…IF YOU’RE GOING TO PLAY A PARLAY
Smart, and profitable bettors know that parlays are a bad play. True, they’re alluring, and offer insane payouts, but the math has told us for a long time, they are “sucker bets”. Be it the lottery or any Casino, the entire gambling industry is built on the fact. However, this piece isn’t about berating those that make parlay wagers, but providing an insight on how to do it smarter. Think of it like the sex-ed class that showed you how to use a condom, vs the one that said you’ll burn in hell after your first handjob. The bottom line, IF you are going to play a parlay, play a 3 or 5 outcome parlay.
For those new, and lucky enough to have never been enticed by the temptress that is Parlay Betting a Refresher (veterans of parlay play can skip to the next paragraph):
A parlay wager is a singular bet on multiple outcomes of multiple events (2 or more), that all must succeed to payout winnings (ignoring progressive parlays, but that’s a tale for another day). In exchange for NEEDING to have ALL selections hit, parlays pay out at a higher rate than those bets individually. Think of it this way, say you were betting on the side of an evenly balanced coin flip. In this example you select “Heads” and would assume the payout of the bet to be about the equivalent as your stake, about 2:1 since you have a 50% chance of being correct (ignoring juice, or the fee to the sportsbook for creating the market between bettors, for now). But, what if you thought the next 3 flips would result in a “Head” and wanted to wager as such: you would be making a parlay bet, and since the odds of 3 continuous, unique “Heads” coming up occur 12.5% (½^3) of the time. With those odds, you would expect a higher payout, knowing that ANY “Tails” result of the 3 flips would mean that you lose. This is a parlay bet. Practically, in sportsbetting you wager on 2 or more outcomes that ALL must succeed to result in a “higher than x payout”, where x equals the payouts of each individual payout NOT TIED together, for you to collect your winnings.
Now, remember the best rule of life: People aren’t stupid, and books/casionos follow suit. This means that although you choose to wager a Parlay Bet, and are predicting MULTIPLE outcomes, expecting a higher return, those books/casinos know how to price them. That is, BEYOND the normal juice/vig you can expect from an ordinary bet, sportsbooks know the strength of the demand for “low-risk, high return” bets like parlays can be “juiced” an additional amount. So in other words, although parlays offer high payouts, given the low likelihood of success, they are priced BEYOND a typical vig of an individual bet, and never offer a positive likely outcome for the bettor–hence a sucker bet.
Below you can see what 4 common sportsbooks are currently pricing the respective parlays at, and what they WOULD be priced at if it were an evenly expected outcome. The first colum is number of teams in your parlay, the second the average current payout for a (roughly) 50/50 bet (-110/-100 bet), the third what the payout would have to be to give you a 0 net expected value (see below) and the fourth being how much money you are losing on every $100 bet over the long run.
|# Tm.s in Parlay||AVG PAYOUT||FAIR Payout||Net|
Most experienced bettors know this, but if you do not, STOP before ever making any bet before you read the following about expected value or EV:
Simply put, for ANY bet (not just parlays, but we will get back to that):
What am I expected to make, less what am I expected to lose. Mathematically this is broken down as follows:
(Probability of success x Payout) – (Probability of loss x Risk/Bet amount)
*Given all sportsbooks know this, you will never find a positive EV on a standard bet ASSUMING a “normal likelihood”. That is, completely even teams playing each other will never pay out 2:1 odds, because there will always be a fee/juice/vig, hence all of success in making money in sports betting is twofold:
- Predicting an outcome(s) at a better than the market (although the game is priced as an even matchup, but you predict one team has a 60% chance of winning, not 50%). Note: this is very difficult to do, even the best in the world hit at about 60% long term.
- Understanding of Estimated Values of every bet made, and ensuring a long term exposure is consistently in the positive spectrum
This brings us back to the problem at hand: Parlay Betting. As broken down above, NO parlay (assuming a normal predictability of outcomes) will ever have a positive EV. HOWEVER, oddly as the quantity of games wagered in a parlay increases the payouts do NOT increase in a geometric pattern. If you asked me why this is, I wish I had an answer, but with over 10 years of professional wagering, I have still never found an answer. The fact remains, there are BETTER parlays to play. See below, four large sports books and their respective payouts based on size of parlay:
|Parlay # Events||Sportsbook||Draft Kings||Fan Duel||Bovada||AVG PAYOUT||ODDS TO HIT||ODD TO MISS||EV||EV Change|
First look at the “EV” Column, and note that your expected value decreases with every additional event/wager you add/attach EXCEPT between 2 team parlay and 3 team parlay. Again, I have no idea why this is, but IF you wager a parlay bet, make it a 3-team parlay. Your EV is actually the highest of ANY parlay (including being better than a 2 team). Second, I added a column labeled “EV Change” or derivative for those that remember HS Calculus, to help see how the size of your parlay does not evenly/constantly increase your payout (geometrically), and is somewhat sporadic. Yes, for every team you slap on to your parlay card, you are expecting to lose more, BUT notice the 5-team parlay. The difference between EV is negligible at less than a penny on the dollar. In other words, if you are going to make a parlay bet, and you’re going to make a 4-team parlay, why not make it 5 when the change in EVs are almost non-existent?
Again, the moral of the story stands: Do NOT play parlays (like a mother telling a child not to open up “Daddy’s special collection of magazines”), but if you do, play 3 or (to a lesser degree) 5 teams, as they show more/derivative value when adding teams from the 2nd and 4th respectively.