## Game Probabilities - Week 13

Weekly game probabilities are available now at the nytimes.com Fifth Down. This week I also discuss the playoff prospects for the Jets and Giants, and why they may seem pessimistic to most fans.

### 7 Responses to “Game Probabilities - Week 13”

1. James says:

Are you going to crank out the 16-0 projections any time soon?

2. Giles says:

I am having trouble getting my comment to post. This is my 3rd try so apologies if the other comments appear later...

Brian, I was wondering if your predictions took into account teams fielding lesser teams once they had secured a play off spot or teams playing below their abilities to secure a better draft spot?

Or do you show the predicted outcome based upon both teams playing at their full potential?

I realise that the wild card picture is still quite open but I always have trouble predicting the outcome in the last few weeks of the season and I wondered if that was something you took into account.

3. Michael says:

Brian - Have you done any comparisons as to how your model's "expected" number of correct picks compares to actual accuracy.

For example, your model "overperformed" last week by going 15-1, when the expected number of correct picks was 11 (11.4 to be precise). It would be interesting to see how, year to date, your model's expected accuracy compares to actual accuracy. Just another way of gauging if there is any over or under confidence in your model.

4. Andy says:

Here's the note from the 5th down editor saying that Brian is too modest because he went 15-1. Meanwhile, Brian and everyone here is thinking this is probably noise but if it is significant it shows that the model performed poorly by undervaluing probabilities of wins.

(editor’s note: Brian is too modest to say so, but his model went 15-1 last week. The one miss was Ravens-Steelers (Ravens won in overtime), and the decision to sit Roethlisberger might have tipped the balance on that one.)

5. Anonymous says:

where do i find brian's season long record of W-L predictions?

6. Michael says:

I took it upon myself to answer my own question. Based on Brian's win probabilities, the expected number of correct picks from weeks 4 through 12 was 91.4 (out of 128 games). The actual number of correct picks was 92. Pretty impressive.

7. Anonymous says: