The completely predictable Yoenis Cespedes

Stealing bases is hard. Or, at least, it looks really hard on TV. I've never actually tried to steal on a professional catcher and that's probably a good thing all the way around. I can envision a solid face plant slide like this one, a pulled hammy or some other kind of wild, embarrassing failure. But some people are really good at it, and you've probably noticed them. This season, when you think of steals, Dee Gordon, Billy Hamilton or Jose Altuve probably come to mind. Someone who probably doesn't come to mind is Yoenis Cespedes.

Let me back up a minute. Stealing second is tough work. The runner has to read the pitcher, take a lead, likely avoid a couple of pickoff attempts, get the perfect jump, run as fast as he can and slide into second base ahead of the oncoming throw and tag. Yes, there are a lot of variables here, meaning, there's a lot that can go wrong. The throw has to travel at least 127 feet, plus whatever distance the catcher is behind the plate when he releases the ball. The covering infielder has to catch the ball and apply the tag while the baserunner is bearing down on him. Collisions are likely if one's not careful. As we know from the rate at which steals are attempted, they're a risky proposition.

Even more rare is the steal of third base. We don't see it all that often for two reasons. First, by sheer distance, the runner has to cover the same amount of ground while the catcher only has to throw 90 feet, not 127-plus. By the dimensions of the diamond, this is a much shorter throw, meaning the ball presumably arrives considerably sooner, giving the runner less time to get into the base safely. The runner may get a slightly bigger lead as pickoff throws to second are less often attempted than pickoffs to first, but the fact remains that this is a tougher proposition. If a right-handed batter is in the box, the catcher has a bit of a chore to execute in working the throw around the batter, but still, this is something we see far less often than attempts to steal second.

The second reason steals of third are attempted less often has more to do with the value of the outcome of the play. If the runner is safe at third, well, he's 90 feet closer to home, but most runners can score from second on a single anyways, so there's not much of an advantage gained. If the runner is thrown out, the team just lost a runner who was in scoring position, which is a valuable commodity. Let's take a look at some run expectancies to illustrate, using the second table here.

For simplicity's sake, let's assume that the runner on second is the only baserunner. If the runner is successful in stealing third, the chances of the team scoring a run in that inning increase as the table below shows.

Okay, cool, if the runner is successful stealing third with nobody out, the odds of a run scoring in the inning increase by 21.6%. If there's one out in the inning when he steals, the odds go up 25.6%, but if there are already two outs in the inning, the odds only increase 4%. No matter what, it plays into the team's favor to successfully steal third. But, as we know, not all steals are successful. Let's go back to the same scenario (runner on second only) and see what happens if the runner is thrown out.

If there are no outs and the runner is on second, the team has a 63.7% chance of scoring a run in the inning. If he attempts to steal and is caught, the resultant base/out state is bases empty, one out, in which case the team has a 17.2% chance of scoring. By being caught, he's just reduced his team's chances of scoring by 46.5% as opposed to just staying at second, already being in scoring position. If there's one out in the inning, the same attempt is made and the runner is thrown out, he's reduced is team's chance of scoring a run in the inning by 34.4%, and if he attempts it unsuccessfully with two outs, he's reduced his team's chance of scoring by 23%.

If you're counting along with me, those are some pretty big swings in run-scoring odds. In all cases, the cost of getting thrown out significantly outweighs the gain of stealing successfully. Teams know this and have known it for over 100 years, so this is not exactly new. On the season, only 13.2% of all stolen base attempts have been attempted steals of third, because stealing third is risky (and because it's easier to get a runner on first than it is to get a runner on second in the first place). When teams have run this season, they've picked their spots, maximizing their chances. The caught-stealing rate on steals of second is over 37% while the rate of attempted runners stealing third being thrown out is only 31%. There may be some noise in there, like runners being picked off, but we have an idea that teams tend to be smart when attempting to swipe third. Check out the team-by-team breakdown below.

Some teams have been good at stealing third and some have not. Of course, some teams just have better base running personnel, so that's a factor, too. Speaking of base running personnel, let's see who's had the most success stealing third this season.

And this brings us back to Mr. Cespedes, the third base swiping machine. Okay, okay, he's only done it four times over the course of the whole season, so maybe I'm making a bit too much out of this, but it's interesting that Cespedes has only stolen four bases all year and they've all been steals of third. He's attempted six steals on the season and five of them have been running from second to third (he was thrown out once stealing second). When Yoenis Cespedes is on second, pitchers may want to keep an eye on him because, although he doesn't appear a threat on the bases, he may just catch them sleeping and take third base.

In terms of the raw numbers of stolen third bases, two familiar names stand at the top. Billy Hamilton and Dee Gordon are just good at stealing bases, period. They're fast and it's a big part of their value as major leaguers. If they didn't run, their teams might not be able to justify their spot on the lineup card (especially in Hamilton's case). Jonathan Villar has been especially risky, as have Nori Aoki and Alexi Amarista, when compared to the average base stealer. But by and large, runners just don't do this very often, and we can see why from the run expectancy examples: it's difficult and often just not worth it.

Getting caught trying to steal third significantly outweighs just staying at second base in every case. If you run, you'd better make it. We expect the Dee Gordons and Billy Hamiltons of the world to try to pull this off because, well, they're good base stealers. On the other hand, Yoenis Cespedes is not a good base stealer. Unless he's stealing third.

Editor's Note: SB Nation's partner FanDuel is hosting a one-day $18,000 fantasy baseball league for tonight's MLB games. It's $2 to join and first prize is $2,000. Jump in now. Here's the FanDuel link.

*Base/Out info from TangoTiger.net.

*Base running data from Baseball-Reference.com.

...

Jeff Wiser is a featured writer at Beyond the Box Score and co-author of Inside the 'Zona, an analytical look at the Arizona Diamondbacks. You can find his work on craft beer at BeerGraphs and follow him on Twitter @OutfieldGrass24.