Many smart people view parlays as a bad bet. Well they are, just like every other bet, if you do not a have positive expected value.
Here is the expected value calculation for someone who picks about 50%.
Lets use a 100 unit bet per game. They will win 50% of the time and lose 50% of the time. When they win they will win 100 and when they lose they will lose -110. 100*.5 + -110*.5 = -5(EV). What this says is for every 100 units you bet you can expect to lose 5. Over the course of 200 bets in the NFL you can expect to lose 1000 units.
Now if that same player plays a 2 team parlay the results to be expected are as follows.
The probability that a 50% player wins 2 of 2 games is .50*.50 = 25%. Lets say the payout will be 2.6/1 on a two team parlay (it may vary slightly depending where you play it) So 100*.25*2.6 + -100*.75 = -10. For every 100 bet the player can expect to lose -10 on each bet. So that is twice as bad as betting the games straight. Below you can see the expected value for each parlay for the 50% bettor based on 100 unit bets
3 team: -13 units
4 team: -17 units
5 team: -20.75 units
So the more games we add to a parlay the lower the EV becomes. In other words the book take a larger percentage of each bet the more teams you put into a parlay.
Now lets assume you pick at 60%. That is a very high number but one that we have come very close to over 400 plus games.
Our expected value on a straight bet of 100 is 100*.60+-110*.4 = +16. That is 21 better than the 50% better. A huge difference. How do we do on parlays?
2 teams we expect to win 36% of the time with a payout of 2.6/1. 100*.36*2.6 + -100*.64 = 29.6
3 team: +52.5
4 team: +72.1
5 team: +97.2
So you can see that we can flip the tables on the sports books if we can pick at a 60% clip. Now picking at a 60% clip is extremely difficult to do, but the numbers work at 55% also. Betting parlays can allow us to bet smaller amounts with a better EV.
The real discipline comes in the fact that you are not going to win as many bets and you will have more losing weeks than if you play every game straight up. But the winning weeks will be magnified to make up for it. Now you if you go and play 10+ game parlays you may have a positive EV but the chances of hitting will be very slim. At a 60% win rate you can expect to hit that bet 1 in every 58.5 times you play it. Those are fantastic numbers when you consider the payout is much higher than that but you have to remember you may go multiple seasons before you hit one.
Much like in the Kelly system you need to know that you have a proven track record for performance. Once you are confident in that record then you can play some parlays. Parlays are a great way to use a small bankroll and build it quickly. Again provided you are picking at a win rate you have in the past.
Using the Kelly is a good way to determine how much you should wager on each parlay. If you are a flat bettor of 2% of your bankroll and want to add parlays in to your money management I would lower my bets to 1% flat and play 2 team parlays at .25% of my bank. 3 team parlays I would drop to .15 or .10%. This may be conservative but remember if you are flat betting the same games you are parlaying you have added to your risk. If you are following my system you may have multiple 2 team parlays with the same team. If that team loses you will be losing more than the 2% flat bettor. You may want to not flat bet at all and just play parlays. Adjust your strategy accordingly.
Refer back to these articles and build your game plan for the season. The teams are doing that now and so should you. You will have a lot more fun and have a better chance at profitability if you get disciplined.
If you have a specific question in regards to how you are going to plan for this season or on a specific game or wager feel free to email me Statsational@gmail.com and I can try and be of assistance.
Follow on twitter @statsational