MaskedChicken said:
The thing about judging wagers like that is that it's not completely about just expected value. When determining wagers you typically use the Kelly criterion (edge/odds) to assign an amount to a play. In this case, because the odds you're getting are 35 to 1, this severely reduces the negative value of the edge (which is (1/~700)*(36)-1=-.95) down to -.95/35=-.027.
So in terms of expected value, this is like an NFL spread you think wins about 3% of the time. But in terms Kelly criterion edge/odds, this is more like an NFL spread you think wins 50% of the time. You shouldn't make the wager because that's not profitable, but it's not as bad as it seems at first glance, and those 35 to 1 odds are pretty much where they should be for the books to have this prop on the board. Don't play it unless you think it happens more than 3% of the time though (which I doubt personally).
This is a very good point. When you deal with large favs/dogs, a big difference between two ML can seem like a big deal, but percentage wise it's really not. This is why when a book deals a 20 cents line on -110/-110 they have a 4.55% edge. But if they dealt -510/+490, it's still 20 cents but now the book only has a 0.55% edge. The line would have to be -583/+417 in order to keep the 4.55% hold.
This is why you see 20, and 30 cent lines in baseball as the lines get larger. a 10 cent line at -105/-105 is only keeping 2.3% as is. But if you play at Pinny, 5dimes, Betpop (Betjam used to have them) and can get -190/+180, they're only holding 1.22%. 5dimes 5 cent overnight lines at -153/+148 are holding 0.79%.