In yesterday’s primer, I introduced the basics behind World Series win probabilities. Today, we begin looking at the 2011 World Series through this lens.
The first natural use of win probabilities, especially on FanGraphs, is the win probability graph. Observe, the graph of series win probability for the St. Louis Cardinals throughout the World Series:
Click to view the graph in a new window with all the bells and whistles provided by the Tableau software.
The graph provides simply the series win probability by play for every single play of the 2011 World Series. The width of the line represents the single game win probability for the Cardinals at the time. From this you can see that in Games Two and Five, the Cardinals had excellent chances to win before eventually losing, and, of course, they were nearly dead in Game Six before making their legendary comeback.
Specifically, the Cardinals peaked early at 77% with a 1-0 lead in the ninth inning of Game Two and could have gained a strong hold of the series if not for a comeback by the Rangers in the top of the inning. The Cardinals would win their second game behind Albert Pujols‘s three-homer game, but the Rangers gained control behind Derek Holland and Mike Napoli in Games Four and Five respectively.
And then, of course, Game Six. It can be easy to forget the Cardinals actually had an excellent chance to win this game without needing David Freese‘s heroics, holding a 71% single-game win probability and a 35% series win probability before Matt Holliday was picked off with the bases loaded in the sixth. That play allowed the Rangers to escape, and just like that the Rangers were up 6-4 thanks to back-to-back home runs by Nelson Cruz and Mike Napoli.
The Cardinals trailed 7-5 in the ninth inning and their series win probability bottomed out at 2% after Ryan Theriot struck out for the second out of the inning. Here, we see the biggest fluctuations of the entire series. Freese’s game-tying triple skyrocketed the Cardinals up to 31.2%. Then Josh Hamilton‘s home run in the 10th sunk the Cardinals back down to 3.9%. Then Lance Berkman‘s game-tying single in the bottom of the inning put the Cardinals back up to 31.7%. Finally, Freese’s home run off Mark Lowe finished the game, giving the Cardinals a 50-50 shot at Game Seven.
Game Seven is quite simple through the lens of series probability added, as Game Seven’s game graph tells the exact same story. The Rangers drew first blood and the Cardinals win probability dropped to 29%, but they would come back in the bottom of the first and never look back, eventually winning 6-2 and taking the Commissioner’s Trophy.
For the specifics of each play, you can mouse over the line and a box with the series win probability as well as inning, out, score, and description of the play will show up. To see more on each individual play, check out this alternative view:
In this visualization, every single play of the World Series is presented, sorted initially by highest batter series probability added (noted as SPA on the graph). Unsurprisingly, most of the top plays come from Games Six and Seven, but a few especially notable plays from earlier games, such as Elvis Andrus‘s single to spark the Rangers’ rally in the top of the ninth of Game Two, rank highly as well.
One of the first things you will probably notice on this second graph is the presence of David Freese — three of the top five plays by series probability added came off Freese’s bat. Freese took home the World Series MVP award, and by this measure, it was certainly earned.
Again, if you click on the image it will open up the Tableau visualization, and you can get extra details by mousing over the bars. You can scroll through all 500+ plays of the series see how each play impacted the World Series, whether it was hugely important or not at all.
The next step will be to take a look at the full contributions of each player and pitcher. Tomorrow, I will present some looks at the hitters who shaped the World Series, for better or for worse.