The Importance of Sequencing by Dave Cameron April 25, 2013 The St. Louis Cardinals just finished off a sweep of the Washington Nationals, and currently stand at the top of the NL Central with a 13-8 record. At 4.9 runs per game, they’re fourth in the NL in offensive production, which is one of the primary reasons they have outscored their opponents by 27 runs so far. (The other main reason is Adam Wainwright.) Now, though, here’s a fun fact you might not expect; the St. Louis Cardinals currently have a wRC+ of 88, tied — with the Mariners! — for 24th in MLB. Even if we remove pitcher hitting from the group and only look at offense from position players, their wRC+ is 97, still below the Major League average. From just looking at their batting lines, it’d be fair to describe the Cardinals offensive performance to date as sub-par, and yet, they’re scoring nearly five runs per game. This is the power of situational hitting. Behold, the Cardinals offense, by baserunner state: Situation PA BA OBP SLG wOBA wRC+ Rank Bases Empty 460 0.199 0.254 0.308 0.252 59 30th Men On 328 0.317 0.377 0.466 0.364 132 2nd RISP 201 0.349 0.413 0.477 0.383 149 2nd The Cardinals are the Marlins with the bases empty and the ’27 Yankees on steroids with men on base. When you count up how many hits the Cardinals have gotten, it’s not that impressive, but when you add in the sequence of when those hits took place, their offense starts to look a lot more potent. There’s actually a pretty easy way to see just how big of a difference sequential hitting has made for each team using two stats here on FanGraphs: Batting Runs and RE24. Batting Runs is the hitting component of WAR, so it’s basically park and league adjusted wOBA multiplied by the number of plate appearances, with the result being the total run value for the player’s outcomes as a batter. It is designed to be context neutral, and not take the situation into account, so it treats a grand slam and a solo home run as equals, giving both the same credit by simply marking them as a home run. Since the hitter didn’t put those guys on base, he gets credit for what he did himself, and not the performance of his teammates. RE24 gives you a comparable result in terms of batting runs above or below average, but does so by calculating the run expectancy changes between when a batter steps into the box and when his at-bat ends. This serves to give the hitter credit for the difference between the expected runs scored in that situation and how many runs actually score. Essentially, RE24 is Batting Runs adjusted for how well a player or team has sequenced their hits for maximum efficiency. It’s not fully context dependent, as it doesn’t include the score or inning, but it does adjust performance for the situations faced throughout the season. By using the custom leaderboards here on the site, you can easily line up Batting Runs and RE24 side by side, and see how many extra runs a team has created (or lost) through sequencing. Here’s a leaderboard with Batting Runs added to the WPA section, so Batting and RE24 are next to each other on the same page. If you don’t want to click through, however, I’ll make it easy for you and simply reproduce the table below. Team Bat RE24 Difference Cardinals (11.4) 13.2 24.6 Reds 7.1 23.7 16.6 Athletics 21.3 36.8 15.5 Mets 11.7 24.9 13.2 Pirates (12.4) (1.1) 11.4 Giants (1.3) 10.0 11.3 Rangers 1.7 10.2 8.5 Tigers (1.2) 4.7 5.9 Orioles 1.8 6.4 4.6 Marlins (36.7) (35.0) 1.7 Astros 5.2 6.5 1.3 Red Sox – 0.8 0.8 Brewers (6.7) (6.2) 0.5 Diamondbacks (9.7) (9.9) (0.2) Yankees 4.1 3.5 (0.6) Phillies (11.7) (13.1) (1.4) Nationals (7.0) (8.8) (1.8) Rays (6.0) (8.0) (2.0) Royals (5.3) (8.2) (2.9) Twins (3.2) (7.0) (3.8) Cubs (16.7) (22.3) (5.6) White Sox (21.6) (27.5) (5.9) Rockies 11.0 4.7 (6.3) Indians 9.5 2.3 (7.2) Mariners (11.5) (18.7) (7.2) Padres (9.9) (17.7) (7.8) Braves 12.6 4.4 (8.2) Blue Jays (10.2) (21.3) (11.1) Angels 10.0 (2.0) (12.0) Dodgers (1.2) (15.3) (14.1) The Cardinals have scored 24.6 more runs than we’d have expected based on their context neutral performance, simply through the act of bunching their hits together. That’s over one run per game in sequencing. On the other side of the spectrum, we see that the Angels, Dodgers, and Blue Jays have all lost double digit runs due to their situational hitting, and not surprisingly, they’re currently the three biggest disappointments in terms of record versus expectations. Side note – how about the Braves coming in at minus eight runs here? The team that entered play with the best record in baseball has actually underperformed their raw offensive totals. That’s amazing. Okay, back to the topic at hand. Now, this is the kind of gap that could be accurately described as unsustainable. You shouldn’t look at that chart and conclude that the Cardinals are full of clutch warriors while the Dodgers are pansies who melt under pressure. These numbers are going to get closer together as the season goes along, and teams who are at the extremes of the spectrum are going to get pulled back towards the middle over the course of 162 games. However, not everything is going to completely even out. Last year, the Braves were +25 in difference between RE24 and Batting Runs, while the Royals were -13. That’s a nearly 40 run swing from top to bottom, and which is essentially equal to having another All-Star hitter in your line-up. Sequencing can be a big factor in a team’s overall record, especially early in the season when things haven’t had time to regress to the mean yet. And this is one of the reasons why I’ll continue to caution against using pythagorean win-loss records as some kind of stand-in for “true team performance”, especially early on in the year. Pythag record is built on the notion of stripping out the sequencing of when runs are scored and allowed, on the belief that the distribution of runs is mostly random, which has historically been shown to be true. However, the same concept applies to the actual scoring and allowing of runs. The Cardinals +27 run differential in 21 games means that their pythag winning percentage is .637. However, we know that 25 of those runs are due to their offensive sequencing, so if their distribution of hits had been more normal, they’d have a run differential of +2, and their pythag winning percentage would be .512. If the Cardinals had a .512 pythag and a .637 winning percentage, you’d probably see some comments about how they’re headed for regression. Because their .637 pythag is actually better than their .619 winning percentage, you probably won’t see anyone write about the Cardinals over performing in the first few weeks of the season. At the same time, though, we don’t think the Cardinals are really the most clutch offensive team in baseball. We expect their situational hitting will regress as the year goes on. At the same time, we expect that Allan Craig, Jon Jay, and David Freese will hit better than they have so far, so there’s positive offensive regression coming for the Cardinals too. Those might end up offsetting, and who knows, St. Louis may very well average 4.9 runs per game all year long. But, it’s worth remembering that pythagorean record isn’t really context neutral, and quoting it as a true talent measure at this point in the season isn’t really accurate either. There’s a lot of noise in both runs scored and runs allowed too, especially in small sample sizes. Let’s not solely focus on regressing one aspect of performance while holding all the rest steady.