The 2015 MLB playoffs are continuing their scheduled march toward the World Series summit in the usual fashion, but the excitement surrounding playoff baseball has not ascended in quite such a steady fashion. Unless you actually are a fan of a team headed to the World Series, you might feel like the playoffs to this point have been a bit of a letdown, at least recently. The wild card games were a fun first act, and the various divisional series were excellent unto themselves and portended great things to come. Was that the peak, though? Were the semifinals, the league championship series, in their dual one-sidedness, a letdown by comparison?
One way to measure a baseball game’s excitement is by tracking leverage. Intuitively, games with more important moments, in which a single play can swing the game’s outcome, are more exciting than games with fewer such moments. Quantitatively, analysts have described the importance of a particular moment in a game in terms of leverage. FanGraphs has a Leverage Index (“LI”), which tracks the leverage of a given play according to the score, inning, number of outs, and number of runners on base at the time. Interpreting the LI calculation for a particular play is simple: an LI of 1 is neutral, while an LI of 2 is considered a high leverage situation.
To create a general measurement of the degree of excitement of each of the first three rounds of playoff games completed to date, I grouped all of the plays from the FanGraphs Play Logs for each game from each round — wild card, LDS, and LCS — according to their LI, and graphed the percentage of plays in each round that were high-leverage (LI>2), above-average leverage (two buckets, 1.5 < LI < 2 and 1 < LI < 1.5), and low leverage (two more buckets, 0.5 < LI < 1 and LI < 0.5).*
First, for a rough look, here’s a round-by-round comparison of all plays in each round separated into just two categories: those with neutral leverage or less, and those with higher than neutral leverage.
More than a third of the action in the divisional round games came in situations of heightened leverage (i.e., LI > 1), far more than the games in either of the other two rounds.
Next is a more detailed breakout, focusing on the distribution of heightened leverage situations across each playoff round and noting the highest leverage moment (“Peak LI”) recorded in each round.
Finally, here’s a table showing the data as tabulated, including the raw number of plays for each series and situation.
The results are clear: the divisional-round games, particularly the early ones (and especially the Blue Jays’ fourteen-inning loss to the Rangers), featured more situations of heightened leverage, as a percentage of all game plays in the round, than the wild card games or the LCS games, while the LCS round exhibited both the highest percentage of the lowest leverage situations of any round and offered high-leverage situations about half as often as did the LDS games. Four LCS games contained no high-leverage plays at all. Also telling: the highest leverage moment recorded in each round: 2.88 (WC), 5.20 (LDS), and 3.67 (LCS). In other words, in terms of excitement, these playoffs may already have peaked.
This certainly is a coarse-grained look at the excitement of watching the games in each round of this year’s playoffs. Graphing the standard deviation of the LI for these plays likely would show, in even more demonstrative fashion, the high degree of exciting moments in the LDS games, as compared with those in the other two rounds.
Here’s hoping for a contentious, exciting World Series!
*All data is through October 21, 2015, meaning that Toronto still could stage a major comeback in the ALCS beginning tonight, possibly requiring a revisiting of this analysis.
More of AD’s work may be found at ALDLAND.
Next post: BttP Podcast: Ep 37 – Matt Jackson & Seth RubinPrevious post: BttP Podcast: Ep 36 – Ben Suissa & Tyler Baber
Rob Mains
Nice research, AD. It’s good to see LI match our gut instinct: Pretty close DS, followed by CS that had some real snoozers. If I’m reading this correctly, your numbers do include Wednesday’s games but not Friday’s. If that’s right, and I did the math right, the breakdown for the five buckets are 45%, 29%, 15%, 5%, and 6% after last night’s game. So the CS didn’t match the DS, which isn’t a surprise, I suppose. But last night’s game did give us the highest LI moment of the postseason: Josh Donaldson’s game-ending, series-ending at bat in a one-run game with runners on second and third and two out had an LI of 6.84.
AD
Thanks, Rob. And yes, exactly my thoughts regarding the Royals’ clinching game. I may explore this in a separate post, but your comment hits the core of it. Onto the World Series!