BBTF's Newsblog Discussion :: Uchill: The math says count on Patriots victory: How football analytics predict the big game

Baseball Primer Newsblog
— The Best News Links from the Baseball Newsstand

Saturday, February 04, 2012

Uchill: The math says count on Patriots victory: How football analytics predict the big game

I always figured my dancing with Madonna was the low point of her career, but after this Sunday…that might change.

Predicting 60 percent of games correctly is astoundingly high – gamblers need to win less than 53% of their games to make a profit. That is why analytics are so popular among bettors, said Elihu Feustel, professional gambler, one-time casino consultant and co-author of “Managing Risk: Attacking Vegas and Wall Street.”

Feustel, of South Bend, Ind., devotes multiple chapters of his book to betting on football. He does not, however, like football.

“I think it’s boring,” he said. He is still willing to make money off of it.

...Baseball may be better known than football for the use of analytics. The Oscar-nominated movie based on a best-selling book, “Moneyball,” traced the Oakland A’s road to success using advanced statistics , but no plans have been made for “Football Outsiders: The Movie.” But football analytics have their own vibrant research community in universities across the country.

Vince Gerrano, executive director of the Sports Analytics institute at Manhattanville University in New York, said football analytics time has come.

“Baseball came first because it’s so much more difficult to analyze individual contribution from a player in football,” he said. “In football, there is so much interdependency.

Repoz Posted: February 04, 2012 at 10:15 AM | 43 comment(s) Login to Bookmark
Tags: amateur, projections, sabermetrics, steroids

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

1. Tricky Dick Posted: February 04, 2012 at 11:31 AM (#4053386)

I liked the article, but I really get tired of the "pocket protector" mime, which shows up in the first sentence.

2. Curse of the Graffanino (dfan) Posted: February 04, 2012 at 11:42 AM (#4053389)

The math referenced in the article does not "say count on Patriots victory". I know it's just as irresistible to use the word "count" in an article about any sort of numerical analysis as it is to use the phrase "pocket protector", but come on.

3. PreservedFish Posted: February 04, 2012 at 11:56 AM (#4053393)

I met a hipster farmer that was wearing a pocket protector. He looked awesome, actually. Definitely made it look cool.

4. ValueArbitrageur Posted: February 04, 2012 at 12:45 PM (#4053419)

Predicting 60 percent of games correctly is astoundingly high – gamblers need to win less than 53% of their games to make a profit

There is this little thing called "the line"...

5. Bitter Calculus Instructor Posted: February 04, 2012 at 01:07 PM (#4053428)

The sports radio people around here all seem to be split as to whether it will be a tight game, or if the Giants will completely crush the Pats.

6. Pasta-diving Jeter (jmac66) Posted: February 04, 2012 at 01:19 PM (#4053435)

Predicting 60 percent of games correctly is astoundingly high

NOWHERE in TFA is there anything about predicting 60% of the games correctly (and DVOA is useless as a predictive tool--that's been shown over and over again)

7. Slivers of Maranville (SdeB) Posted: February 04, 2012 at 01:24 PM (#4053437)

and DVOA is useless as a predictive tool--that's been shown over and over again)

I don't think that's the case, although perhaps it depends on one's definition of useless. Useless predictive tools would be things like tea leaves, the I Ching, throwing darts at a page...

8. Bitter Calculus Instructor Posted: February 04, 2012 at 01:43 PM (#4053443)

Accuscore has historically been 55% accurate against the line, which is high enough to be profitable. Bet games where accuscore and Vegas disagree. Anyone have the numbers on how good football outsiders has been?

http://accuscore.com/sports-betting-system/nfl-analysis/accuscore-is-the-best

9. tshipman

Posted: February 04, 2012 at 01:44 PM (#4053444)

The nature of sports betting is such that any tool that is predicative vs. the spread for any length of time shortly will not be.

10. Bob Dernier Cri

Posted: February 04, 2012 at 02:10 PM (#4053457)

The math referenced in the article does not "say count on Patriots victory"

And a good thing, because I was fixing to say "that's a kind of math that doesn't understand how small a sample one game is" :)

11. Bitter Calculus Instructor Posted: February 04, 2012 at 02:14 PM (#4053460)

The math says: count on six tenths of a Patriots victory!

12. Pasta-diving Jeter (jmac66) Posted: February 04, 2012 at 02:27 PM (#4053465)

Anyone have the numbers on how good football outsiders has been?

it depends who you talk to and what you mean by "good"; the FO website people claim that DVOA is a decent predictive tool but admit that the raw value of the teams' DVOA cannot easily be converted to a pointspread. I've found it's not even very good at predicting which team will win EARregardless of the spread

13. ray james Posted: February 04, 2012 at 02:34 PM (#4053470)

Feustel, of South Bend, Ind., devotes multiple chapters of his book to betting on football. He does not, however, like football.

“I think it’s boring,” he said. He is still willing to make money off of it.

Anybody seen a photo of this guy and Ace Rothstein together?

14. Pasta-diving Jeter (jmac66) Posted: February 04, 2012 at 03:06 PM (#4053483)

let's check the DVOA records this postseason--

going into the wild card games:
Hou 19.4
Cin 0.5
(win)

Atl 15.5
NYG 9
(loss)

NO 23.3
Det 13.6
(win)

Pit 23.7
Den -11.0
(BIG loss--this was the widest DVOA spread between playoff teams in history)

Divisional:
NO 34.7
SF 17
(loss)

Hou 22.1
Bal 11.4
(loss)

NE 24.6
Den -5.1
(win)

GB 26.6
NYG 8.3
(loss)

Conference:
NE 32.4
Bal 14.5
(win)

SF 19.7
NYG 13.7
(loss)

so DVOA is 4-6 in the postseason just picking straight-up winners

(EDIT: BTW, DVOA :
NE 33.2
NYG 14.1

15. Walt Davis Posted: February 04, 2012 at 07:01 PM (#4053570)

I don't know nothing about DVOA but I gather it's a team-wide performance measure?? I can imagine such a measure would be reasonably straightforward in football, especially in measuring past performance, but how does it account for injuries, etc. in predicting the next game?

(BIG loss--this was the widest DVOA spread between playoff teams in history)

Tebow loves Jesus, Jesus loves Tebow.

Brady clearly has a deal with the devil.

16. Zipperholes Posted: February 04, 2012 at 07:20 PM (#4053581)

The nature of sports betting is such that any tool that is predicative vs. the spread for any length of time shortly will not be.

Only to the extent it is both known to and properly accounted for by the betting community.

17. cardsfanboy

Posted: February 04, 2012 at 07:40 PM (#4053587)

Tebow loves Jesus, Jesus loves Tebow.

Brady clearly has a deal with the devil.

I'm not a Brady fan, didn't understand this Tebow thing at all. But that is a good comment.

As to the superbowl, somewhere on this site I was arguably willing to say, even before the season began, that I would take the Patriots, Steelers or Colts in the Superbowl. I'll do the same thing next year. 10 out of the last 11 seasons, have to love parity in the NFL.

18. Ron J Posted: February 04, 2012 at 08:30 PM (#4053615)

#4 According to Pete Palmer (He did a Hidden Game of Football, had nothing like the impact of the baseball book, but it's not bad as a starting point) you have to believe the line is wrong by 3 points to show a profit in betting football.

19. Cblau Posted: February 04, 2012 at 10:36 PM (#4053660)

Vince Gerrano, executive director of the Sports Analytics institute at Manhattanville University in New York,

Nice copy editing.

20. The DA Baracus Hypothesis Posted: February 04, 2012 at 11:26 PM (#4053672)

As to the superbowl, somewhere on this site I was arguably willing to say, even before the season began, that I would take the Patriots, Steelers or Colts in the Superbowl. I'll do the same thing next year. 10 out of the last 11 seasons, have to love parity in the NFL.

Good luck with that Colts bet.

We've had this discussion before. Parity isn't about equal trophy cases, it's about an equal playing field. Yeah, the NFL is set up to artificially create parity (as are other hard cap leagues), but that doesn't change the fact that it's still there.

21. Kurt Posted: February 04, 2012 at 11:32 PM (#4053673)

but how does it account for injuries, etc. in predicting the next game?

It doesn't.

22. John DiFool2 Posted: February 04, 2012 at 11:47 PM (#4053679)

Anyone care to ##### and moan with me about the football HoF shooting down the candidacies of 3 guys who are top ten all time in receptions? Absolutely indefensible. We ##### and moan all the time here about the baseball Hall and its pernicious biases, but IMNSHO Canton does a much worse job than Cooperstown does, so much so that it isn't even funny.

23. Pasta-diving Jeter (jmac66) Posted: February 04, 2012 at 11:51 PM (#4053680)

We ##### and moan all the time here about the baseball Hall and its pernicious biases, but IMNSHO Canton does a much worse job than Cooperstown does, so much so that it isn't even funny.

don't get me started

24. The DA Baracus Hypothesis Posted: February 04, 2012 at 11:54 PM (#4053681)

It's because these idiot writers have an unwritten rule that they only induct one guy per position, which is mind mindbogglingly stupid. Meanwhile Curtis Martin gets in. Makes no sense.

25. The DA Baracus Hypothesis Posted: February 04, 2012 at 11:59 PM (#4053685)

Tiquan Underwood, the WR with the flat top, was released earlier today by the Patriots. I'm sure the media, in their infinite wisdowm, will realize the Patriots did it because they needed a roster spot to bring in a DL from the practice squad and that they did it now so that the Giants couldn't claim him and to not tip their hand until now about what their active roster would look like* rather than make this about what a monster Bill Belichick is.

*granted both those things are near-trivial "advantages"

26. larkin4HoF Posted: February 05, 2012 at 03:30 AM (#4053747)

Don't any of you people watch ESPN. If you want to know who will win the Super Bowl, you only need to know who is better-Eli Manning or Tom Brady.

27. cardsfanboy

Posted: February 05, 2012 at 04:08 AM (#4053751)

#25..... nfl post season roster rules are seriously ###### up. And the Patriots are absolute ####### the most corrupt organization that doesn't feature a donkey or an elephant as a mascot in the history of the U.S.

They are still going to win the Super Bowl, because they have an actual QB worth a crap(not as a human being mind you, just as an athlete) 27-10 Patriots.

28. The DA Baracus Hypothesis Posted: February 05, 2012 at 10:54 AM (#4053798)

What are you talking about, post season roster rules? They're the same as the regular season. The Pats added a DL from their practice squad, in order to do so they had to move a player from the 53 man roster. Since Ochocinco is going to be active, they cut the guy last on the WR depth chart. The reasons for waiting until Saturday are obvious.

And corrupt? Yeah they're ########, but they also do things differently which at times make them seem like bigger ########. But they are not "corrupt" unless like parity you want to change the definition of the word to suit your needs. Other teams do this, but since they're not in the Super Bowl it doesn't get noticed. Exhibit A.

29. Nate Posted: February 05, 2012 at 11:29 AM (#4053812)

Remember when cardsfanboy said Eli wasn't a Top 10 QB? Good times.

30. Bitter Calculus Instructor Posted: February 05, 2012 at 01:38 PM (#4053868)

[25] What exactly is the issue here? Is it that Underwood was on the team all year and doesn't get to play in the Super Bowl? I suppose you could say that's unfair to him, but it's sports. People get cut.

31. The DA Baracus Hypothesis Posted: February 05, 2012 at 01:42 PM (#4053872)

My point is there is no issue, but the media is making it one. These minds of moves happen every week but they need ammo for their "the Patriots are ########\" narrative that they're slaves to.

32. cardsfanboy

Posted: February 05, 2012 at 02:59 PM (#4053903)

Don't any of you people watch ESPN. If you want to know who will win the Super Bowl, you only need to know who is better-Eli Manning or Tom Brady.

if that is the case, then it's going to be a blow out in favor of the Patriots. Eli, third best quarterback from his draft class :)

of course that is as silly as predicting a mlb post season winner based solely on the pitching staffs, as the 2011 World Champion Phillies can attest to.

33. Walt Davis Posted: February 05, 2012 at 07:58 PM (#4054007)

And the Patriots are absolute ####### the most corrupt organization that doesn't feature a donkey or an elephant as a mascot in the history of the U.S.

The A's are offended.

34. cardsfanboy

Posted: February 05, 2012 at 08:17 PM (#4054020)

The A's are offended.

Considering how they were in bed with the Yankees in the 50's and 60's they don't have much reason to be offended. (ok, I forgot that the a's used an elephant/mule as a mascot for a while)

35. ray james Posted: February 05, 2012 at 08:30 PM (#4054039)

~~The A's~~Any Larussa team ~~are~~ is offended.

Fixed.

36. cardsfanboy

Posted: February 05, 2012 at 09:05 PM (#4054088)

Fixed.

not sure what that means. I'm guessing you are saying that since TLR is a paragon of virtue that he should be offended that one of his teams was compared to the corruptness that is the Republican party(and to a lesser degree the Democrats)?

37. Pasta-diving Jeter (jmac66) Posted: February 06, 2012 at 12:19 AM (#4054345)

4-7

38. Robert in Manhattan Beach Posted: February 06, 2012 at 05:36 AM (#4054401)

And so the Pats finish the year 15-4 and a trip to the Superbowl with exactly one win over a team with a winning record, and that game required a dropped game winning touchdown in the final minute and a shanked chip shot. The Giants finished the year with five such wins. But of course the computers loved the Pats. Makes perfect sense.

39. Misirlou's got a busy day, he's wearing a vest Posted: February 06, 2012 at 09:37 AM (#4054426)

And so the Pats finish the year 15-4 and a trip to the Superbowl with exactly one win over a team with a winning record

They also had 8 wins against teams that would have had winning records except for the loses to the Pats. There were a remarkable number of 8-8 teams this year, and the Pats played and beat 75% of them, 2 of them twice.

40. Russ Posted: February 06, 2012 at 10:49 AM (#4054462)

And so the Pats finish the year 15-4 and a trip to the Superbowl with exactly one win over a team with a winning record, and that game required a dropped game winning touchdown in the final minute and a shanked chip shot. The Giants finished the year with five such wins. But of course the computers loved the Pats. Makes perfect sense.

They don't make the schedule. The problem for the difference between the Pats' quality and their record was the weakness of the AFC East in particular and the AFC in general. Their intra-conference, inter-division games were: IND, PGH, DEN, KC, OAK, and SD, not exactly a murderer's row of teams to play. They played the NFC East so: PHI, NYG, DAL, WAS. They basically dodged the two toughest divisions in the league (AFC and NFC North) with the exception of a tragically flawed PGH team (whom they lost to). The new rotating intra-conference schedule creates this kind of problem, in that if a division is weak or strong from top to bottom, you're either coasting or getting bludgeoned. Steelers were similarly helped this year by a schedule that pitted them against the AFC South and the NFC West, the only difference was that Pittsburgh at least had to go up against a good team twice (BAL) and a somewhat non-easy team twice (CIN).

41. bunyon Posted: February 06, 2012 at 10:55 AM (#4054465)

And the Patriots are absolute ####### the most corrupt organization that doesn't feature a donkey or an elephant as a mascot in the history of the U.S.

I hang my head in shame for I spent a couple of minutes trying to come up with the teams cfb was talking about.

42. SoSH U at work Posted: February 06, 2012 at 11:42 AM (#4054513)

The Giants finished the year with five such wins.

So the Giants had one such win in the regular season?

Of course, the Super Bowl champs didn't go over .500 for good until six weeks ago.

43. Robert in Manhattan Beach Posted: February 06, 2012 at 12:40 PM (#4054590)

They don't make the schedule.

And I didn't say they did. The point is that the computer shouldn't love a team that beat nobody as much as it loved the Pats.

So the Giants had one such win in the regular season?

Of course, the Super Bowl champs didn't go over .500 for good until six weeks ago.

Indeed, maybe after they fix the strength of schedule component they could work on weighting recent games more heavily.

You must be Registered and Logged In to post comments.

<< Back to main

Saturday, February 04, 2012

Uchill: The math says count on Patriots victory: How football analytics predict the big game

Reader Comments and Retorts

Support BBTF

Bookmarks

Hot Topics