Comments on Advanced NFL Stats: Team Playoff Probabilities - Week 17

Oops. Yes, that's what I meant. 0.58 to 0.42 i...

2009-12-30T18:44:19.412-05:00

Oops. Yes, that's what I meant. 0.58 to 0.42 in favor of the Cowboys. Sorry to confuse.

It's a home game for Dallas, so they should ha...

2009-12-30T18:35:05.833-05:00

It's a home game for Dallas, so they should have the advantage.

My numbers are actually going to be .58 to .42 in ...

2009-12-30T18:08:24.040-05:00

My numbers are actually going to be .58 to .42 in favor of the home Eagles. That's just right for HFA in a game between equally matched teams. The more that respective team strength cancels itself out, like it does with equal opponents, the more important other factors, such as home field, will be.

HFA ranges from about 8 or 9% for equal match-ups to about 3% or less for mismatches.

Brian, I don't understand why Cowboys have suc...

2009-12-30T17:56:26.612-05:00

Brian, I don't understand why Cowboys have such an advantage in these stats of winning the division (57 to 43) vs. the Eagles, with both teams having a pretty close GWP (.77 to .76, Eagles favored). I didn't think your stats gave much for home field advantage. Can you explain.

Yes, they'll be up at the NY Times as usual to...

2009-12-30T17:41:41.333-05:00

Yes, they'll be up at the NY Times as usual tomorrow AM. No adjustments for teams resting players, though.

I am not sure whether this is the appropriate plac...

2009-12-30T17:37:57.647-05:00

I am not sure whether this is the appropriate place to ask, but I was wondering, Brian, if you plan on posting game probabilities this week since many of the games will be hard to predict based on stats? Thanks for all the great info.

Chris - I tink I would put it that the lines could...

2009-12-30T17:27:54.128-05:00

Chris - I tink I would put it that the lines could be overly favoring the "non-resting" teams after seeing the debacle in Indy with Painter.

Every game will not be that black and white with and without starters. I mean JT O'Sulllivan is prolly better than Sanchez at this point.

That's what's so cool about the app. You c...

2009-12-30T17:06:42.155-05:00

That's what's so cool about the app. You can move the slider to any probability you want for each game and recalculate the playoff probabilities.

The odds at nfl-forecast have the Jets as a close ...

2009-12-30T17:03:37.383-05:00

The odds at nfl-forecast have the Jets as a close favorite, but their a 10 point favorite due to CIN not playing to win - which is prolly about 80% to win for the Jets.

You need a "regular" probability, and a "resting" probability to account for this effect.

You're right, Johnathan. I was using an old Sp...

2009-12-30T16:16:17.125-05:00

You're right, Johnathan. I was using an old Spreadsheet to covert money lines to probability that was somehow corrupted.

I plugged in oddschecker odds in place of NFL-Fore...

2009-12-30T14:53:22.168-05:00

I plugged in oddschecker odds in place of NFL-Forecast odds to account for the predicted effect of teams resting their starters. Giving Houston a 76% chance of winning vs. NE and the Jets an 82% chance vs. CIN, here's what I came up with:

NYJ 82%
BAL 82%
DEN 25%
PIT 5%
HOU 5%
JAC 0.48%
MIA 0.06%

Also, New England moves up to an 86% chance of being the 3 seed.

Chris, are you sure about that? On oddschecker.co...

2009-12-30T14:46:02.749-05:00

Chris, are you sure about that? On oddschecker.com the Jets have closer to an 81-83% chance of winning. The Ravens have an 84% chance.

Interestingly, Vegas is giving both the Ravens and...

2009-12-30T12:18:33.580-05:00

Interestingly, Vegas is giving both the Ravens and Jets a greater than 95% chance of winning their respective games and clinching playoff spots. I understand that the Bengals will likely be resting some of their starters, but I'm wondering if these lines are the result of wishful thinking on the part of the betting public.

I'm reminded of the 1999 Chiefs, 9-6 at the time, missing the playoffs by losing to the 7-8 Raiders in Arrowhead.

In 2006, a 9-6 Denver team lost a win-and-in game to a 6-9 San Francisco team in Denver. Several other teams that could have made the playoffs with a win lost earlier in that day.

These are just two games that are etched in my mind as a Chiefs fan. It seems like week 17 upsets of teams that need a win to get in the playoffs are far too common to be giving out odds so far out of line with what the actual play on the field would justify.

I never consider ties in forecasting the remaining...

2009-12-30T11:55:44.565-05:00

I never consider ties in forecasting the remaining season. When ties do occur in actual games, I do consider the effect of ties in evaluating the tiebreakers.

If Pats beat Hou and Jets-Cincy TIE, don't Jet...

2009-12-30T11:46:42.525-05:00

If Pats beat Hou and Jets-Cincy TIE, don't Jets have non-zero chance of being No. 6 seed?

Yes. JAX has a non-zero chance of making the 5 see...

2009-12-30T11:27:46.215-05:00

Yes. JAX has a non-zero chance of making the 5 seed. If you want the extra decimal places, please visit NFL-Forecast.com.

I left out the extra digits just for readability, but I also don't believe we can really be that precise in our estimates to begin with. I wish there were an extra numeral between 1 and 0--something that means 'really, really tiny but not zero.'

We could name it Jaguero, after Jacksonville's chances of making the playoffs.

I assume that Jacksonville's numbers are off d...

2009-12-30T11:19:40.613-05:00

I assume that Jacksonville's numbers are off due to rounding errors? It shows a 1% chance of making the 6th seed, but a 2% chance of making the playoffs altogether.

Before anyone asks, Miami has a 0.2% chance of mak...

2009-12-30T11:10:37.454-05:00

Before anyone asks, Miami has a 0.2% chance of making the playoffs.