21 July 2012

What the Orioles 10-2 Record in Extra Innings Means

It has been suggested by sum that a team proves itself in extra inning games.  That scenario is one where every plate appearances carries a great deal of importance and that this importance is in some way similar to how a team should perform in a tight race to win a division or wildcard and how they will fare in the playoffs.  This idea provides an engaging narrative that humanizes the sport and gives creedance to the idea that the 2012 Baltimore Orioles have a great chance to make the playoffs.

As of the morning of July 21st, the Orioles are 10-2 in extra inning games.  They win 83% of these and win them against good and bad teams alike.  To declare that this means that the Orioles are a tough team to face and are heading in a playoff direction ignores what we have discussed in a previous post: as fWAR is concerned, the Orioles are a bad baseball team.  Two potential measures: one heavily contextual and the other is stripped from context.  Do you believe baseball is largely linear (events arise from proceeding events) or is it mainly nonlinear (events generally occur by chance)?

To explore the validity of extra inning winning percentage on how successful a team can be, I decided to take into consideration all American League teams from 2009 to 2011.  I compared their overall record in non-extra inning games to their overall record in extra inning games.  The data set did not appear robust to me, so I decided to merely do this as an eyeball test.

Non ExInn

>.500 <.500
ExInn >.500 14 8

<.500 9 11
What we see here is that team that are above average in non-extra inning games have a 61% chance of being above average in extra inning games.  For teams with losing records in non-extra inning games, they have a 42% chance of being above average.  In general, this makes sense.  You might expect a good team in the first 9 innings to be a good team after 9 innings have been played.

Another way to look at the data to check how paired data points relate to each other.  That is, pairing a team's extra-inning record to the non-extra inning record.  To provide a better context, it might be useful to also compare a team's Pythagorean record (based on team run differential) with their non-extra inning record.

This is a somewhat sobering.  With an R-squared of 0.00002, it is fair to say that within this data set that extra inning record does not relate to non-extra inning record.  It also suggests that the Pythagorean method does an incredible job in projecting non-extra inning game record.  The Orioles have a -47 run differential (they have given up 47 more runs pitching than their offense has produced).  This is the third worst run differential among the American League's 14 teams.  Although the Orioles stand at 49-44, the Pythagorean method suggests they should be at 42-51.  The +7 "luck" is the greatest in the American League.

Dealing for a player like Omar Quintanilla may not appear to be an incredibly interesting move, but it is one that makes sense for a team that appears to be outperforming their talent level.  Quintanilla is a solid defender who has excelled in the minors, but has never been able to figure out Major League pitching.  He is a player who is quite useful for a mid-range or bottom tier team to kick the tires on.  He is not useful to a contending team, like the Mets, who may not have the leeway to figure out whether a 30 year old player can surprise them.


Ptcello said...

Any reason for interest in Rick Ankeil?
His strike to the plate(google it) earlier this season was nothing less than amazing.

Jon Shepherd said...

Hits righties well, ok range, great arm.