April 9, 2010

Fantasy Focus

Stolen-base Redux

by Marc Normandin

Printer-friendly

Stolen bases are the most contentious statistic on the offensive side of things in fantasy baseball. Some owners will do crazy things for steals, like drafting players that hurt them in every other statistic. Unlike, say, a player who has a low batting average but hits home runs, there's no carryover effect that increases RBI or R (or OPS in some leagues)—just the stolen base, and maybe some runs, though that isn't guaranteed, either.

That's why last season I took a look at the trends for steals to see if it was worth it to do those sorts of things to your team, or if you were just causing more trouble than it was worth. My plan today is to update those results using 2009 numbers and check if anything has changed in a year's time.

The data begins in 1998, as it was the year the major leagues expanded to 30 teams, and ends with the most recent season:

Steals were up in 2009 by 171, with 98 additional foiled attempts. Attempts to steal were over 4,000 for the first time since 2002, though, and we'll take a look at that in more detail later. As you can see from the rest of the years, that kind of fluctuation in the numbers isn't foreign—from 2006 to 2007, they jumped 150 then fell 119 between 2007 and 2008, only to rise again last season. While the success rate dipped again, it's still higher than it was a few years back, and much higher than at the end of the 90s and beginning of the aughts.

Using the PECOTA weighted-mean spreadsheet (which is adjusted for playing time), we see 3,113 steals forecasted for 2010, with 1,246 caught stealings. I asked Clay Davenport, who formulates PECOTA if the projected increase from 2009 was a trend or did it have more to do with the nature of a league-wide forecast? Clay said that the increase is in line with expectations when you sum the majors together like I did for the purposes of this article. It's a little inflated over what it will actually be, but that's just the nature of projections. What this tells me is that PECOTA thinks a repeat of 2009 is more likely than another 2008 as far as steals.

One thing to remember is that this data is from the entire majors leagues. If you're in an NL- or AL-only league, that kind of detail may be necessary, but for those in mixed leagues, a much smaller percentage of the player pool is going to come off of the board. That's why we will take a look to see how much of the stolen bases from 2009 were wrapped up in the top 30 and top 60 players in that category:

The top 30 had a drop of over 3 percent, the largest of any of the years in this table. The top 60 had a similar drop, and now finds itself at 2004 levels. The conclusion after looking at this data prior to 2009 was that the best base stealers in the league were picking up more of the stolen bases and attempts as managers became smarter about sending the right guys in the right situations, basically an awareness of things like the rate at which steals become productive. Some of that seems to have disappeared, at least in 2009.

(Att% is Attempt Percentage) A larger percentage of attempts went to players outside of this group in 2009. In addition, while the top 30 picked up 30.1 percent of all steals, those players accounted for just 8.7 percent of all caught stealings. For the top 60, it's 14.1 percent. The top 30 stole at a 79.1-percent success rate, and the top 60 was just slightly lower at 78.5. You get a sense of how successful those outside of the top 60 were when you remember the major leagues as a whole were at a 72.3 percent clip.

So, there were more steals to go around outside of the top 60 than in the recent past, though, we're not talking about a whole lot of steals here, and they were spread out amongst the 411 players who stole a base last year. Player No. 100 on this list, Melky Cabrera, stolen 10 bases, so you can go deeper than the top 60 to pick up some steals. Melky was tied with six others (Nos. 94-100). The first player under 10 steals was Akinori Iwamura (and eight others) with nine thefts.

As for their actual performance, I'll continue what I did last year and use Marginal Lineup Value rate as well as its positionally adjusted form to see how the average performance stacked up against previous seasons:

Here is where things get interesting. The performance of the leaders in steals, for both the top 30 and top 60, jumped considerably from 2008, and is at its highest point since 1998-99. There are a few obvious reasons for this. Michael Bourn (.229/.288/.300 in 2008) improved his performance significantly (.285/.354/.384 in 2009), Carlos Gomez (.258/.296/.360 in 2008, ranked 15^th in steals) didn't even make the top 60 in 2009, Cesar Izturis similarly dropped from the top 60, Joey Gathright disappeared. Notice a trend here? Steals-only, no-power guys either dropped from the top 60 or improved their games. That may not be enough on its own to sway the average MLVr and PMLVr as much as we see above, but it's certainly a start.

So, better offensive players accounted for the top 30 and top 60 in steals in 2009, meaning there is less incentive for you to waste a roster spot on someone like Gomez, Gathright or Izturis just to pick up your own share of the base thievery. Whether or not this continues in 2010 is something we will see.

Maybe you're saddled with one of those types of players right now, wouldn't be such a bad thing if you found someone to make a trade. Or maybe you could pull the trigger on a free agent who can produce more, players outside of the top 60 are picking up a higher percentage of the major leagues' steals.