keyboard_arrow_uptop

On Monday, Orlando Hudson stole his 10th base of the year, placing him among the league’s leading larcenists. Prior to this season, Hudson had never been a prolific basestealer; in fact, the Twins let him walk over the winter precisely because they wanted to replace him with someone speedier. With that May 2nd swipe of second, Hudson matched his career high for single-season steals in less than a month’s worth of action. It’s unlikely that the 33-year-old has gained a step at this stage of his career—in fact, his hamstring has already lodged a formal complaint—so why is he suddenly burning up the basepaths?

If Hudson’s basestealing ability hasn’t improved, perhaps the explanation for the uptick can be traced to an increased incentive to put his skills to work. Hudson has always been a decent percentage basestealer—he entered the year with a 73.2% career success rate, which has since risen over three percentage points on the strength of his 10-for-10 streak to start the season—but he’s clearly decided in his tenth major-league campaign that quantity might make a nice complement to quality.

Hudson’s more aggressive approach on the basepaths owes a lot to his off-season change of address. He probably wasn’t possessed by the spirit of Rickey Henderson after putting on a Padres uniform—after all, he’s still speaking in the first person, and he’d already played for two of Rickey’s other former teams without a similar metamorphosis—but he was forced to adapt to an adverse offensive environment. The Padres are baseball’s lowest-scoring team—only partly because they play in the majors’ most cavernous park—and as a result, they’ve attempted to gain an edge with their legs, pacing the majors with 37 steals. As Hudson himself explained, “[Here] we get on and we just have to do what we can to get into scoring position…I’m doing whatever I can to get that extra base.” Padres bench coach Rick Renteria expressed the same philosophy in a slightly more measured tone: “The reality is we’re trying to gain some benefit from that risk that we’re taking. But the risks we take are calculated. They’re not just reckless.”

The Padres might be the team least capable of scoring and most enamored of stealing, but they aren’t the only club packing it in at the plate and burglarizing bases these days. The truth is that there’s a little Padre in a lot of players this season. As you’ve probably read, scoring is down for the sixth straight season; even though we had to wait a couple of weeks longer for our first no-hitter, teams managed only 4.29 runs per game this April, compared to the 4.55 runs per game they scored a year earlier, in the inaugural month of the so-called “Year of the Pitcher.” This season may already have usurped 2010’s “Year of the Pitcher” title, but low ERAs aren’t the only hallmark of this most recent offensively challenged era. This latest Year of the Pitcher might come to merit another moniker: the Year of the Stolen Base.

As Sky Kalkman noted on Twitter earlier this week, many more bases were stolen this April than in April of 2010, despite a substantial drop in league on-base percentage, which fell from .331 to .319. However, there were also over 100 more games played. To avoid the problems posed by a comparison between months with differing numbers of both games and baserunners within those games, we can look at steals on a rate basis, rather than in a cumulative sense. Last season, runners attempted to steal in 5.5 percent of opportunities (defined as situations with a man on first, a man on second, men on first and second, and men on first and third). That was roughly the same rate as 2009’s, which was itself a significant increase over 2008’s rate of .051.

Before last night’s game, runners in 2011 had stolen in 6.3 percent of opportunities. That small percentage increase may not sound like much, but in the exciting world of stolen base attempt rates, it’s quite large. Since fewer runners are reaching base, the full impact of the rate hike may not register in the raw totals, but the increase from last year’s rate to this year’s rate would result in roughly 600 more steals over a full season, given a number of opportunities representative of recent campaigns. This isn’t anywhere near a historical high—just as current scoring levels aren’t anywhere near a historical low—but it does seem to represent a sea change of sorts.

When league-wide run-scoring is low, the value of a stolen base increases, since runners can no longer take conservative leads and expect to be forced home by flurries of hits. Naturally, you’d expect teams to run more when the win value of a steal is high, since the potential payoff is greater. (As it turns out, some curious trends show up when we examine whether they actually do behave this way, but we’ll save that topic for an upcoming article.) The graph below illustrates the historical relationship between the average win value of a steal and the offensive environment in which it was recorded. (The outlier on the left represents the notorious 1968 season, when steals were especially precious.)

For those scoring at home, that’s a strong correlation of -.89, which indicates that as league-wide scoring decreases, the value of a steal tends to rise, offering increased incentive for a runner to go. The higher the potential payoff, the less often a runner has to be successful in order to attempt a theft. Consequently, the so-called “break-even point”—the rate of success at which stealing makes sense—is tied fairly closely to league scoring levels:

Even though the attempt rate has risen since last season, the success rate has barely budged, dipping slightly from 72.4 percent to 71.8 percent. One might have expected a more sizeable drop in light of the higher percentage of runners on the move, but any decrease at all is somewhat out of character, since the game has treated its would-be basestealers increasingly kindly over the past several decades:

There’s no definitive explanation for this slow-but-steady shift toward successful larceny, and like many difficult-to-decipher effects, it’s probably attributable to a combination of factors. Dan Fox speculated about the reasons for this trend a few years ago, crediting the same “ever-increasing level of play” that he identified in last week’s selection for the BP Wayback Machine. Dan speculated that baserunners’ speed may be improving more quickly than catchers’ throwing ability; perhaps it’s also outpacing pitchers’ innovations in the art of holding runners. It could also be that pitchers are taking longer to deliver the ball as they leverage every inch of their frames in an effort to contend with more skillful foes at the plate. Another possibility is that the greatest strides have been made in the dugout, as managers have gained a better sense of when not to send their runners.

The increase in stolen base rate this season likely reflects a sound strategic adaptation by huddled masses of hitters yearning to breathe free against the Halladays and Harens imposing a new offensive order, another expression of the game’s self-regulating system of checks and balances. It may take players some time to wake up and smell the run environment, so we could also be seeing the belated effects of a lag in adjusting to last season’s on-base outage—players’ scoring and stolen-base rates picked up this April much closer to where they left off last season than where they started it. Of course, there’s a limit to how far teams and players should take this behavior: even the 58.3 percent break-even rate in 1968, the real Year of the Pitcher, wouldn’t have been low enough to make giving Juan Pierre and his current 6-for-14 success ratio—plus a pickoff last night—the green light, although that isn’t stopping Ozzie Guillen.

Thanks to Dan Turkenkopf and Sky Kalkman for research assistance.

You need to be logged in to comment. Login or Subscribe
rawagman
5/05
I wonder if the increasing SB attempt trend could be at least partially attributable to a changing in the guard of MLB managers. For example, in Toronto, John Farrell has been a much larger proponent of the running game than last year's skipper, Cito Gaston, ever was. In fact, in many circles, Farrell has been soundly castigated for this tendency. Could the year-over-year numbers be teased to isolate the effect's of different managerial tendencies? (All under the assumption that the vast majority of stolen base attempts are either explicitly called by the boss, or under green-light provisos).
junglechef
5/05
Nice use of "larcenists".
eas9898
5/05
I for one like the increase in running and hope it sticks around. Among other things, it really adds excitement to a game, when there is a guy on first and everyone just knows he wants to take second. Watching a bunch of fat guys go station to station has gotten old.
dollarbill13
5/05
.063 percent of the opportunities? I'm guessing you mean 6.3%?
bornyank1
5/05
Right, sorry about that. Fixed.
BindleStiff
5/05
The run environment in 1968 was roughly 3.4 runs per game. You note that this correlates with a SB Breakeven Point of approximately 59%. While we haven't seen that type of low overall run environment in a while, we still see late-game situations against top closers where the run environment would presumably drop to 3.4 runs per game or even much lower. For example, Billy Wagner last year had a 1.61SIERA (and a 1.43ERA... oh Billy why did you have to go and retire...) If we assume that Wagner was "creating" his own run environment of roughly 1.75 runs per game, then would this drop the SB Breakeven Point for an opposing baserunner running against Wagner to a level a very low level? If you do a straight-line extrapolation using the SB Breakeven Point/Runs Per Game relationship illustrated in the second table above then the SB Breakeven Point for a run environment of 1.75 runs per game would be approximately 52%. Does this mean that Charlie Manuel would be correct in giving Raul Ibanez the green light even if he estimated that Raul only had a ~52% chance of success against a Wagner/McCann battery? This seems right to me, other than maybe having to factor into the run environment the small possibility that the game would go into extra innings. And perhaps the breakeven level dips below 50% if Wagner is facing the bottom of the Phils order (say 2010 Dobbs/Schneider/Valdez) rather than the meat of the Phils order? As an ardent long-time sim-leaguer (shout-out to any other APBA BBW fans on BP!), and a fan of in-game strategy more generally, this kind of stuff really interests me.
alangreene
5/05
Fantastic post. This is a quibble I've long had for years with any analysis done on run expectancy and decision-making. It's always done on league average. In many cases, this is not useful. Your team, the ballpark, the opponent, etc. all play a major role in decision-making, and using league average info can lead to a poor decision.
dturkenk
5/05
This analysis is actually done on average win value, so it takes into account the fact that stolen base attempts tend to occur in higher leverage situations than average. But your point is a good one that looking at things in aggregate can be very misleading in the particular. I've long thought about manipulating win probability added based on a given batter/pitcher matchup but then it doesn't necessarily sum to 0 and means something else. Plus it's a lot of hard work.
jlebeck66
5/05
First of all, good post. Second of all, a shout back at ya from another APBA BBW player.
BindleStiff
5/05
@R.A.Wagman Since the Jays have actually been pretty successful on the basepaths (34SB:11CS overall) why is it that Farrell's tendency drawing criticism?
rawagman
5/05
from all I've read, the raw SB:CS numbers aren't reflecting a few pickoffs that were busted by the defending team. Also, a few of the caught stealing/pickoffs/thrown out trying for an extra base scenarios have been just odd - like a few days ago, when Juan Rivera was caught trying to steal 3B - why run Juan Rivera?
rawagman
5/05
Why run Juan Rivera - try saying that fast three times in a row
BindleStiff
5/06
Great article/analysis by the way. Really made me think. I have one other question that is not addressed directly in the article. Does higher or lower leverage actually change the SB Breakeven Point? I wouldn't think it would change the SB Breakeven Point simply because when leverage is higher the average win value of a stolen base increases, but the average (negative) win value of a caught stealing also increases. So I would have thought that the SB Breakeven Point would stay more or less constant no matter what the leverage is. (All other things remaining equal of course). I could be quite wrong on this though... anyone? I loaned my copy of The Book to a friend, otherwise I would have attempted to figure out the answer to this by referring to the tables therein (which are not available on the web as far as I know).
mikefast
5/06
I'm not sure which tables you are referring to, but Tango's win expectancy tables are available here: http://www.tangotiger.net/welist.html We have our win expectancy reports available here: http://www.baseballprospectus.com/sortable/index.php?cid=110844
WaldoInSC
5/08
...or perhaps Hudson's newfound baserunning philosophy is his way of informing Twins brass that they might consider inserting their leadoff spot in their colonic orifice.