Heavy Favorite Parlays: We Walked Six Seasons and They Lose

The pitch is everywhere: parlay a few heavy favorites, cash most nights, bank a slow and steady profit. We tested it on six seasons of MLB and NHL moneylines at five US sportsbooks. The singles roughly broke even. The parlays lost, and the longer the ticket, the worse it got.

The Pitch

A sales video we were sent in October showed a record of 43 wins and 5 losses in a month at an average price of -336, built from parlays of heavy favorites. The idea sounds safe. A -300 favorite wins about three quarters of the time, so string two or three together and you cash most nights while the payout on each ticket stays near even money. Forty-eight bets is a month, and a month is not evidence of anything, so we did what we do with every pitch: wrote down the test first, then ran it on every season we hold.

What We Tested

Six seasons of MLB and NHL, 2021 through 2026, at the five regulated US sportsbooks we track. A side qualified when its best price across the books sat between -200 and -699, which is the band our own Heavy Favorites lane uses. Each day we built one ticket per sport from the shortest prices first, the way a favorites bettor would, and we staked one unit at the product of the legs' best prices. That is a generous assumption, because one sportsbook rarely holds the best price on every leg, so a real ticket pays a little less than our numbers. If you want the arithmetic behind how a parlay works, we have a plain guide to it. The results are below exactly as they came out.

The Results

Singles first, because they are the baseline. Heavy MLB favorites went 69.6% over 1,836 bets and lost 0.95% of the money staked. Heavy NHL favorites went 71.0% over 1,762 and lost 0.92%. That is what a fairly priced favorite looks like: it wins a lot and returns the stake minus the fee. Two-leg parlays won 48.3% of the time in MLB and 49.7% in the NHL, and lost 4.3% and 6.6%. Three legs won 27.0% in MLB and 35.6% in the NHL, and lost 22.7% and 6.3%. Building the tickets from the longest-priced qualifiers instead of the shortest made MLB worse, at 8.9% and 18.5% lost. Five of the six parlay cells lost between 4% and 23%, and the one cell with a positive stretch, three-leg NHL tickets from 2024 on, sits beside a losing full sample and a losing two-leg cell, which is what noise looks like.

Why the Math Goes Wrong

Two things happen at once. First, the win rate is not the favorite's win rate, it is the product of the legs: two 70% favorites win together half the time, three win together a third of the time. The one property that made the favorites attractive is gone after the second leg. Second, the sportsbook fee inside each price multiplies through the ticket instead of being paid once. A single -300 favorite carries a small fee; three of them carry that fee three times over, compounded, on a bet that now wins a third of the time. The pitch also leans on prices like -600 that barely exist in these sports: across six seasons we counted 13 MLB games and 119 NHL games where a side reached that price at any book. Most heavy favorites are -200 to -350, which pays about even money on two legs and nowhere near the safety the video implies.

What We Do Instead

We post heavy favorites, as singles, in a lane with its own record: one unit, MLB and NHL moneylines in the same -200 to -699 band, at the best of the five books, never parlayed. We say plainly on that record that it is a high win rate with a return near zero, because that is what the six-season test showed and what the lane has done since. A high win rate is a real thing to want, especially for a bettor who hates losing streaks, and we would rather sell it honestly than wrap it in a parlay that turns a fair price into a losing one.

How to Read a Favorites Pitch

Ask for the sample, in bets and in months, and ask what the same picks did as singles. If the singles are near break-even, the parlay cannot be better than them, only more volatile. Ask whether the tickets were priced at one book or at the best of several, because the best-of-five price is a ceiling no real bettor gets every time. And notice whether the service ever passes. A page that always has a ticket is selling tickets. We publish a No Bet when nothing clears, and we published this test knowing the answer was not the one the pitch wanted.