Would you hedge USC, if you were up on them from pre-season?

[Black Aces 518]

whipsaw- Re: USC #1 in the polls. Yep. And Texas gains first-place votes every week. And I meant feeling among fans and commentators, which isn't factored in directly, but makes vote-changing more and more likely. Texas could win eleventy billion to zero and some people are not going to remove USC from the #1 spot until they lose. But they are certainly playing the best and are the only team in the top 10 in both offense and defense.

As far as your location shot, I'm an Oklahoma fan, not a Texas fan, and I'm being objective about it. I'm not a "homer" when it comes to sports analysis. I just bet on Baylor to cover vs OU last week. Texas would have to be WAY behind Va Tech in the computer rankings to overcome the substantial lead they have in the human polls, which count 2/3 of the formula.

[whipsaw]

[ QUOTE ]
eleventy billion

[/ QUOTE ]

Wow, that's the exact # of units I've lost on baseball this playoffs. [img]/images/graemlins/grin.gif[/img] [img]/images/graemlins/grin.gif[/img]

[ QUOTE ]
Texas would have to be WAY behind Va Tech in the computer rankings to overcome the substantial lead they have in the human polls, which count 2/3 of the formula.

[/ QUOTE ]

They will be WAY behind by the end, if VaTech beats BC, Miami, and someone in the ACC Champ game while UT plays girls lacrosse teams the final five weeks. Texas will lose votes to VaTech, especially after the Miami game, enough for VaTech to make up ground in the human polls.

[llabb]

Thanks for the good input in general, guys. Some things I like here, some I'm not sure about.

[ QUOTE ]

I would definitely go with Texas +222 as their hedge.
It depends on what you think the chances are of USC losing before the RB. IMO, not enough to pass up +222 for +152.

[/ QUOTE ]
It's not just the risk of USC losing before the Rose Bowl. It's also the risk that Texas is not the one to play them. You may feel that risk is miniscule, but the board seems split on what happens if VT also wins out.

How is the math below?

(1) Likely lines imply that the chances of USC winning out are somewhere around: 97% against Wash St., 92% against Stanford, 87% against Cal, 92% against Fresno St. 80% against UCLA. (I actually think their chances against Cal and UCLA are a little worse, but I gave SC the benefit of the doubt for this exercise).

Multiplying all of those winrates only gives them a 57% chance of winning out. Is that the right way to calculate their chances of being undefeated?

(2) Meanwhile, on another thread, posters are estimating Texas' chance of winning out at 85-89%. Even if they do win out, there's still the VT risk. Let's say there's a 50% chance of VT winning out, and if they do, a 25% chance that they pass Texas in the BCS (although some posters here seem to think that chance might be closer to 50%). That makes a 12.5% chance Texas is not in the Rose Bowl, even if they win out.

So Texas' chance of being in the Rose Bowl is their likelihood to win out (87%) times the likelihood of not being in the Rose Bowl even if they do win out (87.5%), which equals 76%.

(3) USC must win out, and Texas also must be in the Rose Bowl, in order for the Texas hedge to be better than the USC "No" hedge. Thus, the risk of either of those happening is 1-(0.57*0.76)=57%.

(4) Okay, here's where I don't know how to calculate things. It seems clear to me that if the Texas hedge loses 57% of the time, then the USC "No" hedge is much better. But what is the math behind this? How do I calculate how often a bet needs to win, in order for +222 to be better than +152?

[Black Aces 518]

OK, well let's look at the math of it. Would you agree that the worst-case scenario for Texas in the computer polls would be an average of 4th place? That's assuming they are behind every unbeaten team at the end of the year, and that there are 4, which is the maximum. And the best-case for VaTech is 1st in every one? So that would be a spread of .120 in the BCS (1.000 for VaTech's computer component, .880 for Texas'). So VaTech would need to be at least within .06 of Texas in each poll. They are currently .05 and .04 behind. So it is close with the polls as is, assuming a best case scenario for VaTech in the computers. It is unlikely that SC or Georgia/Bama will pass UT in the computers. If Texas can maintain #2 rankings in at least 6 of the computers, VaTech would have to gain immense ground in the polls. Even if every computer poll had it Va Tech at 1, SC at 2, Texas at 3, Texas would still lead Va Tech right now due to the human polls.

[Black Aces 518]

llabb,

I would say that you are overstating SC's chances of losing any of their remaining games. Cal is too low, they are not a good team. I'd put UCLA more in the low 90s area. Fresno 95, Cal at 95, Wazzu at 99.9.

Same for Texas. OSU is 99.999999, as is KU. Baylor is 99, and A&M is 97. Big XII title game is 95.

I gave the math on VTech overtaking Texas. Barring major struggles by Texas and dominating wins by VTech, it's just not happening IMO.

[llabb]

[ QUOTE ]
There are three reasons to make a bet that hedges a previous bet:

1. The second bet - the hedge - has positive expectancy itself.

2. Your risk is so large, that you'd get sick to your stomache (or your wallet) if you lose.

3. The hedge itself allows you to play other related plays (that have positive EV) that you found too risky to play combined with your current position. But the combination of this new positive EV play and the hedge, must have positive EV.

In your case, it seems the only relevant issue is #1 - does
the hedge have positive EV. So, would you bet against USC now if you didnt' have the USC bet to begin with?. I'll leave that answer to you and other posters.

[/ QUOTE ]
Great post, King Yao. I think there could be a couple other reasons as well:

4. Your estimation of the odds have changed, in the time elapsed between the original wager and the hedging opportunity; or your assessment of which side of the bookmakers' odds you prefer to be on has changed.

For example, if I thought USC winning the Rose Bowl was +EV at +175 pre-season when I took the wager, but no longer think they are a good bet at -170, I would seek to hedge it, and lock in the winnings I've gained from the odds differential. Even if your hedging wager were slightly -EV, the resulting combination could be +EV compared to the EV of letting the original wager ride. That is what I am contemplating now.

5. To reduce variance. Even if I am comfortable with the risk, and losing the amount wagered would not make me sick, I might be more comfortable reducing volatility, especially when I am booking a win and not simply hedging a loss.

I believe that few sports bettors actually wager less than 1% of their bankroll per play, in the disciplined manner that many suggest should be done. Even for those who are disciplined with their poker bankroll, and do not play above their limits, it seems that they are often wagering a higher % of their bankroll than they should.

This leads to a higher risk of ruin. Longer odds, by definition, mean more likelihood of loss. Hedging to book a win, while forgoing profit potential, also minimizes risk, reduces variance, and enables you to more steadily grow your bankroll, perhaps leading to higher unit amounts and greater future earnings.

[llabb]

[ QUOTE ]
llabb,

I would say that you are overstating SC's chances of losing any of their remaining games. Cal is too low, they are not a good team. I'd put UCLA more in the low 90s area. Fresno 95, Cal at 95, Wazzu at 99.9.

Same for Texas. OSU is 99.999999, as is KU. Baylor is 99, and A&M is 97. Big XII title game is 95.

I gave the math on VTech overtaking Texas. Barring major struggles by Texas and dominating wins by VTech, it's just not happening IMO.

[/ QUOTE ]

I simply took the lines for this week's games, and calculated the win probability. USC moneyline is -4000, while Wash is +2800, which means the linesmakers' estimation is for USC to be -3400, or 34:1. 34/25=97%.

For UCLA, I looked at current lines on games. A line of -7 translates to -265, or 2.65:1 = 73%. A line of -14 translates to 5.35:1 = 84%. USC vs. Notre Dame's line was somewhere between there, if I remember right. UCLA's line will be even closer, if they are still undefeated when they meet.

This is a more accurate estimation of win %'s than estimating them yourself. It is very difficult for most individuals to distinguish the difference between a 95% and 97% likelihood of winning.

Your 99.9% for USC over Wash St. means that Wash would not win 1 out of a 1000 times. There are enough intangible variabilities alone to cover that. And your Texas-OSU at 99.999999% is an obvious overexaggeration.

------------------------

Your math for the BCS standings, however, I am very interested in. For the posters who favor V-Tech, do you have any input on Black Aces' post?

[MCS]

[ QUOTE ]
OSU is 99.999999, as is KU.

[/ QUOTE ]

No they're not, and while I know (hope?) you're joking, it actually matters if you change 99.999999 to 99, which IMO is still too high.

[King Yao]

[ QUOTE ]
4. Your estimation of the odds have changed, in the time elapsed between the original wager and the hedging opportunity; or your assessment of which side of the bookmakers' odds you prefer to be on has changed.

[/ QUOTE ]

I think this is in Rule #1 that I wrote. If you think the true odds have changed, you shouldn't be betting the hedge unless it has zero or positive EV.

[ QUOTE ]
For example, if I thought USC winning the Rose Bowl was +EV at +175 pre-season when I took the wager, but no longer think they are a good bet at -170, I would seek to hedge it, and lock in the winnings I've gained from the odds differential.

[/ QUOTE ]

Let's assume you bet it at +175, and think it is now worth -160. You wouldn't want to bet it again if you saw USC -170, but at the same time, you should not be looking to bet the hedge unless you could get fair value: +160.

This is not a case where the risk is all that great. It's close to an even money bet. If the risk is really too much for you to stomache right now, then you overbet your bankroll in the beginning of the year.

[Black Aces 518]

[ QUOTE ]
[ QUOTE ]
OSU is 99.999999, as is KU.

[/ QUOTE ]

No they're not, and while I know (hope?) you're joking, it actually matters if you change 99.999999 to 99, which IMO is still too high.

[/ QUOTE ]

I was joking. I do think 99 is about right for KU, but still too high for OSU. They are AWFUL AWFUL. They had to escape Montana St (I-AA) and Arkansas St in their own home stadium. They have lost every Big XII game by at least 27, save the Missou game, where they rallied from a giant deficit to make the final respectable. Plus, their starting QB is out. Their backup has 11 turnovers in 2 games against mediocre defenses, and now faces one of the best defenses in the league.

Put it this way, I wouldn't bet my life that Texas beats Okie St, but I'd bet my house.