Making the Most of What You’ve Got
A little while ago, I wrote something about how teams score runs by getting runners in scoring position and subsequently knocking them in. The gist of the article was this graph, which illustrates how many at bats each team had with runners in scoring position, and how well they batted in those situations. This version of the graph is updated for the end of the year stats:
In the meantime, people continue to discuss how teams score runs, and even Bill James has updated his Runs Created formula for the umpteenth time, to increase its efficiency.
The discussion of how runs are really created might well go on forever, because it’s a fascinating subject with as many angles as an M.C. Escher print. While I’m not the type to construct the ultimate Runs Created formula, I do like to measure some of the angles along the way. Call me a human baseball protractor. And the angle I’ve been exploring lately is whether teams line up their offensive capabilities as well as they can.
Everyone knows OBP (On Base Percentage) is good, right? Getting on base leads to moving into scoring position which leads to runs being scored. But getting on base early in an inning is much more powerful than getting on base later. According to Tangotiger’s Run Expectancy Chart (based on stats from 1999 through 2002), a team with a runner on first and none out can expect to score .953 runs, vs. .251 runs when a runner is on first with two outs. That is a big difference and, as far as I know, no Runs Created formula captures that information.
So here is a list of each team’s 2004 OBP, compared to its OBP leading off an inning. You would think these two OBP’s would be roughly equal over a full year, but there are some significant team differences. And you’ll notice that teams tended to have lower OBPs when leading off an inning. For the record, the overall AL OBP was .337 vs. .331 when leading off an inning; in the NL, the numbers were .333/.325. I think this is due to a number of causes, such as the pitcher not having to pitch from the stretch.
Anyway, take a look, and we’ll discuss the results below.
Leadoff Team OBP OBP Diff STL .344 .359 .015 SDP .342 .354 .012 MIN .332 .343 .011 SEA .331 .341 .010 TBD .320 .329 .009 MIL .321 .328 .007 LAD .332 .335 .003 PIT .321 .324 .003 BAL .345 .347 .002 FLO .329 .330 .001 ANA .341 .340 -.001 NYM .317 .315 -.002 CHC .328 .325 -.003 ARI .310 .305 -.005 KC .322 .316 -.006 TOR .328 .322 -.006 BOS .360 .353 -.007 DET .337 .328 -.009 CHW .333 .322 -.011 CLE .351 .339 -.012 HOU .342 .329 -.013 COL .345 .332 -.013 NYY .353 .338 -.015 SFG .357 .339 -.018 CIN .331 .312 -.019 OAK .343 .323 -.020 ATL .343 .318 -.025 PHI .345 .318 -.027 TEX .329 .300 -.029 MON .313 .282 -.031
Man, the Expos not only had just about the worst OBP in either league, they were particularly bad when leading off an inning. Good thing Jim Bowden signed Cristian Guzman! On the other hand, some of the best offensive teams in either league — Texas, Philadelphia, Atlanta and Oakland — were right on the Expos’ tail. A couple of these teams didn’t have particularly good leadoff batters, and a couple (PHI, ATL) also had singularly good OBP batters who batted third in the lineup. #3 batters are actually less likely to lead off an inning than any other lineup slot.
The Giants are also near the bottom of this table, which is what happens when your best hitter sets the Major League record for intentional walks and OBP.
On the other hand, the list is led by the St. Louis Cardinals. The Cardinals had three guys with OBP’s over .400 smack dab in the middle of their lineup. This sort of OBP excellence, spread over several consecutive lineup spots, makes it more likely that one of those guys will lead off an inning. Another way to go is the Minnesota/Seattle path, in which you clearly put your best OBP man first. Or, you can go at this the way of the Padres, by having at least a .330 OBP in every spot of the lineup except the pitcher’s.
Overall, the best simple determinant of how often teams get runners into scoring position is still straight OBP, but leadoff OBP can improve the accuracy of your model a bit.
How about the other side of the equation, batting with runners in scoring position? When we look at Batting Average with Runners in Scoring Position (BA/RISP), are we seeing batters “rise to the occasion?” Or are we observing the residue of design, with the team’s best batters coming up most often with RISP?
Well, I compared each team’s overall Batting Average to the Batting Average of the players who accounted for roughly the top 50% of the team’s At Bats with Runners in Scoring Position. That sounds complicated, but all I really did was calculate the relative strength of the batters who came to bat with RISP most often for each team. I didn’t analyze their actual performance with RISP; I looked at whether the best hitters (highest overall BA) were the ones most likely to be at bat when it counted.
Most teams had about five batters account for 50% of their team’s At Bats with RISP. And virtually every team managed to have their best hitters up with RISP. But some teams were better at it than others:
Team BA Top5BA Diff ARI .253 .284 .031 SEA .270 .299 .029 SDP .273 .301 .028 CIN .250 .277 .027 LAD .262 .286 .024 ANA .282 .305 .023 STL .278 .300 .022 TEX .266 .287 .021 BAL .281 .302 .021 COL .275 .294 .020 PHI .267 .286 .019 HOU .267 .286 .019 NYY .268 .286 .018 PIT .260 .278 .017 DET .272 .288 .016 MON .249 .263 .015 CHC .268 .282 .014 TBD .258 .272 .014 BOS .282 .295 .013 FLO .264 .275 .011 ATL .270 .281 .011 KC .259 .269 .011 MIN .266 .275 .009 MIL .248 .257 .009 CLE .276 .284 .008 OAK .270 .278 .008 TOR .260 .265 .005 SFG .270 .275 .005 CHW .268 .272 .004 NYM .249 .252 .003
I figured we would see the NL teams at the top of this list, what with the pitcher batting and all. Well, really, just with the pitcher batting. But the top two teams are an interesting contrast. The Diamondbacks did have their pitchers bat and a lot of other lousy hitters to boot. But they also had a few good hitters, who accounted for 50% of scoring position at bats, and these hitters stood out against the rest of the crew. Remember, this is a measure of relative batting skills.
On the other hand, Seattle surprisingly tied for the top of the list, thanks again to Ichiro. In the AL, the leadoff batter is more likely to bat with runners in scoring position than in the NL, making Ichiro the ideal AL leadoff batter (high BA/high OBP).
The Mets are at the bottom of the list, which is what happens when a .231 hitter (Mike Cameron) has the most at bats with RISP on the team. But if you’ve been following along, you’ve noticed the huge surprise at the bottom of this list: the Chicago White Sox!!
Go back to the top of the page and check out the White Sox’s position on the graph. Yes, the White Sox batted .291 with RISP, which was SECOND IN THE MAJOR LEAGUES. But they were the second-worst team at getting their best hitters to the plate with RISP. As opposed to, say, the Angels, who managed a high BA/RISP thanks to getting their best hitters up to bat, the White Sox did the opposite. Their batters quite simply rose to the occasion when it counted.
Here’s a list of the ten batters who batted most frequently with runners in scoring position for the Sox:
Player AB/RISP Total BA BA/RISP Diff P Konerko 140 .277 .314 .037 J Crede 122 .239 .238 -.001 A Rowand 107 .310 .290 -.020 J Valentin 105 .216 .248 .032 J Uribe 99 .283 .323 .040 W Harris 76 .262 .276 .015 T Perez 68 .246 .397 .151 F Thomas 66 .271 .273 .002 R Gload 58 .321 .397 .076 M Ordonez 54 .292 .352 .060
All but two of these guys batted better with runners in scoring position — some batted significantly better. You absolutely have to give the Sox hitters credit for this surge in clutch hitting. But I can almost guarantee you that it will not happen next year. If Ken Williams is constructing his 2005 Sox based on 2004’s runs scored output, he’s going to be sorely disappointed.
References & Resources
If you want to get into some real Run Creation detail, you should read Tangotiger’s seminal study “How Runs are Really Created.”