June 2, 2004
The Modestly Productive Column
While watching baseball this Memorial Day, I started to think about how different each team's offense is. Watching the Mariners lose this season has been tougher because they're so goddamned boring. There's no variety of in the outcomes: Pretty much, they're going to slap a single, or they'll make an out.
So I ran the stats and...yeah, the Mariners are one of the most boring teams in baseball. When a Mariner batter comes up to the plate, there's a 17% chance he'll hit a single. Of all their hits, 72%--seventy-two-freaking-percent--are singles. Only the Padres matched that. The second most frequent event in a Mariners event is the exciting strikeout (ooh, Boone really took a cut at that one).
Which is the most boring team to watch, though?
Leading Most Event Frequent Team Percentage Event ------------------------------------ Cincinnati 20% K Milwaukee 20% K NY Mets 19% K Atlanta 19% K Colorado 19% K Boston 18% K Baltimore 18% 1B Detroit 17% K/1B St. Louis 17% K Los Angeles 17% 1B Pittsburgh 17% K Oakland 17% K Anaheim 17% 1B Toronto 17% K San Diego 17% 1B Philadelphia 17% K Seattle 17% 1B Chicago Cubs 17% K Florida 16% K Texas 16% K/1B Cleveland 16% K/1B Houston 16% K/1B Chicago Sox 16% 1B Kansas City 16% K/1B Minnesota 16% 1B NY Yankees 16% K Tampa Bay 15% K/1B Arizona 15% K Montreal 14% K/1B San Francisco 14% K/1BLots of ties on that list. What if we look at the two most frequent things a team does?
Top Two Team Event Percentage ------------------------------- Detroit 35% Milwaukee 34% Pittsburgh 34% Cincinnati 34% Boston 33% Toronto 33% Oakland 33% Los Angeles 33% Colorado 33% NY Mets 33% Atlanta 33% Houston 32% Seattle 32% St. Louis 32% Texas 32% Cleveland 32% Kansas City 32% Florida 32% Chicago Cubs 32% Anaheim 32% Chicago Sox 31% Philadelphia 31% Minnesota 31% Baltimore 30% San Diego 30% Tampa Bay 30% NY Yankees 29% Montreal 29% Arizona 28% San Francisco 28%That jumbles things up a little, but there's not much useful information to be gleaned there. The most crowd-pleasing teams so far in 2004 are the White Sox and Rockies, who hit home runs in 4% of their plate appearances. Fans love home runs, right?
And fans love seeing the ball put into play, even if it results in an out. We're told it's how you make things happen by nearly every color commentator.
Montreal leads the league in these modestly exciting outs: plate appearances that did not result in a hit, a walk, a strikeout, hit batter, or any kind of sacrifice. Montreal batters have grounded or flied out in 55% of their plate appearance. Meanwhile, over in Cincinnati, only 44% of their outs are modestly exciting. Modestly exciting out. That's kind of catchy. I wonder if it correlates to run scoring.
Holy mackerel, would you look at that? Modestly exciting outs do correlate to run scoring. This is going to shake baseball to its core: Teams that make modestly exciting outs in a higher percentage of their plate appearances score fewer runs. It's scientifically provable. Just look at these lists:
Top 5 Teams in Modestly Exciting Outs ------------------------------------- Montreal Tampa Bay Arizona Minnesota San Diego
Bottom 5 Teams in Modestly Exciting Outs ---------------------------------------- Milwaukee Detroit Colorado Boston CincinnatiOh, I know what you're going to say: "Derek, does the world really need another junk stat that's just a barely modified proxy for on-base percentage? If a team doesn't make outs, they'll score more runs. That's just common sense."
My response: I can't hear you, and you should type up a nice email I can use in a future column.
What about looking at Modestly Exciting Outs as a ratio to other outs? (Did you notice I'm capitalizing it already? In fact, I'm so worried that Elias is going to steal my idea that I've already copyrighted it in book form, like those helpful guys who name stars after people.)
I looked at each team's ratio of Modestly Exciting Outs (MEO) to other outs. What I found is that teams with the lowest ratios of other outs to Modestly Exciting Outs score--barely--more runs than their Modestly Exciting Out-producing opponents. It's a weak correlation, sure, just .377, and I already see that it's deeply flawed in a couple of ways. But if the big stat companies can promote stuff even less applicable than that, then so can I.
Joking aside, this is actually interesting. Ground-fly outs/strikeouts has a strong correlation to run scoring: the more strikeouts as a percentage of a team's total out share--I mean, uh, Out Share, a new innovative statistic, oh, I give up--the more runs that team scores. Now, correlation is not causation, as we'll tell anyone who listens to us (or even anyone who doesn't listen to us). Shannon Stewart joining your team before a nice run doesn't make Shannon Stewart the cause of that run.
For instance, power hitters generally strike out a lot, walk more than average, and score more runs. All three of those things, no matter how you measure them, will have a high correlation with run scoring. All this really shows is that if putting the ball in play was so important, teams that made most of their outs that way would score just as many runs as the teams that swung hard and often.
The most obvious objection here is that if putting the ball in play often is successful, generating lots of hits, then the infrequent strikeout would generate a disproportionate number of outs.
Let's check that out, and look at balls put into play by contact (MEO+H) versus strikeouts. Again, the correlation is negative and weak. Teams that put the ball in play more as a percent of at-bats, whether it results in hits or outs, are somewhat more likely to score fewer runs than patient, whiffing teams.
It's weird that poking into how boring my hometown team is can lead to something useful, but that's what's really cool about these kind of investigations. Now if I could only find some way to keep my blood pressure low while I'm watching the Mariners rap out those singles, I'd be in business.