A Comparison Between the Wild Card Games

Do you guys know Jaack? That’s not a very good introduction. We’ve run live game chats during the wild-card playoffs the previous two nights, and Jaack is the screen name of at least one participating commenter. This is Wednesday’s live chat, and this is Jaack, at 11:05pm EDT:

Comment From Jaack
There needs to be an article about how much better last night’s game was. Like inning by inning breakdown.

Jaack’s the best. Thank you, Jaack! Following is such a breakdown.

Tuesday’s game, of course, set an impossible standard. It feels like it’s an all-time kind of playoff game, and while I’m fully aware of recency bias, I felt the same way after Rangers/Cardinals Game 6 in 2011, and that one’s stood up. We can debate how amazing it was to watch the A’s and the Royals, but there’s no debate that it was some kind of amazing. So there was no chance at all that the Giants and Pirates would follow that show with at least an equivalent show of their own, but Wednesday was a total dud. The saving grace was that Madison Bumgarner pitched and was awesome, but for the most part he was awesome without any danger, and when the outcome feels decided, the entertainment value plummets.

This isn’t about the quality of the baseball. This is about the quality of the show. We already know that Tuesday’s show was several times better, but now let’s put some actual data to it because what else do you have to do for the next few minutes? If you started reading this, you can finish reading this.

This is going to deal with our win probability data. You probably know what that’s all about. We’re able to determine approximate winning odds for each team for any given game state. Win probability also informs a statistic known as Leverage Index, and while I could give you the actual mathematical definition, let’s leave it at this: Leverage Index (LI) is basically a measure of how stressful and important a situation is. When a closer is trying to nail down a one-run save, the LI is high. When a middle infielder is pitching, the LI is low. LI measures the potential win-expectancy swing from the range of potential plate-appearance outcomes.

Our win-expectancy plots also feature LI plots. Here are the LI plots for the two wild-card games, and I won’t tell you which is which, because I don’t need to:

One of them changes color. One of them disappears! If you’re a visual learner, that’s an astonishing image. Now here are the win-expectancy plots, one pasted over the other:

The x-axis, of course, is time, and while one of these games went more innings than the other, you get a real good sense of the flow. The American League game was a roller coaster. The National League game was a level walk, with a step up one step in the middle. It was quite the step; after the walk, you remember the step. You also remember thinking, outside of that step, that wasn’t much of a walk. Maybe the scenery was nice.

Now we should look at some actual numbers, instead of a visual representation of numbers. Okay, so, the average event in the AL game swung the win probability by 5.2%. The median was 2.8%. The average event in the NL game swung the win probability by 1.6%. The median was 0.7%. And, after the fourth inning? After four, the Royals were up 3-2, and the Giants were up 4-0. From the fifth inning on, the average event in the AL game swung the win probability by 5.8%, with a median of 3.1%. In the NL, the numbers were 0.4% and 0.3%. Another way of saying that is, after the fourth inning, in the NL game, nothing really happened. The Giants doubled their run output, but it didn’t do anything for the viewers. Even Giants fans might’ve felt the beatdown got excessive, and Pirates fans never got anything to cheer about.

Switch on over to Leverage Index. Similar approach! The average event in the AL game had an LI of 1.66, with a median of 1.12. The average event in the NL game had an LI of 0.50, with a median of 0.21. And, after the fourth, the average event in the AL game had an LI of 1.93, and a median of 1.26. The average event in the NL game had an LI of 0.13, and a median of 0.09. Have I already noted that an LI of 1.00 is average? An LI of 1.00 is average. The AL game, start to finish, was 66% more stressful than an ordinary game, and the NL game was half as stressful. And after the first few innings were over, the difference was far more pronounced. Which you knew, if you watched, but now you have sweet sweet numbers on your side.

For the sake of comparison, let’s look at some pitchers. As noted, the average AL game event had an LI of 1.66. That matches the average event LI this season for Aroldis Chapman, Jenrry Mejia, and Rafael Soriano. The NL game matched the average event LI this season for Seth Rosin, one of the Jose Ramirezes, and one of the David Carpenters. From the fifth inning on, the AL game matched the average event LI this season for Kenley Jansen and Glen Perkins. The NL game matched the average event LI this season for a Fornataro, a Campos, an Adams, and a Scribner. Position players tend to pitch with the LI at 0.00 or 0.01. The NL game wasn’t that low, but it did eventually get there. And we’re looking at several innings, instead of a couple plate appearances or so.

As the final and maybe most effective way to put this over the top: in the American League wild-card playoff, there were 19 separate game events that swung the win probability by at least 10%. In the National League wild-card playoff, there was one such event. You know exactly what it is — it was the one thing that happened. Baseball’s gotten to be more and more of a regional game, but the AL game captured the attention of the whole country. Impartial observers came away thinking, that was spectacular. The NL game was the opposite side of baseball, the side where nine innings still take as long as a usual nine innings, but there’s little point in watching past the first few. One of these games, people will talk about for years. Decades, in some areas. The other’s just a hell of an achievement for Madison Bumgarner, who’s already been outstanding on a far bigger stage.

We’ve had two playoff games. One of those games was Baseball! The other of the games was baseball. On average, then, we’ve had pretty decent theater.