Is it possible to have a parlay lose one leg and win the other leg (in a 2 teamer) less than 50% of the time? Consistently?
I'm looking into something that could be very profitable. Say my parlay is correlated with the game (ie: If duke can cover the spread, it will need to be a high scoring game, so I'll play duke -the points and the over) vs. if the other team covers the spread, it will likely be a defensive game so I'll play that other team and the under.
What I'm looking at is not to play one of each as I explained above, but instead to bet duke's spread, and the over 2.2 units to win 2 for both plays. Then also parlay the opposing team and the Under. This would pay out 13/5. so a 1.7unit bet on this parlay to win 4.4 units.
Total units wagered = 6.1
If both individual plays win, 4 units (and the original 4.4 units wagered) are payed out. 4units - 1.7unit parlay = a profit of 2.3units
If the parlay wins, 4.4 units (and the original 1.7units wagered) are payed out. 4.4 units - 4.4 units = even money.
This model could be modded to make the parlay and the individuals pay out the same, but I think it's better to have the individual bets pay out more.
What I need help for, is the math and the likelihood of losing 1 leg, and winning the other.
this split means there are still 6.1units wagered, but we only get a payout of 2 units. This is a loss of 4.1 units.
Can this model be made so to reduce the chance of this split in such a way that it will happen 33% of the time or less?