Game probabilities for the divisional games are up at the New York Times' Fifth Down.
This week I break down the Seattle-Atlanta game. It should be a close one.
Division Round Game Probabilities
published on 1/10/2013
Subscribe to:
Post Comments (Atom)
I've read your article on how you determine game probabilities. I have several questions:
a. How often do you update the model? How drastically do the weightings change?
b. Do you publish your calibration data (i.e. when you forecast a team to have a 70% chance to win, do they actually win that often)?
c. Have you considered a Monte Carlo simulation approach to game probabilities? You've got so much data about individual teams' performance in very specific situations, it seems like you could randomly generate a game's worth play-by-play outcomes, and repeat 1000 times to get a good idea of who will come out on top.
While I am a huge Packer fan, and very biased here, I can;t help but think the chances for GB are significantly higher than your numbers suggest for several reasons, most of them injury related.
Does the fact that Justin Smith may be out or limited significantly coupled with GB's season long injury situation that is only recently immensely improved change your outlook of the game much?
What about the fact that Cap has only 7 games worth of experience, and his Int% is unsustainable for a quarterback who throws downfield as often as he does (if Barnwell is to be taken seriously) and the data has no way of capturing that?
If you really believe in those numbers, you should bet a lot of money on Houston, and little bit on Seattle and maybe San Francisco, too.
But the difference to the betting lines for Houston is enormous. Vegas gives them roughly a 20% shot.
Actually that´s what i did. Betting straight up against NE. Gives me 1:3.8
I took that great chance, especially when that idiot Dan Shaughnessy gave extra motivation for the Texans. It wouldn´t be the first time that NE chokes early in the playoffs...
Karl, Germany
Hey! I know this is kinda off topic but I'd figured I'd ask.
Would you be interested in trading links or maybe guest authoring a blog post or vice-versa?
My website addresses a lot of the same subjects as yours and I believe we could greatly benefit from each other.
If you happen to be interested feel free to send me an e-mail.
I look forward to hearing from you! Wonderful blog by the way!
Visit my site Movie Discussions
You have the teams on your live win probability graph reversed.
The early start time hurts Seattle
Mbchoe
You STILL have the teams on your live win probability graph reversed.
b. Do you publish your calibration data (i.e. when you forecast a team to have a 70% chance to win, do they actually win that often)?
--------------------------
John, i've taken a very small sample and looked at them, and they had been consistent with that.
and, a bit off topic, but nate silver does the exact same thing predicting states and their win probability (for democrat or repub), and he shockingly does not at all hit his percentages, yet he always gets praised for being "right" because his favourite always wins. Brian's model seems reasonable, whereas nate silver's election model is clearly wrong.