UPDATE: The only line that changed was the Raiders line – which moved down to +14.5 from +16, which means the computer thankfully wants to risk just a little less on the Raiders. Some of Gamblor’s bet amounts changed due to fluctuations in the models, but nothing more than a few cents either way. There’s STILL no line for the Seattle game, but the computer sees these teams as pretty evenly matched so it will be taking the points in this one regardless of who’s getting them. A final summary sheet is here.

Last week was the worst in Gamblor’s season so far, both in terms of weighted win percentage (29.9%) and money lost (-$70.33). My own picks were fine, going 8-6, but the computer’s finished at 5-9. Surprisingly, I don’t feel all that bad about it. The computer is going to have some bad weeks, and I’d prefer that they happen when there’s not all that much money on the line. Plus, its biggest pick (Detroit) was looking pretty damned good up through halftime and until about 4 seconds into the third quarter. I’m still up $120 for the season so far, and the season’s weighted win percentage is still at a perfectly healthy 57.9%, so I’ll shrug off last week and move on.

About once a week, the predictions of the Hedge model and the Flip model coincide and both models agree on an underdog in one of the games. This means that the outcomes from the various simulation schemes range from less than 55% to sometimes as high as 90% in favor of one team. But this isn’t an indication of supreme confidence, all that it’s actually showing is that the statistics for one (or both) of the teams are pretty volatile. So far this has turned up four times this year, and it’s gone 2-2. Nothing special. But the weighted win percentage is out of this world at 78.5% (both of the winners were big bets and both of the losers were little ones). If we look back through all of last season, the weighted win percentage was still pretty incredible — 71.1% over all seventeen weeks. This week, unless the spread moves quite a bit, both models will be agreeing on Atlanta as an underdog of 2.5 points at San Francisco. So with these kinds of percentages, why don’t I consider this to be a special situation and put down a huge bet on Atlanta?

Simple. The “Agree” bet doesn’t provide any real advantage. Sure, in 2008 it came in at 71.1% – but in 2007, it performed at a simply awful rate of 37.4%. In 2006 it was strong again, at 65.3%, but in 2005 it was a bottomless pit of despair at 36.2%. Throughout the lifetime of the data back to 1998, this particular situation performs at a weighted win percentage of 55.0%, almost exactly identical to the Gamblor’s overall performance with the weighing scheme I’ve chosen to use. The fallacy of betting based on — or against — trends is what I’ll be talking about in this week’s feature.

(To watch this video, right-click and hit €œplay€)

Top 5 NFL Betting Trends You Should Be Wary Of:

Trends are a very dangerous temptation in sports betting. In my opinion, paying attention to trends against the spread in the NFL makes about sense as looking for trends at a roulette table. They don't exist. If the ball has landed on black nine times out the last ten, the likelihood that it will land on red (or black) on the next spin is not altered in the slightest. For every arcane trend you can find that virtually guarantees a victory for one side, you can find another equally cromulent trend that will tell you the exact opposite. You should no more trust an ongoing trend to continue than you should bet against it under the assumption that Vegas has spotted the trend and is taking measures to reverse it. With that in mind, here are five NFL trends you should know about -- and then proceed to completely ignore:

5. Playing Surface. I haven't programmed Gamblor to make predictions about the over/under, but everything it needs is there. I'm just not sure whether it's possible to gain any kind of edge. A lot of gamblers believe that paying attention to field conditions can be pretty lucrative, particularly when concerned with the over/under. The 3-0 disaster between the Pittsburgh Steelers and the Miami Dolphins at a sodden Heinz Field in 2007 was a classic example of how a messy field can shut down scoring in a football game. Since the Arizona Cardinals moved into their new stadium in 2006, the games have been played on a grass field that grows outside in the perfect Arizona sun and then gets moved inside the dome for the games. It's a unique situation which has led to a 19-7 record for gamblers who bet on the over. This week, the Cardinals are hosting the Houston Texans. The Cardinals defense is remarkably porous, having only held opponents to less than 17 points just eight times in their last 67 games (from 2005 to present). Houston has a potentially explosive offense and a similar philosophy of indifference when it comes to stopping their opponents. The over/under on this contest is at 48 points, which is on the upper end of the range -- but still rather appealing considering the way these teams match up.

4. Teams Coming Off a Bye. When I first programmed Gamblor, I tried to give a lot of consideration to teams that were coming off a bye week. I thought that these teams would play better as a result of the rest, or that bettors would overcompensate for this and there would be an advantage in the opposite direction. I was wrong on both counts - statistically, the bye week had no impact whatsoever on teams that were playing, or the teams they were playing against. This week Atlanta, Philadelphia, Arizona, and Carolina all had the previous week off. Of these, Philadelphia has been red-hot in their return to action, going 11-0 straight up and 9-2 against the spread. Philadelphia has been looking great so far this season, but this trend is running headfirst into two other trends as the Eagles host the winless Tampa Bay Buccaneers.

3. Big Time Underdogs. There are only three or four games every year where the spread exceeds two touchdowns -- so the fact that we'll see two such games this week (Oakland at New York, Tampa Bay at Philadelphia) is quite an unusual occurrence. Shocking upsets are very rare in this situation, but a lucky play at the beginning of the game can be all it takes for the underdog to keep things close enough for the spread to factor in, and laying so many points as a favorite leaves a lot of room for a backdoor cover after the starters have finished playing. It's always very stressful to bet the underdog in what's expected to be such a lopsided game, but underdogs playing on the road that are benefiting from a spread of more than 14 points are 14-4 since 2005. Tampa Bay is getting a 15 point boost in Philadelphia, and the Raiders are getting 16 points against the Giants. I'll be nervous if the spread drifts any higher for either game because I'll interpret that to mean the books are trying to scare money off of the favorite -- but for spreads of 16 or more, the underdog has covered the last seven times in a row.

2. Winless Teams. There's a philosophy that teams who have made it four or more games into their season without winning yet tend to get great value against the spread, because nobody wants to bet against such a worthless crew of losers. Plus, the winless team will be playing their hearts out against an opponent that is most likely phoning it in. The philosophy behind this makes sense -- and the numbers back it up quite well. Since 2000, teams with records of 0-4 or worse are 52-28 against the spread. With a win percentage of 65.0%, that's a pretty seductive trend. This week Tampa Bay, St. Louis, Kansas City, Tennessee, and Cleveland will all be striving to lock down their first win. I wouldn't really say that any of them have a realistic chance of winning outright, but everyone except Tennessee is facing a spread of 9 or more (and the Titans line, which is only 3.5 against the Colts, looks an awful lot like a trap!).

1. Hello, Good-Bye. The incomparable Walter alerted me to this one, and it's pretty hard to ignore. It's been 2-0 so far this season, and it's been 33-6 since 2000. So what's this magnificent trend? It's to bet on favorites of 6.5 or more points in their last week before a bye. A spread of 6.5 or more points indicates that the favorite is expected to fully dominate the game. And with nothing on the immediate horizon to distract them, players can focus entirely on the game at hand. Furthermore, the players know that if their coaches see them dog it against a clearly inferior opponent, they'll be doing nothing but running drills for the next five days. This week, the Cowboys are headed to Kansas City to face the winless Chiefs and although the Cowboys' season has been pretty disappointing thus far, they can avoid a lot of punishment in practice the following week by putting together a solid win over the Chiefs.

Gamblor is loving underdogs again this week, and frankly, I'm glad. Favorites have been ruling the season so far (34-27 against the spread). It's very unlikely that we'll ever see a repeat of 2005, known as the "year of the favorite," when favorites covered the spread at an astonishing rate of 55% and many sportsbooks flirted with bankruptcy. So you can expect the underdogs to bounce back with a vengeance as the season wears on. Here's the summary of bets for Week 5. Good luck!

CLEVELAND BROWNS at BUFFALO BILLS Computer's Pick: Cleveland (+6) Bet Amount: $13.00 (Hedge)

WHY IT'S WRONG: Just because Cleveland kept things close with Cincinnati doesn't mean all that much. They're still one of the worst five teams in the league, while at this point I consider the Bills to be simply average. Cleveland played a full extra quarter against the Bengals and lost, and they'll be flat coming into Buffalo. The Bills are no doubt looking to bounce back after last week's embarrassing loss to the Dolphins, and should have enough talent to get the job accomplished against the Browns.

DALLAS COWBOYS at KANSAS CITY CHIEFS Computer's Pick: Dallas (-8.5) Bet Amount: $7.73 (Flip)

WHY IT'S WRONG: It seems to me like Kansas City is getting way more points than it deserves here, and Gamblor agrees. The only team that Dallas truly dominated was the Buccaneers in Week 1, and that was at home. While I warned against paying too much attention to trends above, I don't think that Kansas City is going to let Dallas walk into their stadium and thoroughly manhandle them. The Chiefs will lose, but they'll make it close.

MINNESOTA VIKINGS at SAINT LOUIS RAMS Computer's Pick: St. Louis (+10) Bet Amount: $2.71 (Hedge)

WHY IT'S WRONG: Gamblor's affinity for St. Louis is such a huge pain in the ass. It has bet on them every single week, and they're 1-3 against the spread so far. They haven't actually been all that expensive, but it's no fun to keep pissing away money on them. I'm glad I get to round this one down and save my money this week. Minnesota looked absolutely fantastic against Green Bay, and if Mark Bulger manages to leave the field on his feet and not on a stretcher he should consider himself very lucky. It's true that Minnesota will be looking ahead to more worthy opponents, but they could put together a team composed of the players' wives and it would still be enough to beat the helpless Rams by two touchdowns.

OAKLAND RAIDERS at NEW YORK GIANTS Computer's Pick: Oakland (+14.5) Bet Amount: $20.04 (Hedge)

WHY I AGREE: How pathetic is your team when a computer model considers a 16 point spread to be just about right? There's a million things wrong with the Raiders right now, and they'll be headed into New York with absolutely no plan whatsoever. Nobody in their right minds would take the Raiders, right? Well, considering that the Giants are headed to New Orleans in a week, there's no way their minds will be focused on this game. If the Giants were facing another weak opponent next week, or had a bye, I'd feel comfortable laying the points. But with the unstoppable Saints on the horizon, there's no way they'll be able to focus on this game properly.

TAMPA BAY BUCCANEERS at PHILADELPHIA EAGLES Computer's Pick: Tampa Bay (+15) Bet Amount: $0.90 (Hedge)

WHY I AGREE: The Eagles have looked great so far, but it will take them the first quarter to get their heads together after the bye. I don't really see it mattering much whether Donovan McNabb or Kevin Kolb is under center, Tampa's biggest problem will be figuring out how put points on the board against Philadelphia's defense. It's hard to imagine that Tampa Bay will be able to get anything accomplished, but with all the injuries Philadelphia is facing I expect them to play a tight, conservative game and once they have a lead they'll keep the ball on the ground and play things safe.

PITTSBURGH STEELERS at DETROIT LIONS Computer's Pick: Detroit (+10.5) Bet Amount: $21.66 (Hedge)

WHY IT'S WRONG: I'm glad that Detroit got that one win under their belt, because I don't foresee them getting another one for a while. After playing a tough first half against the Bears, their special teams fell to pieces and their star quarterback got hurt. They stand absolutely no chance against Pittsburgh.

WASHINGTON REDSKINS at CAROLINA PANTHERS Computer's Pick: Washington (+3.5) Bet Amount: $23.47 (Hedge)

WHY I AGREE: The Redskins and the Panthers are passing each other on the downslope of mediocrity, so it's tough to forecast who is going to want this game less. Since I've got to make a pick, I'll go with Washington, but only for the sake of getting the points.

CINCINNATI BENGALS at BALTIMORE RAVENS Computer's Pick: Cincinnati (+8.5) Bet Amount: $27.62 (Flip - 90%)

WHY I AGREE: Cincinnati had a tough time with Cleveland, and they're probably exhausted. But they'll be playing extra hard here as a way of saying "suck it" to their faithless coach Marvin Lewis, who had to be persuaded by the team to let them take a shot on 4th and 11 at the Cleveland 40 with a minute left in overtime last week. Gamblor thinks this line is a trap, and I actually do too. Given how complete of a team Baltimore has been, especially at home, they should be able to cover a larger spread than this. I think there will be a surprising number of tight games this week, and this will be one of them. Don't be surprised if the spread drifts down to 7.5 by the end of the week to induce more action on Baltimore.

ATLANTA FALCONS at SAN FRANCISCO 49ERS Computer's Pick: Atlanta (+2.5) Bet Amount: $18.05 (Hedge and Flip)

WHY IT'S WRONG: This is another tough game to call. I don't think Atlanta's offense is strong enough to put up many points against San Francisco, but I don't think San Francisco has many weapons to work with, either. I'm pretty enthusiastic about the 49ers this year, so I'll let that cloud my judgment and just take San Francisco here, damn the points.

JACKSONVILLE JAGUARS at SEATTLE SEAHAWKS Computer's Pick: Jacksonville (+3) Bet Amount: $17.87 (Hedge)

WHY I AGREE: I think that my editor Jeff has accepted that the Seahawks really are going to be bad this year (with or without Hasselbeck). Jacksonville, on the other hand, has been looking better every week. This line is dependent on whether Hasselbeck plays, which he probably will. It seems like a gift to get these points here, but I'll just gratefully accept them and I won't be looking back.

HOUSTON TEXANS at ARIZONA CARDINALS Computer's Pick: Houston (+5.5) Bet Amount: $19.86 (Hedge)

WHY I AGREE: It's really hard to know what the hell Houston is going to do during any given week. The last time these two teams played the Texans won (one of their only victories in 2005), so it's possible that Arizona will be motivated by a revenge factor. Houston has many more offensive threats than the Cardinals do, so in the expected shootout I think they'll have the edge.

NEW ENGLAND PATRIOTS at DENVER BRONCOS Computer's Pick: Denver (+3) Bet Amount: $14.44 (Hedge)

WHY I AGREE: I'm wondering that it will take for Denver to earn themselves a measure of respect. Certainly their victory in Week 1 was a fluke, but until someone beats them I'm going to give them all the credit they deserve. They've got to be royally pissed that they're hosting a team with a worse record than their own and are still considered an underdog, so you're going to see Denver players hurling themselves all over the field at the unwelcome Patriots just to make a point. I expect them to come into this game with a ton of intensity and dominate a New England team that only barely survived a scare from Baltimore last week and is still bruised and sore.

INDIANAPOLIS COLTS at TENNESSEE TITANS Computer's Pick: Tennessee (+3.5) Bet Amount: $14.08(Hedge)

WHY IT'S WRONG: Despite this sucker line, I'm still all over Indianapolis. They are making virtually zero mistakes right now and until I see them start to falter, I'm with them all the way.

NEW YORK JETS at MIAMI DOLPHINS Computer's Pick: Jets (-2) Bet Amount: $11.73 (Flip - 90%)

WHY I AGREE: Despite losing last week, it's important to remember that the Jets defense held Drew Brees and the potent New Orleans offense to just ten points. Miami, by contrast, looked like a completely different team than in their first three weeks as they obliterated Buffalo. Their luck will run out here, and will be gleefully gobbled up by Mark Sanchez. I'm glad that Gamblor is going with the Jets here, because I'm pretty excited about them too.

For legality purposes, this website does not promote or advocate gambling. This is solely for entertainment purposes only.

Tags: Football, Gambling, gamblor, NFL, NFL picks, Yup it turns out Seattle really does suck, Zack Dennis

This entry was posted on Thursday, October 8th, 2009 at 7:30 am and is filed under Featured, Featured - Frontpage, Football, Sports. You can follow any responses to this entry through the RSS 2.0 feed.You can leave a response, or trackback from your own site.