Live In-Game Win Probability 2.0

Live Win Probability graphs are back with tonight's season opener. This year, the graphs have more features, are easier to read, and update much quicker compared to the 2008 version. They feature the new format debuted earlier in the year for NBA and NHL.

For those of you new to the site, here is a primer on what you'll see. The graph charts the probability each team will win the game play by play. The closer the line gets to the top of the graph, the more likely the visiting team will win. The closer it gets to the bottom of the graph, the more likely the home team will win. Think of it as an NFL Richter scale or maybe a cardiogram of a 'football heart attack.'

The win probability is calculated based on 2,400 of the most recent NFL games since the 2000 season. It factors in possession, time, score, down, distance, and field position. Over the next few days, more features from last year will return. Those include play descriptions and additional advanced stats, such as 1st down probability, current drive scoring probabilities, and current expected points. Future features may also include real-time 4th down recommendations and a "leverage index," which indicates how critical the current play is.

The WP graph can be a useful tool to help understand the flow and rhythm of the game. It tells us which plays and situations really contributed to the outcome. It can estimate how costly a blown call by the refs was, or how badly a critical penalty hurt a team. It tells us exactly where the teams stand in terms of their chances of winning at each point in the game. Often, the game is out of reach far before the TV announcers and analysts will tell you. It also tells us which games were really competitive and which games were the biggest comebacks. Additionally, used together with other tools, it can help analyze coaching decisions, including 4th down decisions, 2-point conversion attempts.

As always, the link is

  • Spread The Love
  • Digg This Post
  • Tweet This Post
  • Stumble This Post
  • Submit This Post To Delicious
  • Submit This Post To Reddit
  • Submit This Post To Mixx

12 Responses to “Live In-Game Win Probability 2.0”

  1. Anonymous says:

    Thanks for providing this service. I stumbled across this site during the NBA season and was blown away by your WP graphs.

    Looking forward to the NFL season and keep up the good work!

  2. bruddog says:

    I remember the guys at PROTRADE (Jeff Ma and company) had come up with a similar model and used to have it on their web page. I think ESPN even used it for a season to highlight the plays that affected winning percentage the most. But then PROTRADE decided to have a much less statistical bent to their website and got rid of it.

    I've enjoyed looking at these over the past year and glad someone is doing it again! Kudos!

  3. Anonymous says:

    Can someone please tell me when Brian posts his selections for the week? Thank you

  4. Anonymous says:

    They usually don't start until week 3 or 4. The system needs a few weeks of data before they're reliable.

  5. Brian Burke says:

    Bruddog-Thanks. I came across remnants of their stuff when I was researching WP models. There are a few archived ESPN articles out there featuring play of the week-type analysis based on it. The guy who developed the model it occasionally comments here.

    There is another one that's still up and running called It's live, like mine, but it's not always working. Commenter JimA contributed to its development.

  6. dfan says:

    Here's a good one; should the Steelers go for it on 4th and 1 down by 3 at the 10 yard line with 3 minutes left?

  7. Anonymous says:

    I'd love to see a derivative of the WP line... sort of an excitement meter.

  8. Anonymous says:

    I am also very interested in the expected value of that 4-1 decision down by 3.

  9. Anonymous says:

    why do the charts start at 50%? unless the spread is pick'em, one team has a greater chance than the other. why not use the winning probability of each time based on the moneylines. for e.g. right now miami/atl is 2.81/1.498 which corresponds to atlanta starting with a 65.23% prob of winning. thanks for your great work!

  10. ben says:

    Using Brian's WP Calculator, the Steelers had a 35% chance of winning at 4th and 1 on the Ten 9 yard line with 3:46 left.

    Their running game had been terrible all evening but 1 yard still isn't a lot to ask. Let's say that they had a 65% chance of converting (Big Ben sneek anyone?).

    If they failed on 4th down their win percentage drops to 25% and if they succeeded their win percentage goes up to 55%. So (.65*.55)+(.35*.25)=.445 so a 44.5% chance of winning if they go for it.

    After kicking their field goal their win percentage actually dropped to 32%. They were 12.5% more likely to win if they went for it on 4th down (assuming 65% chance to convert).

    What if their chance to convert was 50%? Then it's a WP of 40% if they go for it and still 32% if they kick the field goal. The chance to convert would have to drop to 23% for that fieldgoal to break even.

  11. Brian says:

    I realize that the model was not developed with college football in mind, but is there chance that live wp graphs will ever be available for some NCAA games, even if you still use the NFL model? They're just so fun to watch.

  12. TheEdge says:

    why not start each game's graph off at the probability you predict in your weekly predictions? surely each team doesnt have a 50% chance of winning at the beginning of the game.

Leave a Reply