PHOENIX, AZ - MAY 21: Justin Upton #10 of the Arizona Diamondbacks is tagged out by infielder Justin Sellers #12 of the Los Angeles Dodgers as he is caught stealing during the sixth inning of the MLB game at Chase Field on May 21, 2012 in Phoenix, Arizona. (Photo by Christian Petersen/Getty Images)
"We're an aggressive team and we're going to be aggressive... In the end, we hope to tighten it up. We'll get into some games down the road that become more meaningful, (and) we'll be much sharper at it. That's the goal.:"
-- Kirk GIbson
Going by last night's contest, that goal does not seem to have been achieved. In the first inning, Adam Eaton was caught trying to take third base, on the front end of a double-steal - a crucial out in an inning where the Diamondbacks had three walks plus a hit, but failed to score at all. It was the team's 46th time caught stealing this season, just one behind the league-leading Pirates, and their success rate, of 64%, is also second-worst in the NL. But stolen-bases are just one aspect of base-running; so how do the Diamondbacks stack up overall?
Find out after the jump. Quick answer. It's not pretty...
As raw data we use the NL Baserunning page on Baseball-Reference.com. There are a lot of columns present for each team there, so let's start by explaining the meanings of the ones we'll use going forward. SB, CS and SB% should not need any explanation. PO are straight pick-offs - where the pitcher throws over, and the runner is tagged trying to get back. POCS are pick-offs caught stealing, where the pick-off throw leads the runner to try and advance, and he's tagged out at the next base. Those count both here and in CS. OOB are other Outs On Basepaths: not pick-offs or caught stealing, but outs when trying to take an extra base, on a sacrifice fly, etc.
The columns beginning with XBT% refer to that extra base, for example going from first to third on a single. 1stS, for instance, is the number of times a runner was on first when a single was hit. 1stS2 is how often thy ended up on second; 1stS3 is how often they were on third. The other columns are similar: 1stD refers to being on first when a double was hit, where you can end up on third or home; 2ndS is being on second when a single is hit. You should note that 1stS2 + 1stS3 will typically be slightly less than 1stS, because neither include the times a runner tries to reach third, but makes an out. XBT% is the overall percentage, across all cases, where an extra base is taken.
What we can do, to obtain an overall picture of base-running competence, is combine the base-stealing and base-running stats, to give us two numbers. One is for all the bad things which can happen, which we'll call TOOTBLAN - as you know, this stands for Total Outs On Basepaths, League Average Normalized. :) This is worked out as CS + PO - POCS (because those are the pickoffs that have already been counted in CS, remember) + OOB. For the good things, we simply add together SB plus XBT, the number of extra bases taken. Let's call this WATBLAR - Whizzing Around The Bases Like a Rocket.
FInally, we can compare TOOTBLAN and WATBLAR for each team, against the league average figure, and see how many bases were gained and outs given up, compared to that league average. Add the two differences (low TOOTBLAN good, high WATBLAR good) together and you get an overall measure of how good or bad a team was at base-running, compared to the rest of the franchises. For example, we have 11 more TOOTBLAN than league average, and a dozen less WATBLAR, giving us a score of -23. Here's the table which results - with the final column that overall measure, for which I still haven't got a name. Suggestions are welcome.
As you can see, it's not very good for the Diamondbacks, who are ranked 13th overall. Still. there are worse out there. I mean, look at the Pirates, twice as bad as the D-backs, all the way down at -46. Hahaha! But hang on. Because not every team has had the same opportunities to steal a bag or take an extra base. The three clubs below Arizona have a season on-base percentage of .301, .303 and .304 respectively, compared to our .325. So it's no surprise they've WATBLAR'd less often - they're simply on-base less often.
Fortunately, the data tells us the number of stolen-base opportunities (SBO) and extra-base opportunities (XBO) each team has had. This allows us to work out an expected number of stolen bases, pick-offs and extra bases for a specific club, based on the chances they had. From there, we can work out expected values for TOOTBLAN and WATBLAR, and use those as the comparison mark for their actual performance, rather than just league average. This makes a huge difference. Expected WATBLAR for the Pirates, for instance, is only 264; so their actual number of 246, while still poor, is a lot closer to that than league average (291). Here's the chart with these new stats.
Yuk. When you compare bases gained and outs given up, to the projected number based on opportunity, things look even worse for the Diamondbacks. While their expected TOOTBLAN is unchanged, they should have half a dozen more extra bases, based on the chances they've had. That, combined with improvements for the teams below them, due to their target baselines dropping, is enough to drop the Diamondbacks into last spot in the National League.By this measure, we are indeed the worst team on the basepaths in the NL.
There are already existing metrics for this, such as UBR from Fangraphs (shown as BsR on its charts). This paints a radically different picture, listing the Pirates as the best base-running team in the league, with the D-backs slightly above average and the Nationals the worst. However, that measure excludes stolen bases and caught stealing (which are instead folded into another number, wOBA). A closer parallel may be Baseball Prospectus's Team Baserunning number, that does include thefts, successful and failed. Their numbers for 2012 show the Diamondbacks ranked 29th of 30 in the majors, ahead of the Nationals. The Marlins lead there as well.
My study is by no means a definitive metric. For instance, it values an extra base and an out as the same, when the latter is probably more damaging [that's why your stolen-base percentage needs to be well over 50% in order to be acceptable]. You could also figure out the expected number of extra bases for a team better, based on the specific situations they have - it's easier to go second to home on a single than first to third. I'll probably leave that for now, but what I do plan to do at the end of the season, is break down the Diamondbacks' numbers to an individual level, and see who are the best WATBLARs and worst TOOTBLANs on the team.