Last year, we tracked the odds of the Patriots going 16-0, which eventually ascended to 100%. Although it's very unlikely that that'll happen again, we do have two teams who are still working on a "perfect" season through nine weeks. The 8-0 Titans are trying to follow in the footsteps of New England, while the Lions have a chance to become the first team to go 0-16. So, which is more likely?
To figure out the predicted lines for each game, I used the LVSC power rankings, and considered whether the game was home or away. Using the lines, we can estimate a team's chance of winning (or losing) each game. Here's Tennessee's schedule:
The average over/under for Tennessee games so far this year has been 37.5, so I used that number for each game. In Week 13, if they were 10.5 point favorites against the Lions, the expected score would be 24.0-13.5; applying the Pythagorean formula to that gives the Titans a 79.6% chance of winning, which is what you see in the table.
Tennessee has two pretty difficult road games coming up over the next couple weeks, but then they get a break, with three games they win more than 3/4 of the time. After that it gets tough again, with two more tough road games, and a matchup against the Steelers. Add it all up, and they have only a 2.9% chance of going undefeated.
Now, for the Lions:
Detroit's average O/U thus far has been 44.5, so that was used for their scoring environment. (Except for the game against Tennessee, where I averaged the two teams' totals and used 41, for both tables.)
This is an incredibly difficult schedule. The average LVSC rating of these teams is 139.8, nearly a field goal above the league average of 137.2, although the Lions do have five of their final eight games at home.
Detroit's easiest remaining game comes when they host the Vikings in Week 14. That shows just how difficult their schedule is; Minnesota is the LVSC's #17 team. Overall, the Lions have a 7.2% chance of finishing the year winless, which is more than twice as high as Tennessee's undefeated percentage. This isn't too surprising, as the Titans' success has been viewed as a bit fluky, while Detroit is just really, really bad.
Sunday, November 2, 2008
What Are The Odds: Another Perfect Season
Subscribe to:
Post Comments (Atom)
11 comments:
I'm sure there's an explanation that I'm not mathematically inclined enough to figure out, but why does Tennessee have a 79.6% chance of winning at Detroit, while in the 2nd table, Detroit has a 75.8% chance of losing that game? Shouldn't those numbers be identical?
No good reason. Was just using different scoring environments for each. Which was dumb, fixed it. Thanks.
Can you do a post on these numbers for college basketball?
http://rushthecourt.wordpress.com/2008/11/02/vegas-odds-preseason-check-in/
Am I right in agreeing with the poster that UConn and Wake Forest are steals here?
I don't see any value with UConn. I thought about it at 20:1, but decided against it because of TVOM and how hard that conference is. Definitely not at 14:1.
Maybe on Wake, not sure. Still working on getting back into college basketball mode.
While we're on the topic of CBB, back in May/June you recommended UCLA at 30-1. I finally took that bet at sportsbetting.com 3 weeks ago. Now, the odds have been slashed to 8-1. Just curious if you have any insight as to what may have happened.
These numbers are far from perfect (I sincerely doubt the Tits will be only a 1 point favorite @ Jax, but I could just be dumb), but I enjoy this type of analysis anyway. Thanks.
Interesting way of determining win probability. When I do these sort of analyses, I use the half point calculator to back out the money lines and get it that way. I'm not sure one is more correct than the other, but I'd be interested to hear your thoughts.
On one hand, I think using the half point calculator would completely ignore the scoring environment for a team, which is not good.
On the other, I think I might be overrating it; Tennessee's lines will take into account that their games are more low-scoring, so the lines used here might be too high. For example, if they were 37.5 points better than some team, automatically doing it my way would give the Titans a 100% chance of winning, which would obviously be unrealistic.
My guess is that averaging the results from the way I do it here and your way using the half-point calculator would be the most accurate. Either way there probably wouldn't be much difference.
You know, 67.3% of all stats are made up...
What are the odds of the Titans going 16-0?
By that I mean a million to one ten million to one, etc?
I'm not trying to be too condescending here, but do you know anything about probability and gambling?
2.9% = ~33.5-to-1
Post a Comment