What should you use a profit boost on?

Every app hands you the boost and then goes quiet on the only question that matters: what do you put it on? Most people default to one of two answers. The heavy favorite, because it usually wins. Or the moonshot parlay, because the payout looks incredible. Both defaults are wrong, and the reason both are wrong is one short formula.

If you’re still on the question of whether boosts are worth using at all, start with are profit boosts actually +EV?. This article assumes the answer is “sometimes” and asks what deserves them.

One formula decides this

A profit boost multiplies the profit on a winning bet. Take a bet at decimal odds d and boost the profit by b, and the boosted decimal becomes 1 + (d - 1)(1 + b). If the bet is fairly priced, your win probability is p = 1/d, and the expected value collapses to something almost rude in its simplicity:

EV = boost % x (1 - win probability)

The boost times the probability you lose. That’s the whole thing. The full step-by-step, with the algebra written out, is in how to calculate a profit boost’s EV.

Read the formula again, because it’s telling you something most bettors have backwards. A boost is worth more the more often the bet loses. Favorites win a lot, so there’s barely any profit for the boost to multiply. Underdogs lose a lot, and the boost pays you for exactly that.

The decision table

Here’s a 50% boost at a $25 capped stake, across the odds ladder, with every price assumed fair:

Same boost, same stake, and the expected profit more than doubles from top to bottom. The table embarrasses the play-it-safe crowd immediately: burning a boost on -200 because “it’ll probably win” takes $4.17 of expected value when $10.00 was on the same menu. The moonshot crowd has to wait one section for their turn.

Worked example, the +150 row. Decimal odds 2.50, so a 50% boost makes it 1 + 1.50 x 1.5 = 3.25, which is +225. If +150 is the fair price, the bet wins 40% of the time: EV = 0.40 x 3.25 - 1 = +30%. On a $25 capped stake, that’s $7.50 of expected profit for tapping a different button than the favorite. Run any candidate you’re actually holding through the boost EV calculator and it does this in one line.

Three things fight the table

If the table were the whole story, the answer would be “always the longest eligible odds.” It isn’t, for three reasons, and they matter in order.

Longshot prices are the least trustworthy prices in the market. Books shade longshots hardest, because casuals love them and rarely check. That +400 on the board might be a +550 shot wearing a +400 price. Every row in the table assumed the price was fair, and that assumption fails worst exactly where the formula looks best. Your defense is to stop trusting the board and price the bet yourself: take a real two-sided market, strip the vig with the no-vig calculator, and see what the bet is actually worth before any boost touches it. A +40% EV computed from a fake fair price isn’t +40%. It’s a guess in a lab coat.

Variance climbs the whole way down. A fair +400 loses 80% of the time even when your math is perfect. Over a season of daily boosts that’s fine, the edge shows up in the sum. Over one week it can look like you’ve learned nothing. The math doesn’t owe you a smooth ride, only a direction.

The terms bind. Stake caps, minimum odds, eligible markets, sometimes a required number of legs. The best bet on the board is irrelevant if it isn’t eligible, and a min-odds rule can shove you toward exactly the longshot zone where point one lives. When a boost demands extra legs, the desk meets the requirement with as little added variance as the terms allow, because every unnecessary leg is vig you volunteered for.

The EV-first answer

So the answer to “what should you use a profit boost on” is not an odds range. It’s a procedure: the eligible bet with the highest verified post-boost EV.

Verified is the load-bearing word. It means the fair price came from devigging a real two-sided market, not from a gut read, and definitely not from the book’s own boost copy. Then the boost gets applied to that verified number, the terms get checked, and the biggest surviving edge wins.

In practice, that procedure lands on moderate underdogs more often than on either end of the table. Long enough that the boost has real profit to multiply, short enough that the market still prices them honestly. Not because moderate dogs are magic. Because that’s where the two things the formula needs, a big chance of losing and a fair price you can trust, overlap most often.

The check takes about a minute per boost once it’s a habit. Or skip the minute: the free Discord posts one boosted play a day when the math clears the bar, fair odds and edge shown, join free. Every result lands on the public ledger, the $4.17 days and the losing weeks both.