Another game in Cleveland, another battle against poor officiating

The Oakland Athletics have a long history of being involved in games decided by controversial officiating. From Derek Jeter’s "flip play", to Angel Hernandez’s call on what should've been a game-tying Adam Rosales home run on July 8th, 2013. One way or another, the A’s always seem to find ways to put the game in the umpire's hands.

Last Friday, it happened again. This time, however, there were no fireworks. Nobody was tossed from the contest and social media wasn’t aflame with less than flattering remarks about Angel Hernandez. This time it went completely unnoticed. Why, you ask? Because this time it was perceived to be just "part of the game".

For the past 5-10 years, people of all walks of life have debated just how much umpires impact the game of baseball. This is, in fact, a silly question because their impact on the game is everything. All you have to do is look at a hitter’s average in different counts to see that the outcome of an at bat fluctuates dramatically depending on what the count is. The real question becomes this: of the calls umpires get incorrect, how much do they matter in the grand scheme of things?

Before we delve into this question, we must consider the following: general managers calculate a certain amount of uncontrollable variables or "luck" when putting together rosters. That more or less means GM’s are constantly trying to minimize the margin for error. That’s how Oakland Athletics GM Billy Beane has made a name for himself. He exploits any little bit of information that could potentially minimize the risk he takes when evaluating players.

When Billy Beane looks at a hit, he doesn’t just see a player who hit the ball well enough to get himself to 1st base without the opponent having committed an error. He looks at each and every hit differently. Some hits are harder than others to different parts of the field. Some hits barely miss an outfielder’s glove, while others would almost undoubtedly be a hit given most circumstances. Moreover, a double off the wall is more valuable than a double that squeaks by a corner infielder. This can be summed up as follows: not all hits are created equally.

If hits are examined and dissected to such an extent, why don’t we examine balls and strikes in the same manner? Well, we do. If you look up a player’s batting average on a 1-2 count versus a 2-1 count, you’ll most likely find a dramatic difference. All of the data that I’m about to show you can easily be accessed on the Internet.

In a study done by Beyond the Box Score writer Scott Lindholm (@ScottLindholm), we find that umpires get the calls wrong on pitches within the strike zone 13.2% of the time, and 15% of the time on pitches outside the zone (his article can be viewed here). When looking at it from the school of thought that says an umpire gets a call wrong every (roughly) 8.5 pitches, it doesn’t seem like a big deal. What if that one pitch decided a ballgame (2 in the case of the situation we’re about to examine)? Then it obviously becomes everything. As we stated earlier with the importance of a ball and a strike, it depends on the situation.

Let’s take the game last Friday, for example. The Oakland A's are playing the Cleveland Indians in the first of a 3-game set before the All-Star break. A’s right-handed starter Kendall Graveman has pitched a terrific game until he runs into trouble with 2 outs in the 6th inning. A string of events occur where Graveman loads the bases, forcing A’s manager Bob Melvin to bring Drew Pomeranz out of the bullpen to face Indians’ All-Star second basemen Jason Kipnis. Pomeranz throws two good looking spike curves that should’ve been called strikes (as shown here via fangraphs.com), but instead home plate umpire Laz Diaz calls them both balls. That’s 2 bad calls in a row. Instead of Pomeranz being ahead in the count 0-2, he finds himself in an uncomfortable 2-0 deficit.

To look at how this may have decided the game, we look at Kipnis’ probability of getting a hit in different counts. In 0-2 counts this season, Kipnis bats .209. In 2-0 counts, he bats .375. Pomeranz then throws a fastball up and in to fall behind in the count 3-0, in which Kipnis bats .500 (you can find Kipnis' splits here). Now lets look at Pomeranz’ chances of executing in that predicament. When Pomeranz falls behind in a count 3-0, there’s a 57.9% chance that his opponent will reach base. When an opponent falls behind 0-2 to Pomeranz, his opponent’s chances of getting aboard fall to 14.6% (Pomeranz' splits). Kipnis does what he was predicted to have done: he gets on base via a walk. This gives the Indians the lead, which the A’s would never get back, losing by the final score of 5-1.

Instead of praising Bob Melvin for his managerial smarts in that situation, the general public now is angry with what they perceive as Pomeranz "blowing the game". When, in reality, it was the umpire who, according to statistics, strongly influenced the final outcome.

That’s how much a ball or a strike can mean and that’s how badly we need electronic strike zones. To put in a strike zone with 100% accuracy is to make a GM’s job that much easier, hitters that much more deserving of huge contracts, and pitchers worth the gaudy sums clubs tend to pay them these days. The human element has no place where it isn’t needed. That’s why computers do a lot more for Fortune 500 companies today than they did 10 years ago.

To my knowledge, there isn’t anything out there that will measure how much a bad call will affect the game as a whole. However, we can take scenarios like the above, analyze them, and present them as proof that umpires dramatically affect the margin of error that GM’s take into account when filling 25-man rosters.

I personally feel that too often we dismiss poor officiating as something that’s "part of the game". Too often do we overlook statistical probability and how much it is imperative to get calls right at critical junctures during play. In a game of inches, you can’t have faulty measuring equipment, if you know what I mean.