Pebble Hunting: How Bad Can Clogging the Bases Be?

We’re not great at holding a lot of small details in our brain for a long period of time, so we summarize and categorize, often remembering only the nut graph of a story rather than the specifics. I think we do this for baseball teams, too, and I’m sure I do it for baseball teams. I know a little bit about every Tiger, but when I think about the Tigers I mainly think along the lines of these bigger, summarizing narratives:

• Good team;
• In a weak division;
• Justin Verlander;
• Bad defense;
• They’re fat.

The Tigers' lineup weighs 211 pounds more than the A's lineup. That's a free Alex Avila. That's not fair!

— Sam Miller (@SamMillerBP) October 7, 2012

The last one, in particular, is the sort of thing that probably takes on way more importance than it needs to. Teams get labeled based on their strengths and weaknesses, but the strengths and weaknesses are just ingredients playing off other ingredients in very complicated dishes. Garlic is terrible in cake, but in soup it’s great. You don’t judge the dish by what ingredient is in it but on whether the whole thing is good or bad. Considering submitting this amazing analogy as my MacArthur Genius Grant application.

Anyway, the point is that the Tigers are large and slow. Listed player weights aren’t very reliable, but Detroit had the heaviest third baseman in baseball last year (minimum 100 games played), the heaviest shortstop in baseball, and the heaviest first baseman. Their new DH is a former catcher who has stolen a base every 272 games in his career and is now coming back from knee surgery. They were one stolen base better than the worst-stealing team in baseball last year, and 21 of their steals came from a player (Quintin Berry) who is now much lower on their depth chart. Matthew Trueblood asked us to talk about this on Effectively Wild (the daily podcast from BaseballProspectus.com), but I’d rather talk about it here. Says Matthew:

Speed is not a critical part of an individual's value. Maybe you steal 50 bases, but how often do you get caught? All of that. But on a team level, I think the research has generally concluded that having a lot of guys who run well can create a compounding effect and make an offense more efficient. The Angels and Rays consistently seem to outscore a third-level expectation given their individual outputs, partially because they pressure defenses and take a lot of extra bases.

The Tigers of 2013 look like the polar opposite of that, maybe the slowest team in 20 years. Fielder, Cabrera, Martinez, Avila, even Peralta, everyone knows are shockingly slow. But even their fast guys aren't fast. Andy Dirks and Austin Jackson are average runners, maybe a little better. Jackson was 12-for-21 stealing bases last year though. Then you get to Torii Hunter and Omar Infante, who are both average runners, maybe a little worse, and both are also past 30 now. The Tigers grounded into more double plays than any other team in the AL last year, and the two big additions to their offense—a returning Martinez and Hunter—grounded into a combined 35 double plays in their last full seasons.

Is this a problem? Will it matter, on a team so thoroughly capable of crowding the bases, if none of them are good at circling them? Just wondering.

So what follows is going to be a quick attempt to answer this question. I want to note up front that every step adds a bit of noise (you’ll see) that makes it harder to draw conclusions, so this is more about thinking through small questions rather than definitely answering the question. The question, by the way, if simplified to the most basic “Can a team as large and slow as the Tigers succeed?” was basically answered in affirmative about four months ago. Of course that team can. It just did.

First, we need to define speed. There are various attempts at scoring speed out there, and a few years ago Ben Lindbergh took Bill James’ attempt and measured how closely it hews to BP’s baserunning runs, on a team level. As it turned out, it hewed fairly closely: a correlation of .63 from 2007 to 2009. To put that in perspective, it’s roughly the correlation between team runs scored and team winning percentage: strongly suggestive, if not nearly the whole darned thing.

Because I want to go quite far back, I’m going to use the baserunning runs listed on our site as a proxy for speed. Not perfect! But at least suggestive.

So the first and most basic thing is that there is a positive correlation (going back to 1954) between baserunning runs at a team level and pythagorean record, though a weak one: about .21. That long of a timespan complicates things, because of different length seasons, different run-scoring environments, etc. So I further limited it to an era where run scoring was reasonably consistent, and moderate: 1970 to 1980. During that stretch, the correlation between BRR and pythagorean record was .28. Not very strong, but stronger.

But then, of course baserunning runs would show some positive correlation to pythagorean record: baserunning runs are runs, and pythagorean record is based on runs. What we really want to know is whether a fat and slow team has disadvantages aside from baserunning. If the disadvantage were limited to baserunning, then the answer to the original question would be very clear, with the gap between the best and worst baserunning teams usually no larger than about three totals wins.

So next step: Removing the baserunning runs from a team’s offense and seeing if speed now correlates to the baserunning-neutral pythagorean record. The correlation on that is weaker still:

• Overall correlation:  .12
• 1970-80 correlation: .20

but there is still some residual correlation.

Unresolved is where that correlation actually is. Some reasonable possibilities, most of which aren’t answerable by the spreadsheet I cobbled together for this:

• Better team defense;
• Better team health;
• Smarter overall team approach, which shows up elsewhere as well
• Younger players/more players in physical prime;
• Fewer double plays, more infield hits;
• More athleticism leading to more positional flexibility;
• Outscoring a third-level expectation given their individual outputs, partially because they pressure defenses and take a lot of extra bases.

The one of those explanations I can basically rule out is the link between baserunning runs and team defense; there’s a very weak correlation between the two. That doesn’t mean that speed and team defense aren’t correlated. We already established that baserunning runs are an imperfect proxy for speed score, and we know intuitively that speed score itself is an imperfect proxy for speed. Which is to say that we might be so far removed from actual speed at this point that the correlation between speed and team defense is entirely unknown within the context of this article.

So Detroit’s slowness might hinder it a bit, beyond simply the measurable baserunning disadvantage. The good news for the Tigers is that plenty of teams, in addition to the 2012 squad, have won a lot of games while baserunning awfully. Of the 195 teams since 1954 that have given back at least 10 runs on the bases, 32 of them won at least 90 games. That’s 16 percent, which is less than the overall frequency of 90-game winners (around 24 percent) but not by much.

Just 20 teams have been 20 runs worse than average on the bases since 1954, and if you remove the baserunning runs from the equation, those 20 teams had a .500 pythagorean record. Four of those teams won at least 90 games:

2001 Giants: -22 baserunning runs, led the league in TAv (by a lot)
2003 Giants: -21 baserunning runs, led the league in defensive efficiency
2004 Red Sox: -25 baserunning runs, second in TAv, third in FRA
2012 Nationals: -21 baserunning runs, led the league in FRA

Which is all to say that even if the Tigers are as bad on the bases as you might reasonably expect them to be, it’s merely a disadvantage. It’s not a threat to the entire plan.