Art of the Steal

Ichiro had one of the best basestealing seasons ever in 2006 when he succeeded on 45 of 47 tries. Kevin P. Casey/AP Speed Kills The following are the best (minimum: 200 stolen bases) and worst (minimum: 100 stolen bases)? career thieves among active players. MOST EFFICIENT Name, team Steals Attempts PCT. Carlos Beltran, Mets 270 307 87.9% Jimmy Rollins, Phillies 286 346 82.7% Carl Crawford, Rays 302 366 82.5% Ichiro Suzuki, Mariners 312 382 81.7% Dave Roberts, Giants 242 299 80.9% LEAST EFFICIENT Name, team Steals Attempts PCT. Luis Gonzalez, Marlins 128 215 59.5% Cristian Guzman, Nationals 116 176 65.9% Brad Ausmus, Astros 101 153 66.0% Vladimir Guerrero, Angels 173 259 66.8% Damion Easley, Mets 114 170 67.1% Related Links BALLARD: Rickey Henderson and Jimmy Rollins discuss their base-stealing secrets

It begins with a puff of infield dirt, a tiny smoke signal sent up from near first base. Translated for the pitcher, it might read, you are toast. With that liftoff, the crossing of left cleat over right, Philadelphia Phillies shortstop Jimmy Rollins, the best base stealer on the best basestealing team in the league, is headed for second. Let's freeze it right there.

See the catcher? The poor guy is helpless. With elite runners like Rollins, catchers generally have very little control over the play. What matters is how quickly the pitcher delivers the ball to the plate. Faster than 1.2 seconds is excellent; slower than 1.4 and the would-be thief gains such a head start that, as Phillies baserunning coach Davey Lopes puts it, "The catcher can have a bazooka back there and it won't matter." Regardless, the man behind the plate is the one whose reputation is on the line. His caught-stealing percentage is what people look at, so he's bracing, hoping to uncork a perfect throw. With Rollins, he'll need to.

Next check out the shortstop. With the lefthanded-hitting Chase Utley at the plate, he's already moving to cover second base, creating a fat hole on the left side of the infield through which Utley can send a line drive or a hard grounder. This is one perk for hitters batting behind a speedster like Rollins; another is that they'll likely see an increase in fastballs. (The quicker the catcher gets the ball, after all, the quicker he can throw down to second.) The end result: According to Baseball Prospectus, a hitter's batting average jumps 15 points when a basestealing threat is on first, with a slight bump in power as well.

On to the fellow on the mound. At this point in his delivery he's already committed, for better or worse. If he's unfurling a slow, looping breaking ball, then Rollins is as good as safe; a fastball and it might be close. Might be. Either way, the pitcher's already lost the game within this game, the goal of which is to keep Rollins tethered to first by slide-stepping, throwing over or holding the ball. (Do it long enough and a runner's legs, coiled to sprint, start to ache.) In this case, however, Rollins correctly read the pitcher's "tell" just as a poker player might -- it could have been a slight dip of the shoulder, an incline of the knee or even an unconscious glance downward -- and knew he was initiating his delivery. That's when Rollins took off.

Finally turn your attention to the short guy in pinstripes causing all this commotion. Head down and knees driving, Rollins is drifting ever-so-slightly toward the infield grass as he runs, a habit that drives Lopes nuts. (During practice, Rollins invariably runs straight.) At this point Rollins has one of two thoughts on his mind -- "Either, I'm going to be safe, or Oh, s---, I better get going," he says. In a few more steps he'll sneak a glance toward the catcher and evaluate the situation. Is the ball in the dirt, making a slide unnecessary? If not, should he go in straight or fade to the right to avoid a tag? And when should he ease off the throttle, lest he repeat the mistake he made against the Washington Nationals in August, when he beat the throw to second but kept on sliding, as if he'd suddenly hit a patch of ice, until he was stranded a few feet from the bag, baserunning's equivalent of a beached whale? That one still rankles him, and understandably so. It's one of only two times this season Rollins has been caught. The other came when Atlanta Braves catcher Brian McCann threw him out. ("Just a damn good throw," says Rollins.) Every other time, all 38 of them through Sunday, he's been safe.

If 38 of 40 sounds like an impressive ratio, that's because it is. Only once in the last 86 years -- in 2006, when the Seattle Mariners' Ichiro Suzuki went 45 for 47 -- has a player stolen more bases while being thrown out so few times. So while Rollins's steal total may not be as gaudy as, say, Willy Taveras's league-leading 66 for the Colorado Rockies or Rickey Henderson's eye-popping totals from the past, his thefts may be more valuable. Sure, Rickey once stole 130 bases in a season, but he was also caught 42 times that year. Forty-two! At his current rate of 95%, Rollins would steal 798 bases before he got caught 42 times. And in this post-Moneyball age, it's not how many you steal that matters, but how efficient you are.

Why? Well, consider that every stolen base increases a team's expected runs by .25 per game according to Baseball Prospectus, whereas getting thrown out reduces them by .64. In other words, to make it worthwhile, a team needs to be successful on roughly three out of every four attempts. By that measure, even though players are running less than in decades past, we are in something of a golden age of theft. The average efficiency in the majors this season is 73.4%, second in baseball history only to last season's 74.4%, and the most prolific basestealing teams this season are all playoff contenders: in order, the Rockies, Rays, Mets, Dodgers, Phillies, Red Sox and Angels. They're also the most effective; Rollins's Phillies lead the majors in stolen base percentage, at 83.8%, followed by the Rockies (81.1%) and the Red Sox (79.6%). At a time of year when an extra run here or there can mean the difference between an October spent playing ball and one watching it, that can be a huge advantage. "It's a great weapon, but the key is to use it correctly," says Lopes, who stole 47 bases in 51 attempts for the Cubs at the age of 40. "It's a matter of becoming a student of baserunning. People associate speed with basestealing, but I've known a lot of guys who had great speed but were terrible base stealers. There's an art to it."

So what is that art? Well, everyone knows how not to steal a base: jump too early and get picked off, go too late and get gunned down or just be plain dirt slow. In the annals of bad baserunning, there are plenty of cautionary tales. The Mariners' Harold Reynolds was nabbed 29 times in 64 attempts in 1988, and the Cardinals' Garry Templeton went 28 for 52 in 1977. Then there's former Giants first baseman Will Clark, a sublime hitter who nonetheless put together perhaps the most unsightly basestealing season in history. In 1987, Clark broke for the next bag 22 times. Only five times did he successfully make it. (In his defense, Clark explains that it was partly a managerial thing. "We had Roger Craig as our manager and we did a ton of hit-and-runs," says Clark. "I mean, he'd hit-and-run with anybody, and he said, 'If I put a hit-and-run on, whatever you do, don't get picked off.' So you rarely got a good lead.")