It's a good thing that no one can duplicate that stub unless they have a reasonable facsimile thereof.

cfreak: Hmm quiet in here tonight. Where is everyone?

On the way to camp outside the casino with their newly created ticket just in case.

You know, CFP's (College Football Poll) rankings have consistently produced >.500 prediction ATS over the last decade. Granted, for every dollar you put in, you get ~1.05 out. Still, if you have 10K to blow, that's a lot of money compounded over the 14 weeks of CFB and bowl game season. Assuming you spread out your 10K equally amongst all the games, and that you lose your money in a loss, and double it in a win, you would have $14943.95 through week 7 of this season.

Of course, I'm not entirely sure how the payout scheme of sports betting works. I am assuming it's double or nothing. Even if it's lower, you should be coming out ahead.

In other news, absolutely dumb luck.

Pratty: video man: You know, CFP's (College Football Poll) rankings have consistently produced >.500 prediction ATS over the last decade. Granted, for every dollar you put in, you get ~1.05 out. Still, if you have 10K to blow, that's a lot of money compounded over the 14 weeks of CFB and bowl game season. Assuming you spread out your 10K equally amongst all the games, and that you lose your money in a loss, and double it in a win, you would have $14943.95 through week 7 of this season.

Of course, I'm not entirely sure how the payout scheme of sports betting works. I am assuming it's double or nothing. Even if it's lower, you should be coming out ahead.

It depends on the line.

ATS means against the spread.

His error is in assuming double or nothing. Everywhere I've ever seen sports betting (legal or otherwise), for a straight pick ATS you bet 11 to make 10 (any any multiple thereof). If you win, you're up $10. If you lose, you're down $11.

To break even, then, you'd need to win 11 out of 21 games, which is slightly more than 52%. I don't know how much that cuts into his supposed profit margin.

Your Average Witty Fark User: Dear Jerk: What kind of odds maker says 'this team is a 500/1 shot, but if they do make it, they will once again have only a 500/1 shot to beat the AL team. It's not like they'll be on a hot steak or something.'

You said "hot steak".

It's more like a lukewarm meatloaf, with those odds.

dartben: Pratty: video man: You know, CFP's (College Football Poll) rankings have consistently produced >.500 prediction ATS over the last decade. Granted, for every dollar you put in, you get ~1.05 out. Still, if you have 10K to blow, that's a lot of money compounded over the 14 weeks of CFB and bowl game season. Assuming you spread out your 10K equally amongst all the games, and that you lose your money in a loss, and double it in a win, you would have $14943.95 through week 7 of this season.

Of course, I'm not entirely sure how the payout scheme of sports betting works. I am assuming it's double or nothing. Even if it's lower, you should be coming out ahead.

It depends on the line.

ATS means against the spread.

His error is in assuming double or nothing. Everywhere I've ever seen sports betting (legal or otherwise), for a straight pick ATS you bet 11 to make 10 (any any multiple thereof). If you win, you're up $10. If you lose, you're down $11.

To break even, then, you'd need to win 11 out of 21 games, which is slightly more than 52%. I don't know how much that cuts into his supposed profit margin.

Well, the computer's percentage has been 53.4% through the last 7 weeks.

Recalculated, with your helpful information, the after the last 7 weeks one would've been sitting on $10790.50, given a 10k initial investment.

On the flip-side, after the 10-11 season, given a 10k initial investment, one would be sitting on much less 10K. It was also the computer's worst year ever.

Going back another season, you would've been sitting on 6250.26. Hm. While the compounding weekly idea seems good, it's not. The compounding went OK until week 14, where the computer only guessed .294 of them right. Ouch.

If you divided it out over every game of the season, which is a safer option, you would have winnings of xDollars for the following years:

2010: -$664.54
2009: $270.90
2008: -$130.00
2007: -$244.54
etc...

Ok, that's not good either. You only made money on one season. The computer seemingly does the bowl season usually very well though:

2010: $900.90
2009: $1225.45
2008: -$1008.18
2007: $137.27
2006: -$454.54
2005: $1588.18
2004: $5005.45 (HOLY SHIAT)
2003: $900.90
Total on (Easily) available data:
$8295.43 over 8 years worth of bowl games.

The site says it has bowl predictions dating back to 1993, but I can't find the seasonal one's before 2003. The percentage of correct ATS guesses in bowl games though 93'-10' is .541, which means one would average 328.18 in winnings every year if you bet just 10K on each bowl season.
That's, of course, assuming your 'Bet $11, Win $10 or Lose $11'. Anything worse, and you lose money.

/...and if my math is right. Which I'm pretty sure it is.

Yeah, that 10% vig is a killer.

I'm sure the computer probably does alright, simply because the spread isn't meant as an actual prediction of the final margin of victory, it's more of a prediction of what the public thinks will be the actual victory margin. Since the latter inherently involves taking into account irrational fans who will bet for their team without caring about the spread, while a computer doesn't, stands to reason it'd have to be a better predictor.

Always rememebr that the perfect spread from the casino's/bookie's point of view is one which will get equal money on both sides of the line, because they make their money on the vig (collect $11 from each side, pay out $21 to the winner, keep the left over dollar). They don't care if it's accurate as to the game itself.