The Milwaukee Brewers' seven stolen bases last night exposed the worst-kept secret in baseball: the Pittsburgh Pirates do not care about throwing out base stealers. They haven't all year, really. Runners have stolen 84 bases in 90 attempts against Rod Barajas and they've stolen 53 bases in 61 attempts against Mike McKenry. That's 137 stolen bases in 151 attemps. That's embarrassingly bad; the Pirates' CS% is about 9%, which is a third of the league average in 2012 of 27%. Barajas has a generally strong reptuation as a defensive catcher; in the five years prior to 2012 he's had caught-stealing percentages of 34%, 15%, 15%, 15%, and 25%. His career rate is 28%. Even if you take that skein of 15% seasons to mean his skills are declining, what we've seen this year is a stark departure from what was expected. The same goes for McKenry, who caught 36% of stealers when he was a minor league catcher. What that means is this: the problem goes well beyond the catchers. The Pirates have told their pitchers (and to an extent, presumably, their catchers) to ignore base stealers.
There's been a lot of wailing and tooth-gnashing over this at various points in the season, none moreso than right now, of course, as the Pirates hit rock bottom. It's undeniable that letting runners have free bases unimpeded is a bad idea, but there's more to it than that. Neal Huntington runs a front office that most people would consider to be a smart front office. He's got a stats team run by Dan Fox, who's certainly a sharp guy. Clint Hurdle (especially) and Ray Searage seem to be more old school baseball guys, but neither of them ever seem to complain about the carousel around the bases that's been turning singles into doubles all year against the Pirates. That says to me that whatever the reasoning is behind this decision, it's got both the front office and the coaching staff on board. That means it deserves some investigation. Why are the Pirates ignoring base stealers? What are the benefits? What does it actually cost them?
Let's start with the simplest question: How many extra runs have the Pirates allowed by virtually ignoring base stealers this year? There's an easy formula to calculate stolen base runs (hat tip: @nvasconcelos): SBR = (0.3*SB)-(0.6*CS). That is, if you steal two bases for every time you're caught, you break even. That means that the Pirates opponents have racked up in the ballpark of 36 runs on stolen bases this year. If we assume that the Pirates an average amount of base stealers this year (41 of 151), that number drops to about six. The Pirates have allowed 30 more runs on stolen bases this year than the average team, which translates neatly to about three wins. Since we're mostly talking in theoretical terms here, I'm going to be kind to the Pirates and say that they've cost themselves between two and three wins by ignoring base stealers; Barajas may well have been below average at catching stealers even with the benefits afforded other catchers, plus teams would run on them less often if they appeared interested in stopping them.
Two or three wins is not insignificant, though! It could well end up being the difference between a playoff spot and not a playoff spot for the Pirates this year, and that's true for any team in contention. The assumption on my part is that the Pirates must think that they're reaping some kind of benefit from this that off-sets the loss of those 2-3 games. I'm not privy to the Pirates decision making process, but I'd assume that the reasoning is has a couple of levels. One reason is that throwing from a slide-step is annoying and it messes with the rhythym of a lot of guys, which can affect a lot of things. Another reason is that if the catcher has reason to think a runner is going to steal and he's intent on stopping him, he's probably going to call for a fastball even if it's not a situation in the pitch sequence that he would normally call for one, which could result in hitters having a better chance to hit the ball than they normally would.
It's pretty hard to measure these sorts of effects, but let's try to get an idea. You might expect that pitchers being more comfortable in the stretch would result in the Pirates having a higher strand rate than their opponents. It doesn't; the Pirates' strand rate this year is 72.7% and the league average is 72.4%. The Pirates are basically an average pitching team (their team ERA+ is 97) and their strand rate is basically average.
We could pull this out for a bigger picture look: how do the Pirates pitch in bases empty vs. bases occupied situations as compared to the league as a whole? An easy way to do this is to use Baseball-Reference's tOPS+ splits, which lets you compare one split to total performance. In 2012, National League pitchers have a .705 OPS against with the bases empty and a .745 OPS against with at least one runner on base. That's a 95 tOPS+ with bases empty and a 106 tOPS+ with at least one base occupied (100 is average: since we're talking about pitchers here, below 100 is better than average, above 100 is worse). The Pirates' tOPS+ split for bases empty/occupied is 96/106. In plain English, the Pirates are no better with runners on base compared to bases empty than an average National League team.
It's a little more complicated than this, of course, but our superficial glance here tells us that the Pirates are giving away 2-3 wins this on the bases by ignoring base stealers and reaping absolutely zero appreciable benefits. I could be missing something, of course, but it's awfully hard to see how any sort of rational decision-making process lead to this sort of situation.
Your analysis is flawed in that you're comparing apples and oranges.
Comparing runners on and bases empty tells you nothing.
What you need to do is compare A) how the Pirates pitchers have pitched this year with runners on... with B) how the Pirates pitchers WOULD HAVE pitched by paying more attention to the runners (more fastballs, looks to first, pitchouts, etc.) This would give you strikes/outs/runs/wins saved by ignoring the runners. Compare that to the outs/runs/wins lost by ignoring the baserunners and you'd have your answer.
Obviously this is impossible. But I'm guessing Dan Fox has done some regressions and has proved that paying attention to the runners isn't that important.
@guapo Well, that's why I included the splits. I'll admit that it's possible that the Pirates have a unique group of individuals that are, as a whole, way worse at pitching out of the stretch than every other pitcher in the league, but I think that it's pretty telling that those tOPS+ splits come out at almost exactly league average. We can't say /for sure/ that the Pirates get no advantage from this, but I think assuming something else is even more unlikely.
@whygavs The point is, that as a credential media member you can't print stuff that is absolutely wrong and claim it's correct with the disclaimer that "It's a little more complicated than this..."
Mathematical statistics (not statistics like RBI's, etc) are difficult for a lot of people and if you're not good at them you need someone to check your work before you publish.
Dejan is an idiot, you're not.
What I like about your article is that you've put a 'games lost' number on this fundamental problem. What's even scarier is that you could probably do the save thing for other Pirates problems this year such as 'double steals with 2 outs', 'bad base stealing', 'terrible bunting decisions', etc., etc. The list APPEARS to go on for 'games lossed' due to terrible coaching decisions and fundamentals, no?
It is unclear to my why your analysis of the benefits to pitching leads you to the conclusion that there aren't any. You are assuming that if the staff were to pay attention to baserunners they would still be at or near league average in the stats you cited. There is no evidence to support that assumption. We have no idea what the benefit is without having a baseline to compare.
@rowhoo Every other team in the league holds baserunners, and their pitchers lose about 11 points of tOPS+ when runners are on. The Pirates do not hold baserunners, and their pitchers lose... about 10 points of tOPS+ when runners are on. That is pretty strong evidence that the Pirates aren't gaining anything measurable by ignoring baserunners.
So conversely if there is limited benefit in stealing bases then why is it done? Statistically a runner on second will score more than a runner on first (I can't produce that statistic but it's common sense) ... again 'common sense' tells me that stolen bases have caused the Pirates to lose more than 2-3 games this year as I can come up with at least 2 against the Brewers in the last 2 series.
Most of the studies on the issue suggest that stealing bases is a useful tactic if you're successful a significant percentage of the time, with two-thirds being about the break-even point. So if you run a lot and are successful only 60% of the time, the thought is you're losing ground. If you're successful 90% of the time, as teams are against the Pirates, you're going to score more runs that would just staying put.
Common sense dictates that giving away free bases is self-damaging, and as you point out, there's no hard evidence to contradict common sense here.
I would bet the majority of those 9% of "base stealers" thrown out by the Pirates this year have come on busted hit-and-runs and such.
I don't know if it's a matter of "common sense" (although what you say about giving away bases is perfectly...well, sensible) per se, but it is odd that a front-office which has embraced the so-called "new statistics" would see that there is a measurable loss in what appears to be an organizational philosophy without a concurrent measurable gain in other areas. If that is the case, then how do you justify the action without resorting to an explanation involving something intangible and not measurable, which is (in a large part, anyway)why sabrmetrics came into being in the first place.
@whygavs That's good analysis...certainly appears to answer the question I had about potential benefits