The number values to running aggressively
Including a long bonus ramble on the offensive character of the 2002 American League!
Taking the extra base was one of the frequently-cited qualities about the World Series Champion 2002 Angels to the point of exhaustion, and is less often mentioned as a quality of that franchise in general during the Mike Scioscia years.
The Mariners are attempting a similar strategy to what the Angels are said to do: If you’ve got a single, go for two. A double, go for three. Now hitters do this but generally speaking, they’re pretty conservative about taking the sure base over the chance at greater riches.
Is it worth doing? When does it become counter-productive?
Back to the 2002 Angels. It’s hard to distinguish the attempts at extra bases from a team’s overall character, for a couple of reasons. Particularly, extra bases are a product of the team’s speed (Adam Kennedy’s a lot more likely to take second on a liner to right-center than, say, Frank Thomas is). The Angels stole 117 bases that year (and were caught 51 times). So they’re stealing and (supposedly) taking the extra base. Where does strategy and team composition meet here, and how to divide them?
First, take HRs out of it. How often did an at-bat result in an extra-base hit? Here the Angels were #3 in the AL. This immediately reveals some odd things, though: the Mariners, #2 in steals, fare quite badly at XBH/AB. The Red Sox, who didn’t steal, do well. The Mariners, who tried stealing a ton, don’t, and neither do the Royals. Speedy teams weren’t getting extra bases, and slow teams were. Hmm.
Let’s try something different: what percentage of hits were turned into extra-base hits? That is, (2B+3B)/H. Again, strangeness: the list goes Baltimore-Minnesota-Boston-Toronto-Texas.
Even Edgar and Olerud hit a ton of doubles, though. Maybe the double’s not the best method. The real mark of aggressive baserunning and speedy teams is the triple. Who pulled that off more, as a percent of their non-HR hits? Kansas City, Detroit, Toronto, Tampa Bay… Anaheim’s almost at the bottom.
So we shrug and move on. Some people believe that aggressive baserunning is worth a run a game, but no matter how we look at this (what teams scored more runs than you’d expect, given the number of balls they put into play, or hits they got, or… whatever) there’s no way the effect’s that big. There’s a good Michael Wolverton article from 2004 where he looks at this and finds the difference in 2003 between the best team in the league at not being caught stretching and the worst was 15 runs. The Angels were the second-best team in baseball that year.
15 runs over a season is hard to distinguish from noise. I’d be interested to see if that pattern exists over a long period of time: did Sciosia-managed teams do consistently well over years? That his players were thrown out so rarely in 2003 suggests that they were not as aggressive as widely believed, since Wolverton’s study focuses on the cost of baserunning outs. The Angels could have scored 100 runs by advancing more than they should have — but if they were doing that, even with a tremendous success rate, they’d still have been on the leader board there for their failures.
Back to the point. Stretching an extra base is, from a value standpoint, exactly the same as trying to steal a base. You start with the value of being on the base you could get with 100% certainty, and you’re risking the value of being one base further against the damage of being out. There are obvious real-world differences that affect the decision (where’s the ball, who’s fielding it, and so on), but in the end, it’s the same kind of wager that stealing bases is.
Fortunately for my purposes, a ton of research on this has already been done. For purposes of simplification, make it a generic situation, so I can use a run expectation table.
Say that leading off, Raul Ibanez hits a ball into left-center, and it’s a sure single. In 2005, teams with a guy on first with no outs scored on average .9 runs.
Should he make the turn to second? Teams with a guy on second with no outs scored on average 1.1 runs, which means it’s worth +.2 runs if he makes it.
But if he’s thrown out… teams with one out and no one on scored only .3 runs/inning. So it’s -.6 if he’s caught trying.
This is easy. 75%, right?
Wow, that’s tough. Stretch a double to a triple? The value of being on third with no outs is pretty high — 1.5, a gain of .4 runs — but the difference between being on second with no outs and having one out and no baserunners is also larger — .8 runs.
You need to be able to get there 2/3rds of the time to make it worthwhile.
You see where this is going: as a strict value proposition, trying for the extra base is worth it when the player can do it successfully in the vast majority of situations. Every time they fail, it’s a pretty nasty knock to the team’s chances that inning.
In this sense, it becomes clear that there are players you want trying this: more than speed, you want players who can judge their chances on the fly, maybe able to head back if the fielder who looked like they bobbled it comes up throwing, and you need coaches who can help. If a guy’s going to go every time, they’re going to get killed out there. So there are players who may not be well-suited to this. The trick would be to figure out who those guys are, and get them to knock it off, while the others keep at it.
So if you see some players getting thrown out now, so be it. If they’re thrown out over and over stretching their hits, that’s when you should get really mad about it.
Up soon: the secondary effects. Putting pressure on the defense, keeping the rally going, and so on, and good stuff like that.