Subscribe to Sports Illustrated and Save Over 82% Subscribe to SI Give the Gift of SI
Keep up with the latest news, notes and developments with Fungoes, a daily journal for all things baseball that will last all season long.
4/10/2007 03:56:00 PM

### FanGraphs: Better than a Box Score

By David Appelman

The Mets' home opener against the Phillies was a nail-biter up until the Mets' seven-run explosion in the bottom of the eighth. If you were to take a look at a regular box score, you might dismiss the game as a blowout, but a single glance at a Win Probability graph and it couldn't be clearer that this was a very exciting game:

 FanGraphs.com

Through the use of historical play-by-play data and some additional math, Win Probability tells us a team's chances of winning at any point in the game. The main chart tracks the Win Probability as the game progresses, play-by-play. The lower chart tracks something called Leverage Index, which is the potential importance of each play. The higher the Leverage Index is, the more important the play.

So let's take a look at some of the more exciting moments of the game from a Win Probability perspective:

• Top of the fourth, no outs, bases loaded, 1-0 Phillies. The Mets find themselves with a 23 percent chance of winning. Cole Hamels swings at John Maine's first pitch, the Mets get the force-out at home, and the their chances of winning rise to 29 percent. Next up is Jimmy Rollins and he grounds into a double play. The Mets have avoided a real disaster and find themselves with a 43 percent chance to win going into the bottom of the fourth.
• Bottom of the fourth, two outs, bases loaded, 1-0 Phillies. With the Mets having a 46 percent chance of winning, Jose Valentin singles off of Hamels, scoring two runs and bringing the Mets' chances all the way up to 67 percent!
• Top of the sixth, two outs, runners on second and third, 3-2 Mets. Ryan Howard picks the perfect time to hit his first home run of the season, putting the Mets' chances of winning at just 25 percent.
• Bottom of the eighth, one out, runners on first and second, 5-4 Phillies. With a 55 percent chance of winning and the game on the line, Jose Reyes hits a double-play ball to Rollins, but ... Rollins drops the ball and everyone's safe. The Mets now lead 6-5 and have a 77 percent chance of winning the game, their highest Win Probability thus far.

From that point on, the Mets steamrolled their way to victory, ending the eighth with a six-run lead. They had a 99.7 percent chance to win at that point and the game was effectively over.

While I'm sure you found my play-by-play account utterly riveting, the graph does an excellent job of showing the entire game flow, not just the score. The score alone would not have told you the Phillies blew a great opportunity in the fourth, nor would it have told you Geoff Geary's wild pitch more or less sealed the Mets' victory.

But that's not all Win Probability is good for. You can calculate which players were the most valuable or least valuable to their team over the course of the game. This is done by determining the difference in the Win Probability at the start of the play and the end of the play, and then attributing it to the batter and pitcher. This is called Win Probability Added (WPA). (Click here for the WPA-enhanced box score.)

Using WPA we know that Geary pretty much blew the game for the Phillies by decreasing their chances of winning by 50 percent while Howard did all he could to get the Phillies to win, increasing their chances by 32 percent. No pitcher on the Mets was particularly effective, but Carlos Delgado and Jose Reyes were the real heroes, increasing the Mets chances of winning by 32 percent and 27 percent, respectively.

To see more Win Probability or learn more about it, here are a few places to get you started:

• FanGraphs produces live Win Probability charts for every game so you can track the chances of your team winning as the game progresses. It also offers WPA stats on a game-by-game and full season basis.
• For more information about Win Probability and Leverage Index, The Hardball Times has a few great articles: The One About Win Probability and The Secret Recipes of Leverage Index.
• Tom Tango provided the Win Probability data behind the graphs and is co-author of The Book: Playing the Percentages in Baseball which examines a number of common baseball strategies using Win Probability and other methods. The Book's blog at insidethebook.com is always a tremendous resource.

David Appelman is the creator of FanGraphs.com.

Labels: