Archive

Tag Archives: Football Outsiders

I wasn’t going to post today. I’ve had some bonus Saturday posts the last couple weeks (about Pro Bowl snubs and my emotional state regarding the 49ers-Packers game) and I am still exhausted after yesterday. But then I was all like… nah, check the Super Bowl odds! But I am keeping it quick.

Here are the eight remaining teams’ odds to win the Super Bowl at Sportsbook.com, alongside Football Outsidersestimation of their true probability, ranked by expected payout:

Super Bowl XLVIII Odds- 1/6/2014

Team American Odds Odds To One Break Even FO Chance Expected Payout Rank
CAR 1000 10 9.09% 15% 5.91% 1
NE 750 7.5 11.76% 17.2% 5.44% 2
NO 1500 15 6.25% 6.5% 0.25% 3
SD 1500 15 6.25% 6% -0.25% 4
IND 2000 20 4.76% 2.3% -2.46% 5
SEA 250 2.5 28.57% 24.6% -3.97% 6
SF 500 5 16.67% 8.9% -7.77% 7
DEN 250 2.5 28.57% 19.6% -8.97% 8

Again, Football Outsiders’ model can’t really account for all the Pats’ injuries1, so don’t immediately buy into New England. But the Carolina Panthers are still looking nice and undervalued! Here are teams’ odds to win their conference championship:

2013-14 NFL Conference Champion Odds- 1/6/2014

Team American Odds Odds To One Break Even FO Chance Expected Payout Rank
NE 340 3.4 22.73% 37% 14.27% 1
CAR 400 4 20.00% 27.9% 7.90% 2
SD 700 7 12.50% 14.7% 2.20% 3
NO 800 8 11.11% 13.1% 1.99% 4
IND 750 7.5 11.76% 8.1% -3.66% 5
SEA 100 1 50.00% 41.6% -8.40% 6
SF 260 2.6 27.78% 17.4% -10.38% 7
DEN 62.5 0.625 61.54% 40.3% -21.24% 8

I’m a little more inclined to like New England here; with a home win over Indianapolis, and another Chargers win in Denver (the last of which came less than one month ago), it might not be… well no, it’s still pretty crazy given the injuries, but maybe has a higher chance than 22.72 percent? I’m intrigued by San Diego and New Orleans as well. We know the Chargers can win in Denver, but we don’t seem to believe it (Denver is favored by 10 points); we saw the Seahawks crush the Saints a few weeks ago, and might believe that a little too much. I would certainly not put money on Denver at this point, not with Von Miller out for the season.

Fewer than four weeks until the Super Bowl. Can’t wait!


  1. OH, and there’s been another one! Top run defender Brandon Spikes is out for the year, after playing the whole regular season. Really, really sad stuff. 
Advertisements

Last time I checked the temperature in Vegas was December 26th, last Thursday. Five whole days ago! I remember thinking throwing some money at Green Bay and/or Carolina might not be a bad idea. Rodgers was probably coming back (week 8 of a typically 4-6 week injury), and the playoff odds report from Football Outsiders suggested that New England might have good value (which I ignored due to their injury plague), as well as Carolina (which I didn’t, with encouraging signs from Steve Smith). Here’s a little doohickey of how the early Super Bowl bets looked back then (and here’s everything I wrote last time):

Super Bowl XLVIII Odds- 12/26/2013

Team American Odds Odds To One Break Even FO Chance Expected Payout Rank
NE 1000 10 9.09% 14.2% 5.11% 1
CAR 850 8.5 10.53% 15.2% 4.67% 2
PHI 3000 30 3.23% 4% 0.77% 3
CIN 1800 18 5.26% 5.9% 0.64% 4
NO 2500 25 3.85% 4% 0.15% 5
KC 3500 35 2.78% 2.9% 0.12% 6
PIT 20000 200 0.50% 0.1% -0.40% 7
ARI 9000 90 1.10% 0.6% -0.50% 8
DAL 10000 100 0.99% 0.2% -0.79% 9
MIA 9000 90 1.10% 0.3% -0.80% 10
BAL 10000 100 0.99% 0.1% -0.89% 11
SD 10000 100 0.99% 0.1% -0.89% 11
CHI 6000 60 1.64% 0.7% -0.94% 13
IND 3500 35 2.78% 1.1% -1.68% 14
GB 4000 40 2.44% 0.2% -2.24% 15
DEN 300 3 25.00% 20.3% -4.70% 16
SF 750 7.5 11.76% 5.7% -6.06% 17
SEA 220 2.2 31.25% 24.3% -6.95% 18

As always, the odds are from Sportsbook.com. Break Even informs how often the bet needs to cash for you to break even at those odds; Football Outsiders provides estimates of how often the bet will actually cash; and the long run Expected Payout is the difference. Those Football Outsiders’ numbers use their core DVOA metric, weighted towards the end of the season, and also attempt to account for home-field advantage. I don’t believe they are exactly the true probabilities of each team winning the Super Bowl, but they are (“probably”) close. Are they closer than Vegas’ probabilities? Well, here are the updated Super Bowl odds, as of this afternoon:

Super Bowl XLVIII Odds- 12/31/2013

Team American Odds Odds To One Break Even FO Chance Expected Payout Rank
CAR 1100 11 8.33% 14.5% 6.17% 1
NE 850 8.5 10.53% 14.7% 4.17% 2
PHI 2000 20 4.76% 6.5% 1.74% 3
KC 3200 32 3.03% 3.4% 0.37% 4
SD 5000 50 1.96% 1.8% -0.16% 5
CIN 1800 18 5.26% 4.7% -0.56% 6
NO 2500 25 3.85% 2.8% -1.05% 7
IND 3000 30 3.23% 0.6% -2.63% 8
SEA 260 2.6 27.78% 24.9% -2.88% 9
GB 1800 18 5.26% 2.3% -2.96% 10
DEN 300 3 25.00% 20.2% -4.80% 11
SF 750 7.5 11.76% 3.5% -8.26% 12

Six teams have been eliminated, defrosting the picture a little. Remember five days ago when Green Bay was 40-to-1? And then remember this? Green Bay’s price in the betting world has more than doubled since then, now at 18-to-1. With Rodgers’ return, Football Outsiders made their projections using only data from games in which Rodgers played the majority of snaps. Given that they’ve got the Packers as winning it all only 2.3% of the time with Rodgers, and that Clay Matthews keeps getting surgery, 18-1 is too high a price for me now. But stranger things have happened.

What hasn’t happened is a similar shortening of the odds for the Carolina Panthers. In fact, their odds are longer! So let me get this straight: Carolina beat the 49ers in San Francisco by one point in a game in which the 49ers couldn’t score a touchdown (albeit without Michael Crabtree), secured a first-round bye (and time for Steve Smith to heal), will quite possibly host the Eagles for their first playoff game1, and the 49ers, down in the fifth seed, heading to Lambeau Field to face Aaron Rodgers, are more favored? Oh, I guess it’s because Carolina looked a little shaky last Sunday on the road against the 4-12 Atlanta Falcons, before sneaking out the win. OH WAIT the 49ers did the Exact. Same. Thing. Eight days ago. Remember this? Well I guess it’s because Carolina doesn’t have as good a defense… oh, right. Well, as good a young, mobile quarterback, then… oh, right. What the hell is going on here? This is Drew Brees’ face. And this is his face when he plays the Carolina Panthers. What else do you need?

As I’ve mentioned before, I think the reason Carolina is so low is because they haven’t been here before. Just a couple of months ago they were 1-3, and people were writing articles like this, agreeing when Rodney Harrison said in Week 5 that the Panthers should bench Cam Newton. Well, you don’t lose close games forever. The Panthers are legit. Yes, they could lose the Super Bowl, or in Seattle, or even their home divisional match-up next week. Every bet has risks. But at 11-to-1 (!!!), the Panthers are a risk worth taking.

That’s the Super Bowl odds. But you can also foolishly gamble on the conference champions! Here’s the same table for the conference champions:

2013-2014 NFL Conference Champions’ Odds- 12/31/2013

Team American Odds Odds To One Break Even FO Chance Expected Payout Rank
CAR 500 5 16.67% 27.5% 10.83% 1
NE 350 3.5 22.22% 31.9% 9.68% 2
PHI 1000 10 9.09% 12.3% 3.21% 3
KC 1600 16 5.88% 8.2% 2.32% 4
SD 2400 24 4.00% 4.4% 0.40% 5
CIN 650 6.5 13.33% 11% -2.33% 7
NO 1500 15 6.25% 5.4% -0.85% 6
IND 1200 12 7.69% 2.5% -5.19% 9
SEA 90.9 0.9 52.38% 42.1% -10.28% 10
GB 950 9.5 9.52% 5.6% -3.92% 8
DEN 66.7 0.67 60.00% 42% -18.00% 12
SF 400 4 20.00% 7% -13.00% 11

Again, I like Carolina, and while I don’t really like Philly to take the NFC, I might like them at that price. As for being overvalued, certainly my 49ers seem to be (because of the last two years I guess, or possibly all the Pro Bowlers we don’t deserve?), and I’m guessing the Broncos are because people forget the enormous impact of Von Miller? Who the hell knows.2 Anyway, that’s my gambling outlook at the moment, once again courtesy of Sportsbook.com and the insightful people at Football Outsiders. Check back after Round One for an update. Salaam.


  1. I think the Eagles are the least scary of the NFC teams playing this weekend: 49ers, Packers, Saints, Eagles. It’s just my opinion. 
  2. That was not a question. I’m telling you, who the hell knows. Nobody, that’s who. 

Did you say Super Bowl odds? Like how did we get to Super Bowl odds all of a sudden? Well, we’re heading into Week 17, and 20 of the NFL’s 32 teams will see their seasons end this Sunday. (tear, sniffle) At the moment though, 18 may still cling to their dreams of winning the Super Bowl this February. As they were months ago, Seattle and Denver seem to be the heavyweights, but then, so did the 2007 Patriots six years ago. Seattle and Denver are in fact the favorites in Las Vegas, at +220 (2.2 to 1) and +300 (3 to 1)1 respectively to come out on top in the Big Apple,2 while the Steelers, who need to win and three other teams to lose just to get into the playoffs,3 are currently +20,000 to win it all on Sportsbook.com. Whoo! Football Outsiders has also posted their updated playoff odds report, which factors DVOA (weighted to favor more recent games) and likely home field advantage into a simulation run 50,000 times. They posit that Seattle has a 24.3% chance of winning the Super Bowl, and Denver a 20.3% chance. (Pittsburgh 0.1%.) What’s it all mean?

When gambling, I like to take the same approach I take to poker: I get a little drunk and after losing a while look for ways to steal my opponents’ chips, cheat, or childishly point out their other shortcomings which is to say for every hand (or money line, or spread), I examine the payout and determine how often I’ll have to win for a bet to be profitable in the long run. By doing this correctly, I’m not really gambling anymore, I’m just making money/playing a sound strategy/etc, because in the long run, there is no risk. If my odds of winning are better than the odds of how much money I have to risk relative to the reward, this is a good bet in the long run! (Whoo!)

There are differences between poker and, say, betting on a Super Bowl champion before Week 17. This season is only played once, these games are only played once. It’s not an infinite game, it’s more like a one-shot game. But, if you had complete faith that Football Outsiders’ playoff probabilities were 100% correct,4 and perfectly reflected the true odds, what would that mean?

So you want to bet on the Broncos. They’re +300, or 3 to 1. To break even, this bet needs to come through 25% of the time.5 But Football Outsiders has them winning Super Bowl [Inserting Roman Numeral…] XLVIII only 20.3% of the time. The expected payout of this gamble is negative (-4.7%, in fact). Bummer. Here’s that same breakdown for all 18 teams still currently in the race, sorted high to low by expected payout (aka best bets):

Team Odds To One Chance to Break Even FO Chance Expected Payout Rank
NE 10 9.09% 14.20% 5.11% 1
CAR 8.5 10.53% 15.20% 4.67% 2
PHI 30 3.23% 4.00% 0.77% 3
CIN 18 5.26% 5.90% 0.64% 4
NO 25 3.85% 4.00% 0.15% 5
KC 35 2.78% 2.90% 0.12% 6
PIT 200 0.50% 0.10% -0.40% 7
ARI 90 1.10% 0.60% -0.50% 8
DAL 100 0.99% 0.20% -0.79% 9
MIA 90 1.10% 0.30% -0.80% 10
BAL 100 0.99% 0.10% -0.89% 11
SD 100 0.99% 0.10% -0.89% 12
CHI 60 1.64% 0.70% -0.94% 13
IND 35 2.78% 1.10% -1.68% 14
GB 40 2.44% 0.20% -2.24% 15
DEN 3 25.00% 20.30% -4.70% 16
SF 7.5 11.76% 5.70% -6.06% 17
SEA 2.2 31.25% 24.30% -6.95% 18

Even the briefest of glances reveals that Football Outsiders’ odds are not perfect; they are not adjusted for injures. The Patriots’ projection is weighted towards the recent past, but still includes data from when Rob Gronkowski (and Wilfork, and Vollmer, and the 600 other players the Patriots had go on IR this season) were on the field. It’s possible their true odds are still better than Vegas thinks. I probably wouldn’t bet on them, even at 10 to 1. Green Bay, meanwhile, looks like a terrible bet, but almost half of the data is without Aaron Rodgers. Though his collarbone’s status is still unknown, I would at least consider throwing a little money on Green Bay right now. Or maybe I would, if Clay Matthews weren’t injured as well. But then, there’s the Carolina Panthers. Mm-mm! That’s a leap I find most intriguing. I might be similarly intrigued by Cincinnati if Geno Atkins hadn’t gone out for the year, and I also am of the purely speculative opinion that New Orleans is too low, although Brees would have to bail out his offensive line for that to happen. Back to the Panthers, though.

The last reason I like the Panthers on these odds is because, despite beating the 49ers and their last-second, home win against the Saints last week, I just kinda feel they still may be a little underappreciated. There’s been a lot of coverage on the Panthers’ inability to close out games, how Cam Newton often choked his first two years, etc. Surely people will doubt his ability to “win the big one”. We love to apply narratives to outcomes, to explain things (in fact, I’m doing that right now to supplant my justification of Carolina), but many times it may not depend on Newton’s mental psyche so much as, well, sh*t happens.

Consider elite quarterbacks John Elway and Tom Brady. Elway lost his first four (!) Super Bowls, before winning consecutive championships his final two seasons. Brady won hist first three (starting when he was 24!), before losing two to the Giants. Elway constantly heard about how he couldn’t win “the big one”, but no one said that about Brady after his losses, what with his three rings and all. It’s nonsense. Football, more so than any other professional sport, is a team effort.6 The quarterback is the most important position, not the only important position. There are 11 guys on offense, 11 guys on defense, and still others on sub-packages and special teams. Ironically, in close contests when a quarterback “chokes”, the quality of his teammates matters even more, as the margin between defeat and victory shrinks.

The numbers suggest the Panthers’ odds in Vegas are too long. The above is why I think that might be. Newton’s Panthers have never made the playoffs, and have known only losing seasons. Newton just isn’t a winner yet. Well, no quarterback is, until he wins. Forget the narratives. Elway was a great quarterback for many seasons; that he won championships in his final two might have a little to due with additional experience and wisdom, but more to do with having Terrell Davis, or more favorable match-ups, the technological advancements in scouting your opponents, etc. If a quarterback is good enough to get to the Super Bowl, he’s good enough to win it. Many factors beyond a quarterback’s control come into play, and that we may overlook them does not make such factors any less significant. And hey, sometimes… sh*t happens.


  1. For +###(##) American odds, just divide by 100 to get the fractional odds. The + means how much you win (profit, net, etc) on a $100 bet. 
  2. Well, New Jersey, technically… lame. 
  3. The Steelers need to beat Cleveland at home, have New York (Jets, duh) win in Miami, have Baltimore lose in Cincinnati, and have Kansas City win in San Diego. 
  4. I do not endorse such faith. But they are “probably” pretty close. (Heyyy-eyyyy!) 
  5. 1/(1+3); Every time you win you triple your money, so you can lose three times for every time you win (betting the same amount), so that’s 0.25 = 25%. 
  6. I haven’t looked at this scientifically, but you’ve got around 25-30 guys who see significant time on a football team (47-man active roster). Hockey’s at 19-20 (20), baseball 18-22 or so (but pitchers make it weird, anyway 25 guys on the active roster), soccer 11-13 (18), and basketball’s down about 7-10 (13). 

Two things have been on my mind since yesterday’s NFL action. One, can any team replicate the Arizona Cardinals’ shocking success in Seattle, where they won 17-10 as nine-point underdogs? Two, how is it possible the New York Jets can still finish without a losing record? How did they get to seven wins heading into their final game? As these things aren’t really related at all, I made them related in this post, another feature of “What’s More Impressive?“.

A quick rundown of some numbers I’ll be using to answer:

  • DVOA stands for Defense-adjusted Value Over Average. Football Outsiders created it to determine relative team strength. The idea is that there are only 16 games a season, a small sample to rank teams on. But there are lots of plays in a season, usually more than a hundred every game, so Football Outsiders looks at how successful a team is on an average play, and then compares that to a league-wide average, using percentages. Feel free to read more.
  • PW% stands for Pythagorean Winning Percentage. This uses the idea that the margin of victory (or defeat) is a significant indicator of team strength, especially over the course of the season. The formula is (Points Scored ^ 2.37) / {(Points Scored ^ 2.37)+(Points Allowed ^ 2.37)}.
  • Point spreads have long been used by Las Vegas casinos. They’re designed to get even action on both sides. Most would bet that the Broncos would beat the Texans, so Vegas increases payouts for betting that the Broncos will win by, say, 10.5 points (the line for yesterday’s game). Casinos frequently adjust spreads to ensure that they see half the action on each side. The more a team is favored, the better their chances are of winning outright. (Duh.)
  • HFA stands for Home Field Advantage. In a recent column on Grantland, Bill Barnwell examined a team’s average margin of victory in its current stadium, and compared that margin to their average margin of victory (or defeat) on the road, going back through the 2002 season. The difference between the two average margins, divided by two, is historically the expected “extra” points for the home team, relative to a neutral site.

Let’s start with the Cardinals. They’ve been pretty good this year, sure. They were 9-5 heading into yesterday’s game, still alive for an NFC Wildcard berth. They have been cursed by their company; the league-leading 12-3 Seahawks, as well as the 10-4 49ers, are also in the NFC West division. They had lost to them each once already, heading into Week 16. They’d beaten the Lions (that meant something for most of the season), and the Panthers (who clinched a playoff spot yesterday), and the Colts (who clinched last week), but all at home. The Cardinals were 6-1 at home, and 3-4 on the road, and that home loss came to… the Seahawks. Second-year Seahawks’ quarterback Russell Wilson was a perfect 14-0 at home heading into yesterday. The Seahawks have looked dominant all year, and, well, unbeatable at home.

Cardinals vs. Seahawks

  • Record: 10-5 (66.57%, tied for 7th); Seahawks’ Record: 12-3 (80%, tied for 1st)
  • DVOA (through Week 15): 10.9% (10th); Seahawks’ DVOA: 40.4% (1st)
  • PW%: 60.29% (9th); Seahaw’s PW%: 79.17%% (1st)
  • Record of nine-point underdogs (since 2003): 28-88 (24.1%)
  • Seattle’s HFA (since 2002): 5.2 points/game (1st)
  • Interceptions Thrown By Carson Palmer: 4 (4!!!)

1 2 Here’s the thing: I was considering taking the points in this game, but ultimately decided the Seahawks were just too good. Worst case scenario, Arizona would hang around, but Seattle would eventually cover (see: Broncos-Texans). But if you had guaranteed to me that Carson Palmer would throw four interceptions, I would have put my life savings on Seattle (or, at least some real money). I can’t believe they won with him throwing four picks. I know they did it with a really, really good defensive performance, aided by an injured Seahawk offensive line, but, like, how did they do that?

As for the Jets, they’ve been, uhh, bad. They now have seven wins (!), five of which are against teams with 5-10 records or worse.3 They lost by 25 to the Titans. (Yes, the Titans.) They lost by 40 to the Bengals. (Yes, 40.) They beat the Bills by seven… and later lost to them by 23. But they’re now 6-2 at home, 1-6 on the road, and a remarkable 5-1 in games decided by a touchdown or less. That includes (home) victories over the Patriots by three points4 and the Saints by six points. The thing is that “games decided by a touchdown or less” suggest more luck rather than skill. We love fitting narratives to teams, players, and coaches about how they always pull it out, but a team’s record in close games converges to .500 the more close games they play. The Jets flipped a coin six times and got five victories back. What else suggests the Jets have been lucky?

Jet’s Victories (TB, BUF, @ATL, NE, NO, OAK, CLE)

  • Record: 7-8 (46.67%, tied for 18th); Record of Defeated Opponents: 41-64 (39.05%)
  • DVOA (through Week 15): -13.8% (26th); Average DVOA of Defeated Opponents: -5.27% (Average DVOA Rank of Defeated Opponents: 19.4)
  • PW%: 30.79% (30th); Average PW% of Defeated Opponents: 44.88% (Average PW% Rank of Defeated Opponents: 19.9)
  • Record of teams with Jets’ spread (since 2003) in Jets’ victories: 485-687-1 (41.4%)

That last number is the sum record of teams in match-ups similar to the Jets in their victories, in Vegas’ eyes. The Jets were four-point underdogs in their Week 1 victory over Tampa Bay, and historically four point underdogs have gone 45-91 (straight up, not against the spread). In their Week 3 win over Buffalo, the Jets were 2.5-point favorites, who’ve gone 66-63, etc. Most of the teams the Jets have beaten were bad. But they’re still surprising wins, given that the Jets have seemed even worse (except for, you know, the whole “winning” technicality). What’s more impressive?

Cardinals’ One Win @SEA vs. Jets’ Seven Wins (TB, BUF, @ATL, NE, NO, OAK, CLE)

  • Winning % Gap Between Competition: Cardinals -14.3%; Jets +7.62%
  • DVOA Gap: Cardinals -29.5% (9 ranks); Jets -8.53% (6.6 ranks)
  • PW% Gap: Cardinals -18.88% (8 ranks); Jets -14.09% (10.1 ranks)
  • Historical Winning % Given the Same Spread: Cardinals 24.1%; Jets 41.4%

I feel compelled to declare the New York Jets’ 7-8 record at this point (much) more impressive than the Cardinals’ road victory over the Seahawks, primarily for two reasons. One, ironically this gives the Cardinals more credit. I think they’re a good team, and while I don’t think they’d win in Seattle every time (more like two or three out of ten), this wasn’t a fluke. Two, the Jets have pulled off a bunch of unlikely wins. While the probability of a Jets win in any single one of those games maybe wasn’t as low as the Cardinals’ in Seattle, winning all of them is remarkable.5 Using Vegas spreads, for instance, there was a 24.1% chance the Cardinals won, and on average a 41.4% chance the Jets did. But the Jets won seven times. With an expected winning percentage of 41.4%, the odds you go 7-0 are 0.2%!6 And I think that’s damn impressive.


  1.  http://www.footballoutsiders.com/dvoa-ratings/2013/week-15-dvoa-ratings 
  2.  http://www.teamrankings.com/nfl/odds-history/results/ 
  3. Actually, the Falcons are 5-9. If they beat the 49ers tonight, they would be 6-9. But still. 
  4. Perhaps you recall, this was a bizarre game where the Jets game-winning field goal in overtime actually missed, but then there was some phantom penalty no one had ever heard of on the Patriots, and the Jets got to kick again, from much closer. 
  5. The games are statistically independent, that is, the Jets are a football team with strengths and weaknesses in their match-ups with opponents, and that the outcome of a particular game does not affect the probability (strengths and weaknesses) of an outcome of any other game. 
  6. (414/1000)^7 
%d bloggers like this: