Needed...help with NFL Pick'em Picks

The Game Room
1

I volunteer for a charity, and a friend entered my charity to compete in a Pickem Pool, which has a very nice payout. We could sure put the money to good use, so I am hoping to give our charity the best shot at winning.

I plan to post the match ups, then use a consensus pick for our charity. My competition is mainly horseplayers, who only slightly follow football, so with some expert SDMB help, I think my charity could be the surprise winner.

It is straight up, spreads don’t matter, I just need the winners. What I would like to do is ask each week, then take note of who tends to get the most right, then give more weight to those posters picks as the season goes on.

Here are this week’s match-ups. Next week I will post the selections earlier, but I just didn’t think of asking for help here until now. The deadline for picks this week is Friday, 4:00 PM eastern time.

Sunday, September 14
Tennessee (1-0) at CINCINNATI (0-1) -1.5 Matchup
Oakland (0-1) at KANSAS CITY (0-1) -3.5 Matchup
New Orleans (1-0) at WASHINGTON (0-1) -0.5 Matchup
NY Giants (1-0) at ST. LOUIS (0-1) +8.5 Matchup
Buffalo (1-0) at JACKSONVILLE (0-1) -5.5 Matchup
Green Bay (1-0) at DETROIT (0-1) +3.5 Matchup
Indianapolis (0-1) at MINNESOTA (0-1) +1.5 Matchup
Chicago (1-0) at CAROLINA (1-0) -3.5 Matchup
Atlanta (1-0) at TAMPA BAY (0-1) -8.5 Matchup
San Francisco (0-1) at SEATTLE (0-1) -7.5 Matchup
New England (1-0) at NY JETS (1-0) -1.5 Matchup
Miami (0-1) at ARIZONA (1-0) -6.5 Matchup
San Diego (0-1) at DENVER (1-0) +1.5 Matchup
Pittsburgh (1-0) at CLEVELAND (0-1) +5.5 Matchup
Monday, September 15
Baltimore (1-0) at HOUSTON (0-1) -4.5 Matchup
Philadelphia (1-0) at DALLAS (1-0) -6.5 Matchup

Thanks for any help you can give me!!

  • If you are curious, the charity takes in retired thoroughbreds, evaluates what kind of life would be best for them, then retrains them for a new career, ( or nice pasture with if that is best) and then places them in a well qualified home where they will be safe, and well cared for.
2

To be honest, the best thing you could do to improve your chances of winning (especially if you’re not a football buff, but even if you were) is just to pick all of the favorites every week, and it’s not close.

How many participants are there? The fewer there are the more effective this strategy figures to be – in fact, if there were a half dozen or fewer, and your competition all followed any other strategy, I’d think this would be a near lock. If there are scores and scores of players, however, you’ll be a longshot to win no matter what you do, and the fact that at least a few other players will be doing something similar would lessen the effectiveness. With a league that big, you might have to pick some longshots to increase your variance.

Is there money for finishing in 2nd or 3rd?

3

whoops

4

Sure, I’ll play. I’m curious how well I can do.
Do you want commentary or just picks?

Tennessee (1-0) at CINCINNATI (0-1) -1.5 Matchup
BENGALS capitalize on Vince Young’s injury and redeem themselves.

Oakland (0-1) at KANSAS CITY (0-1) -3.5 Matchup
RAIDERS don’t stink as bad. Won’t be pretty.

New Orleans (1-0) at WASHINGTON (0-1) -0.5 Matchup
Drew Brees and the SAINTS take down the 'Skins

NY Giants (1-0) at ST. LOUIS (0-1) +8.5 Matchup
GIANTS, in the lock of the week.

Buffalo (1-0) at JACKSONVILLE (0-1) -5.5 Matchup
JAGUARS in a tough game.

Green Bay (1-0) at DETROIT (0-1) +3.5 Matchup
PACKERS all the way.

Indianapolis (0-1) at MINNESOTA (0-1) +1.5 Matchup
Peyton’s pissed after last week. COLTS stomp 'em.

Chicago (1-0) at CAROLINA (1-0) -3.5 Matchup
I’m liking the BEARS here, the line and the D look good.

Atlanta (1-0) at TAMPA BAY (0-1) -8.5 Matchup
I wouldn’t bet this spread. I’d take the under (38)
Closer game than you’d think, but the BUCS take it.

San Francisco (0-1) at SEATTLE (0-1) -7.5 Matchup
SEAHAWKS, Hasselbeck’s gonna light that secondary up.

New England (1-0) at NY JETS (1-0) -1.5 Matchup
I hate to bet against Favre, but the PATRIOTS are still a fantastic team and coaching and disclipline will win the day.

Miami (0-1) at ARIZONA (1-0) -6.5 Matchup
CARDINALS. Tough luck for the Fish.

San Diego (0-1) at DENVER (1-0) +1.5 Matchup
Denver looked real good, and Merriman’s gone. BRONCOS.

Pittsburgh (1-0) at CLEVELAND (0-1) +5.5 Matchup
STEELERS will win by 2 touchdowns.

Monday, September 15
Baltimore (1-0) at HOUSTON (0-1) -4.5 Matchup
RAVENS pull out an unexpected road win.

Philadelphia (1-0) at DALLAS (1-0) -6.5 Matchup
COWBOYS are the team to beat in the NFC. This will be a shootout.

My user name has absolutely nothing to do with gambling or sports picks. I can barely choose what I want for breakfast.

5

Thanks VarlosZ,

There are only 10 participants. Your strategy does make sense, I would be curious if the numbers back it up. I only question it because if it was very likely that would work, I think everyone else would be doing that too. And designing a contest like this, not involving the spread, would end up with everyone tying.

Sort of like with horse racing, one third of the favorites win. So a contest where you just were trying to pick the most winners it would be smart to just take all the favorites, and be assured of hitting one third.

But the more advanced players would be able to look at which favorites were “false” favorites, or know that the odds of a 2-5 favorite winning were up to 60%, and in another race, the favorite being 3-1, had less of a chance of winnning. The rule of one third of favorites winning is constant, but the advanced players know when betting against a favorite is the better way to go.

So maybe a strategy would be to go with all the favorites where the spread was very large, and then get some “expert” opinions on the games where the spread was very close, and try to hit some of those where the odds makers were wrong.

I will start counting up the number of favorites that win, each week. And then see if I can find a pattern to when the favorite doesn’t win, and add that into the mix. It might be that odds makers put too much emphasis on home field advantage, or not enough emphasis on key players out with injuries, or something like that. I am also a horse player, so out of habit I guess I am trying to figure out when the favorite more likely be beat. That may not correlate at all to football odd makers though, so I will analyze how they do.

I was thrilled to see someone named Picker replied to my thread…then I read the small print. :frowning:

Thanks for your insight. I am going to look at your commentary and see which games you have picked against the favorites, and use the ones you feel strongly about.

So this week I think I will try a combo of these two suggestions. Go heavy on the favorites, but add a couple slight underdogs who look to have a reason to surprise…
Thank you both so much. I will let you know what the final picks are, so you can follow along…

I do appreciate your taking the time to help…

6

Here is what we went with:

Cincinnati
Kansas
New Orleans
New York
Jacksonville
Green Bay
Indianapolis
Carolina
Tampa
Seattle
New England
Arizona
Denver
Pittsburg
Houston
Dallas

7

The teams I pick to win:
Tennessee (1-0)
Oakland (0-1)
New Orleans (1-0)
NY Giants (1-0)
Buffalo (1-0)
Green Bay (1-0)
Indianapolis (0-1)
Chicago (1-0)
TAMPA BAY (0-1)
SEATTLE (0-1)
NY JETS (1-0)
ARIZONA (1-0)
San Diego (0-1)
Pittsburgh (1-0)
Monday, September 15
Baltimore (1-0)
DALLAS (1-0)

8

Not many people will choose to do it because it’s not fun. It’s a lot more interesting to substitute your own judgement than to open the paper on Thursday and pick every team with a minus sign next to its name. Also, most people are under the delusion that they actually know better, or that there hunches are usually correct, or that they’ve found a system that works – true for a small number, but a mistake for the vast majority.

In theory this is fine, but the problem is that the spread already takes into account a wide range of “expert” opinions, as well as a very wide range of non-expert opinions (which can actually be an even better predictor of future events). Everyone who bets football is trying to come up with a system that tells them when the point spread is wrong, and only a tiny percentage (definitely under 5%, probably under 2%) are able to do it profitably. And if you did find something new and reliable . . . well, screw this contest, spend a year in Vegas and print money at the casinos. :slight_smile:

If you really want to try and come up with something, however, I have two suggestions. First: point-spreads and results for every NFL game going back to 1993. Trying to find patterns based only on data from this season would be futile due to the small sample size.

Second, DVOA (Defense-adjusted Value Over Average), courtesy of Football Outsiders. They take every play from every game, then measure it’s success or failure against the average result for plays in similar situations (down & distance, score, field position, time remaining, caliber of opponent, etc.). Again, this is to get around problems of sample size. Instead of having 16 binary data points on which to base judgements, this looks at thousands of plays for every team, each with dozens of possible outcomes.

I have had success for several years now using DVOA to pick games against the spread. What you might do is take the games with point spreads of less than 3, then look to see if DVOA says the underdog is significantly better than the favorite (and factoring in some common sense things like home field advantage and major injuries). If you do try this, early in the season you’ll want to pay attention to “DAVE,” which is a weighted average of the current season’s DVOA and the site’s preseason projections. Later in the season, you’ll want to pay attention to “Weighted DVOA,” which more heavily weighs recent performance. If you have any questions about this feel free to ask – I’m very familiar with their numbers and how one might use them for gambling purposes.
However. Having said all that, I still have to reiterate my suggestion that you simply follow the spread if you care more about winning than about having a good time competing. No offense, but the odds of you (or any given individual) finding an exploitable inefficiency in the odds are pretty slim, and even you did manage to find one, the margin – over a single season, for your purposes – would be tiny, to the extent that you’d be almost as likely to lose ground exploiting your edge as to benefit. Trust me, there’s no low-hanging fruit out there; it’s all “take my 1% edge and do it a thousand time over five years” stuff.

9

I just wanted to apologize for not responding to this post earlier. I had thought I would the day after it was posted, but Hurricane Ike interupted. We lost electricity for 11 days, and have been busy since with clean-up and repair work.

I had a friend submit my picks for me while I was offline, and followed your suggestion of going with the favorites. Your reasoning convinced me to give it a shot. I just now had a chance to check on the contest, and I won two weeks ago, and tied for the lead last week. ( There are also small prizes for each weekly winner, ties become a push). This week I am tied with one other for the lead, so tonight’s game will decide the weekly winner.

I haven’t had time to really look at the numbers yet, but I will be sticking to your plan, it really is working. When I get caught up, I will spend some time on the sites you linked. That first one fits in very well with horseracing, and how to find value when betting, knowing when “the crowd” might be wrong, so I look forward to learning more about that.

Thanks again for your help. Our charity has already won a small amount of money, and is really in a good position so far to take the whole thing. I would have never have thought that going in. But so far so good… And not having to spend time analyzing everything works great for me now, since my time is needed elsewhere until we get things straightened up around here. So using your plan has been ideal.

Thanks so much for your input. I will let you know how we do. Should we win, I will give credit for the donation in part to VarlosZ!