December 18, 2012
Painting the Black
The Unimaginable Base-Stealing Genius of Coco Crisp
When Coco Crisp stepped to the plate last season, his walk-up song was “Who Gon Stop Me." The chorus served as a legitimate question once Crisp reached base. Not many have showed the ability to stop Crisp in recent seasons. He went 39-for-43 on stolen base attempts in 2012, pushing his three-year total to 120 steals in 136 attempts. That’s an 88 percent success rate on 45 attempts per season. None of the six other players with at least 45 attempts since 2010 succeeded more than 83 percent of the time. Rickey Henderson never had a three-year rate of more than 86 percent throughout his career.
The highest levels of baseball performance tend to have impressive streaks built in. Crisp’s basestealing in recent years has two worth noting. From July 10, 2011, until June 21, 2012, Crisp stole 36 consecutive bases without failure. Ignore pickoffs and Crisp’s streak creeps into August 2012 and encompasses another 15 steals. (The only catcher to stop Crisp in 2012 was Jose Lobaton.) Those high levels of performance leave us bewildered and curious. How is it that Crisp—who had a career 74 percent success rate prior to 2010—has turned into the league’s most efficient and prolific thief? Is it with Billy Hamilton-like speed, or Henderson-like wits and hubris*, or something else?
I watched each of Crisp’s stolen base attempts in 2012 in an effort to find out. My initial plan was to write descriptions of each attempt while timing the pitcher and catcher, then to assign blame to the slow one in the battery. Yet by the time I finished reviewing Crisp’s final 10 stolen bases on the year (I worked backward), it was clear that the catcher was rarely at fault. Opposing backstops managed throws on just three of those final 10 steals, leaving Crisp to swipe seven bags without a challenge. On the season, Crisp stole 20 of his 39 bases without a throw, including four double-steal situations.
As part of the review process, I kept track of a few variables along with the throw. Most notably, I reported the trigger on each attempt. If Crisp took off before the pitcher started his delivery, it was typically after a glance or two, so I recorded it as Crisp running on the look. If the pitcher had begun his delivery by the time Crisp took off, then it went down as Crisp running on the move. Here are the results in full:
Coco Crisp Stolen Base Log, 2012
Excluding double steals, Crisp ran 12 times on looks and 22 times on moves. (One couldn't be determined without a replay.) He drew four throws on look-triggered steals and 14 throws on move-triggered steals. When Crisp went on looks he was near unstoppable, in part because he focused on getting a good jump over getting a good lead. He knew he could make up for the step or two with a running start. Take the steal below against Tim Lincecum as an example:
The outer image shows Lincecum after completing a glance at Crisp. While Lincecum is staring in at the plate, about to begin his delivery, Crisp is dashing toward third. By the time the pitch is crossing the plate, the inner image, Crisp is about three-fourths of the way to the base. Ivan Rodriguez in his prime would’ve held onto the ball. The kicker is that Lincecum knew Crisp was in a running mood because he had swiped second the pitch before. Even with that knowledge Lincecum couldn’t stop him or make it a close play.
Crisp doesn’t just succeed when the pitcher neglects to look back at him. He feasts on patterns. There were instances this season where pitchers would throw over, fake a throw, step off, and offer multiple glances before delivering the pitch. Crisp would soak this information in, along with the position of the shortstop or the second baseman, and then make his decision to run later in the at-bat. Crisp’s confidence in his read is such that he took off on Rafael Betancourt after a fake pickoff attempt:
I left the score bug in to show the importance of the situation. Crisp would have heard about it if he'd been caught stealing, and the A's would have been doomed if Crisp had made that second out. Yet Crisp trusted himself and took off. And not only did he swipe the bag, he did it without a throw. He scored the tying run on a sacrifice fly later in the at-bat.
Who should get credit for Crisp’s stolen base genius? It’s unclear whether he’s a good guesser, a good video scout, or a good reader who took advantage of his team’s apparently über -detailed scouting reports. The most likely answer is a mix of all three. Crisp seems to have an advanced feel for reading a pitcher’s movements and identifying the point when the pitcher shifts his attention from preventing a stolen base to throwing his pitch. Whether that skill comes from study or from attention to the front office doesn’t matter.
While the genesis of Crisp’s brilliance is unknown, his Kryptonite is clear: Three of his four caught stealings were on pickoffs by left-handed pitchers, and in each instance the pitcher used his head in a manner that deceived him. First, Clayton Kershaw turned his head toward home plate while stepping and throwing to first base. Crisp took off on the head turn and was nabbed at second. Then Charlie Furbush got away with a head-nod prior to starting his delivery, the move led Crisp to start his attempt, and the Mariners were able to tag him out. Boone Logan picked off Crisp twice in the same at-bat, though the speedster beat a pickle to return to first safely the first time. A few pitches later, Logan got Crisp again when he used Kershaw’s head-fake:
PECOTA projects Crisp to add 5.5 runs with his legs next season and finish with the best BRR in baseball. So who’s gonna stop Crisp from continuing his assault on historically low caught stealing rates? If anyone can, it will be the pitchers who realize their head can be their best friend or their worst enemy against him.
*The best part of the exercise is when Oakland’s broadcast would show Henderson after a Crisp steal. Henderson, for his part, never appeared impressed.