MONDAY MORNING UPDATE: The computer had a pretty great weekend, going about 50/50 but hitting its big bet on Buffalo.  Of course, all of that pales in comparison to tonight’s bet on Miami, but that really comes down to an internal accounting tweak – if Gamblor wins, it gets an extra $460 to play with, while if it loses I’m taking $460 away from it.  Win or lose, the results will count towards the season total.  For what it’s worth, the line has moved to +7.5, which shows that Vegas is still pretty nervous about having set the line so low in this game.

…and we’re back!

Welcome back to another fun-filled season of gambling, following the prognostications of that neon-clawed nightmare, the ill-tempered tempest, the one, the only, GAMBLOR!  Last year I introduced a new model, which I called Son of Gamblor, and the unholy alliance of these two abominations concluded the season with a small profit of $222.  As usual, I tinkered a little bit more with the program during the offseason.  I was hoping to develop a model for predicting winners that could be useful in betting on the money  line (i.e. predicting the winner or loser of the actual game without the spread factored in).   My efforts delivered a model that could accurately predict winners 59.9% of the time, which sounds pretty good.  But someone else has already developed an even better model, which picks winners at a rate of about 64%.  That’s even more impressive, right?  Not so much.  The best predictor of the actual winner of football games?  The goddamned spread.  That’s right, if you just pick the favorite to win each and every time, you’re going to be right 67.4% of the time.  Should you build a betting strategy based around that?  Absolutely not.  The odds on the money line will prevent you from ever seeing any kind of profit.  But it just goes to show you that Vegas is pretty smart when it comes to predicting who has the edge in the NFL.

Even though this exercise didn’t provide me with any new gambling tools, it did provide me with a new way to show you what Gamblor is thinking in each game.  Gamblor simulates each game close to 150,000 times, and comes up with a distribution of scores for each team.  The predicted score I’m using is simply the median of all these simulations – the value that represents the exact middle of the score curves.  For example, in the very first game of the 2011 NFL season the computer predicts that the Green Bay Packers will beat the New Orleans Saints by a score of 27 to 20.  The graph below shows the distribution of scores that the computer envisions for each team:

In most simulations, the computer predicts that Green Bay will score a lot of points.  In fact, it thinks Green Bay’s chances of scoring more than 40 points are higher than its chances to score fewer than 10.  And Gamblor simply doesn’t believe that New Orleans can keep up with that.  So in the computer’s eyes, it expects a fairly solid victory for Green Bay.  Currently, Green Bay is favored by just four points, which suggests (to Gamblor) that the sportbooks are trying to tempt people into betting on the Packers.  And they’ve been pretty successful – right now about 70% of the money is on Green Bay.  But as I’ve explained in the past, Gamblor is essentially designed to do the exact opposite of what it thinks the books are trying to trick people into doing.  Thus, both Gamblor and Son of Gamblor have relatively small bets on the Saints.

This year, I’ll be including Gamblor’s predicted scores for each game and graphs of the distributions for each along with my thoughts on each game.  As always, remember that Gamblor and Son of Gamblor are better at this than I am.

Away Home Spread Gamblor’s Pick Bet Amount Son of Gamblor’s Pick Bet Amount Overall Pick Overall Bet
NOS at GNB -4 NOS $4.50 NOS $3.38 NOS $7.88
PIT at BAL -2.5 PIT $4.08 BAL $1.33 PIT
DET at TAM -1.5 DET $7.24 DET $7.24
ATL at CHI 3 CHI $0.12 CHI $11.79 CHI $11.92
BUF at KAN -5.5 BUF $35.04 BUF $42.97 BUF $78.00
IND at HOU -9 IND $4.01 HOU $0.63 IND
PHI at STL 4.5 STL $9.18 STL $9.18
CIN at CLE -6.5 CIN $8.82 CIN $8.82
TEN at JAC -3 TEN $21.59 TEN $21.59
NYG at WAS 3 WAS $3.89 WAS $6.41 WAS $10.31
CAR at ARI -7 CAR $14.05 CAR $14.05
SEA at SNF -5.5 SEA $8.03 SEA $11.83 SEA $19.86
MIN at SND -9 MIN $4.56 MIN
DAL at NYJ -4 DAL $5.29 DAL $5.29
NEP at MIA 7 MIA $201.45 MIA $258.44 MIA $459.90
OAK at DEN -3 OAK $21.78 OAK $21.78


Computer’s Pick: New Orleans (+4)
Bet Amount: $7.88 (Consensus)
COMMENTS: It happens all too often that you expect a nationally televised game between two great teams to be a high-octane shootout (accompanied by a sky-high over/under), and it turns out to be dominated by the defenses, ending with a score of 13-10 or 9-7.  While I’m not really expecting that to happen, I won’t be surprised if it does, and in any case four points is a nice treat in what figures to be a close game.  I’m happy with Gamblor’s small bet on the Saints here.
Computer’s Pick: Pittsburgh (+2.5)
Bet Amount: $2.74 (Conflict)
COMMENTS: Pittsburgh has had the upper hand in this rivalry lately, winning six of the last eight contests, and I think that trend continues here.  And furthermore, these games are usually tight – there hasn’t been a difference of more than 9 points since 2007 – so getting the points is a nice bonus.  I think the Steelers will have an easy time here.
Computer’s Pick: Detroit (+1.5)
Bet Amount: $7.24 (Gamblor only)
COMMENTS: There’s too much hype surrounding Detroit right now.  I’m happy to jump on the bandwagon once it gets rolling, but let’s not forget that this was a team that went 6-10 last year, and they’re suddenly favored – on the road – against a team that was 10-6 last year.  I’m not saying that Detroit isn’t primed to have a solid season, but I don’t think they’re going to come galloping out of the gate the way everyone else seems to think they are.
Computer’s Pick: Chicago (+3)
Bet Amount: $11.92 (Consensus)
COMMENTS: I completely forgot that Chicago played so well last year.  But I still think that Atlanta, despite their reputation for problems outside the dome, will be able to dominate this game.  Enough so to cover the spread?  Absolutely.  Enough so that I’d start Julio Jones in my flex spot?  Not quite.
Computer’s Pick: Buffalo (+5.5)
Bet Amount: $78.00 (Consensus)
COMMENTS: The computer really likes the Bills here, and I do too.  They finished last season playing pretty well as a team, finishing 4-4 in their last eight games, and even though they’re susceptible to getting blown out by better teams like the Patriots and Jets, they played well against mediocre opponents like the Lions and the Dolphins.  Add Matt Cassel’s injury to the mix, and you’ve got a great opportunity for the Bills to pull off a surprise upset.
Computer’s Pick: Indianapolis (+9)
Bet Amount: $3.39 (Conflict)
COMMENTS: Without Peyton Manning, the Colts have proven to be helpless.  Kerry Collins won’t make a difference, especially not at short notice like this.  I’d actually expect the Colts would be better with Curtis Painter at the helm; he’s terrible, but I think the rest of the team would ratchet up their effort to account for this. Yet for all that, this spread seems too high for me.  Houston isn’t reliable enough to deliver the blowout that this spread suggests, and they won’t get up for this game the way they would have if Manning had been in the lineup.  Hence, I think the Colts will probably lose, but will cover the number while they’re at it.
Computer’s Pick: St. Louis (+4.5)
Bet Amount: $9.18 (Gamblor only)
COMMENTS: The Rams have a lot to be excited about this season.  Except that they don’t really, but it certainly seems like it.  The Eagles, on the other hand, are going to be under a tremendous amount of pressure to perform right out of the gate.  I’m not sure they’re ready for that yet.  I expect the Eagles to perform much as the Miami Heat did in the early part of the season – poorly.  They’ll get things together soon enough, but I wouldn’t be surprised if the Rams steal this game from them – or at least give them a good scare and cover the spread.
Computer’s Pick: Cincinnati (+6.5)
Bet Amount: $8.82 (Gamblor only)
COMMENTS: Yuck.  It’s such a pain in the ass to have to write about these games – both teams are going nowhere this season – so I won’t bother, except to say that I think Cleveland is the better team here and should cover the spread but probably won’t.
Computer’s Pick: Tennessee (+3)
Bet Amount: $21.59 (Gamblor only)
COMMENTS: Jacksonville just completely imploded.  Even though Tennessee’s defense is nothing special, they won’t have too much trouble stopping Luke McCown.  Jacksonville’s defense will be able to contain Chris Johnson for a while, but if Tennessee is smart and controls the ball well, the Jaguars will wear out soon enough and he’ll have some big holes to bust through.  The spread on this has moved (to Jacksonville -2), so my bet amount will change a bit, but I’m happy to have money on Tennessee here.
Computer’s Pick: Washington (+3)
Bet Amount: $10.31 (Consensus)
COMMENTS: Eli Manning looked pretty terrible during the preseason, but I think he’ll be getting it together in time for the regular season.  The Giants offense should be pretty solid, and I don’t think the Redskins have any chance in this game.
Computer’s Pick:Carolina (+7)
Bet Amount: $14.05 (Gamblor only)
COMMENTS: This is a tough call for me.  Arizona seems like they could really get things together and just tear the Panthers apart, or they could sputter and stall and just knock through a few field goals.  I’m not expecting anything much from Cam Newton; I know he controlled the ball well in the preseason but things will change in a big hurry now that the games actually count.  Look for him to make his first few turnovers here.  I really do think the Panthers haven’t improved at all, so I expect an easy win for Arizona, but I’m not supremely confident that it will happen.
Computer’s Pick:Seattle (+5.5)
Bet Amount: $19.86 (Consensus)
COMMENTS: Again, who cares?  Sometimes it’s fun to look at the English Premier League and see how they relegate teams to lower divisions for playing poorly; I feel like the NFC West is the same sort of thing for the NFC.  None of the teams play at the same levels as the rest of the conference.  I doubt that Seattle has any chance to win this game, but both teams are leaderless and it’s hard to imagine either one building up much of a lead.  In which case I’m glad to that Gamblor is taking the points.
Computer’s Pick:Minnesota (+9)
Bet Amount: $4.56 (Gamblor only)
COMMENTS: Chargers all the way here.  They’re going to wipe the floor with the Vikings.
Computer’s Pick:Dallas (+4)
Bet Amount: $5.29 (Gamblor only)
COMMENTS: While the Jets haven’t changed all that much since last year, I’m not sure they’re going to be all that fired up for this game.  Dallas might be.  With Felix Jones looking better than ever, I think Dallas can keep this tight.  I expect to see an upset, a slim Jets victory, or a push.  But I’m not anticipating the Jets can win this by more than 4.
Computer’s Pick:Miami (+7)
Bet Amount: $459.90 (Consensus)
COMMENTS: Last year I was furious when Vegas set a short line for the Miami-New England game in Week 17, opening up with the 13-2 Patriots favored by a single point at home against the Dolphins, who they had beat by a score of 41-14 in Miami earlier in the season.  Predictably, hordes of gamblers dumped everything they had on the Patriots, and Vegas panicked and moved the line by three points.  It was one of the rare occasions where the sportsbooks made a big – and costly – mistake.  And they’re up to it again here.  This line opened with the Patriots favored by only 3 points.  It’s not like Miami made any major upgrades in the offseason, or the Patriots lost any important players.  So of course, once again, everyone is betting everything they’ve got on the Patriots.  Which means that the books have predictably panicked again and moved the line all the way up to 7 (creating a fantastic window for a “middle” if you got a bet in at 3, incidentally, but I’ll discuss that in a feature column later this season once I’ve had a chance to run some more numbers).  I have so much trust that the Patriots are going to beat the living snot out of the Dolphins that I’m going to bet $459.90 on them.  I’ve always said that it’s Gamblor’s money to do with as it pleases, and this still holds true, but I can’t bear to throw more money away on the Dolphins.  So for this game, I’m the book.  If Gamblor is right, I’m going to give it an extra $459.90 to play with.  If it’s wrong, instead of losing its money to the sportsbook, it’s gonna have to give it to me.
Computer’s Pick:Oakland (+3)
Bet Amount: $21.78 (Gamblor only)
COMMENTS: I don’t think we’re likely to see a repeat of last year’s trouncing of the Broncos.  In fact, it wouldn’t surprise me if the Raiders completely reversed last year’s undefeated record in the division and went 0-6.  With the revenge factor at work, and also the fact that Oakland’s defense is completely and totally helpless to stop any kind of offensive play whatsoever, I think Denver will roll here.
Enhanced by Zemanta