Thursday, March 31, 2011

"Small Ball"


Growing up, I was convinced small ball was an important part of baseball at every level. When I played, I made sure I could do all the things associated with "small ball" because everyone associated with baseball had always told me that small ball was a huge part of the game and a team that could do "all the little things" could beat a much better team.

As I get older, I find this sentiment idiotic. This is not to criticize any of my former coaches, teammates, or even kids I've coached. I don't blame them one bit, because "small ball" works better the younger the players are. For example, in an in-house city league full of 7-year-olds, bunting and stealing are likely to be against the rules. That's because there's going to be a plethora of kids who have no idea how to play baseball and it's simply much easier to just make the game simple at the age level. Also, these people simply didn't know any better; if the majority of 'baseball people' think something is true, chances are casual fans are just going to agree.

Once they do allow bunting and stealing and all the other important parts of small ball, though, small ball gets tougher and tougher as you get older because more advanced players can routinely make difficult plays look easy.

The Twins open up the season in Toronto tomorrow (Friday), and I couldn't be more excited. However, the Twins decided this off-season to build a more "small ball" lineup around their home-run threats. Speed, bunting, bat control, etc. were all cited by the team's front-office and manager this off-season when discussing potential moves. The team traded JJ Hardy because Ron Gardenhire wanted more speed in the lineup; which would be fine, except stolen bases are among the most overrated statistics in all of sports. And Alexi Casilla sucks, but I've been over that enough.

I suspect we will see a lot more bunting this season than last season, and most fans think that's a good thing. Unfortunately, it's not. I recently purchased The Book: Playing the Percentages in Baseball by Tom Tango and others. I wouldn't have found it without the help of Wet Socks blogger Derek Wetmore (who also writes for other places, so check it out) and I appreciate it.

Anyways, "The Book" is fantastic. Run Expectancy is my favorite statistic in the book. Basically, it's the chances a team has to score in specific situations, based on how other teams have fared in the same situations. On average, when a team starts an inning, they will score about .555 runs before the inning ends. Obviously teams can't score half runs, but basically a team scores a little more often than once every two innings. A single to lead-off the inning moves it up to about 0.95 runs before the inning ends. Every situation will change the chances a team has to score, and for each new situation there is a new percentage chance that the team will score.

You may be thinking that these statistics are flawed, though, because obviously every situation is different. You may want to bunt Nick Punto with a runner on first and no outs, while you would never ask Joe Mauer to do that. When teams need just one run late in the game, they always seem to bunt a runner over as long as the guy at the plate isn't a middle-of-the-order type bat. However, this is statistically the wrong decision. Every single season for more than 100 seasons has provided the data necessary, and the fact is more runs have scored with no outs and a runner on first than one out and a runner on second. When you bunt, you are actually LOWERING your odds to score. The only exception is bunting a pitcher, because chances are they are such poor hitters that they are likely going to strike out or hit into a double play.

Since 1969, your chances for scoring one run are also statistically higher with no outs and a runner on first than they are with one out and a runner on second. That's just for one run, which is important. This means that the data isn't being influenced by high scoring early innings--but rather that even if your team ONLY needs one run, bunting is statistically incorrect.

So, basically, not only did the Twins give away an above-average shortstop on a below-market one-year deal for two minor league relievers, they replaced him with a mediocre bench player because said mediocre bench player can run fast and bunt. That means the Twins are going to try to steal more bases, which is worrisome. A team needs to be successful on 75% of their steal attempts to break even. Alexi Casilla over his major league career is a fantastic 35 for 39, which is about a 90% success rate. If he could maintain that, his speed would be an asset. However, those numbers are scattered over five seasons and the sample sizes each year are small enough to suggest Casilla picked his spots (which is smart) and if he runs more often, he's likely going to get thrown out more. In the minors, Casilla was just 164/219, which is a 75% success rate. Because most major league teams are much better than minor league teams at holding runners, usually a minor leaguers stolen base success rate will drop by 4-8% at a minimum. That would put Casilla below the required success rate to break even, which when combined with his below average offense and likely below average defense, makes you wonder why the team seemed so willing to move JJ Hardy.

It's unfortunate, but not entirely unexpected. In a much stronger division this coming season, the Twins gave away multiple wins by replacing Casilla with Hardy, and if they use small ball as much as I expect them too, they'll probably give away three or four wins that way too. I will obviously be cheering for the Twins, but I'm certainly more down on the team than most. I'd love to be wrong, of course, but the numbers don't usually lie.