your reasoning is correct and yields the best chance of getting yourself some WPA points.

The reason I make this point is because what Jerry did last night was in direct contrast to what some of us railed him for in this game.  If you'll remember, in this game, in the 5th inning, Tim Redding was left in with a two run lead, two runners on, and two outs.  He surrendered a run scoring double to backup catcher George Kottaras.  Then, with Nick Green due up, Redding was relieved by Sean Green with the tying and go ahead runs in scoring position.  Green gave up the base hit to Green, and the Sox took the lead, which they would maintain and build on in the following innings.

Last night, you had a similar situation, albeit an inning later.  Livan Hernandez had runners on first and second and two outs after yielding a single to second string catcher Jason Jaramillo.  This time, in contrast to the Redding decision, Livan was not given a chance to close the inning.  And again in contrast, instead of using a specialist like Green to try and record the final out of the inning, Jerry went straight to Bobby Parnell to pitch to Jack Wilson.  In this case, Parnell preserved the lead, at least for the time.  With three of the first four batters due up in the 7th inning being left handed, Pedro Feliciano was brought into the proper situation, removing Parnell from the game and continuing to protect the two run lead.  But the problem didn't come until the eighth inning, when a predictably shaky J.J. Putz didn't have the saftey net of the team's two most effective relievers not named Frankie.  Now Jerry has a major problem.  He has a reliever who's faced five batters without recording an out and while allowing three runs (one inherited) to score.  He no longer has the lead, so on the road, he obviously can't use Frankie.  So he has to rely on Brian Stokes, Ken Takahashi, and Sean Green to try and keep the game close and give his bats a chance to plate some runs.

The moral of this story is this:  Jerry is not unpredictable for the sake of deception.  He is unpredictable because his reasoning is flawed.  These were virtually identical situations.  They required no attempt to deceive the opposing manager, yet he made a completely different set of decisions.  Its quite conceivable that the shift in stratagem between these two games was completely related, and that Jerry was being reactionary to the poor ultimate outcome of the Redding game in Boston.  Maybe he didn't even realize it.  Maybe it was just in his "gut". Or maybe there was some other line of reasoning involved here that we just can't know.  Maybe he just doesn't like the way "5th inning" and "Bobby Parnell" feel together.  For all I know, he flipped a coin.  Either way, at best, he was wrong once and right once, and neither time did he land on a successful outcome.  I wonder what will happen next time a Mets Starting Pitcher struggles in the middle innings of a close game....