how promos are rated

every score comes from expected value, basically how much a promo pays you back on average for each dollar you run through it. here is the actual math i use, nothing hidden. back to promotions

the core idea

a promo is worth whatever it pays you minus what it costs to unlock. i turn that into a percentage of the money you put in.

EV% = ( what it pays back − what it costs to unlock ) ÷ money you put in

the cost to unlock is just the book's margin (the vig) on any betting it makes you do, usually somewhere around 3 to 5% of what you have to wager. everything below is that same idea for one type of promo. when i need a real probability i pull it straight from the price with the vig stripped out, p = 1 / fair odds.

the formulas, by type

boosted odds

EV% = p × O_boost − 1    (p = 1 / fair odds)

a boost is good whenever the boosted price beats the fair one. say fair is 2.00 (p = 0.50) and they bump it to 2.40, that is 0.50 × 2.40 − 1 = +20% on the bet. free money if you were taking that side anyway.

bet and get / free bets

EV = (F × c) − (T × v)

F is the free bet amount, c is what a free bet is actually worth in cash (about 0.7, since you don't get the stake back), T is how much you have to bet, v is the vig on those bets. these are the best ones going. a $500-in for $250-free deal is roughly 250 × 0.7 − 500 × 0.04, so about +$155 on $500, call it +30%.

early payout (2 up, 90th minute)

EV = P(you get the lead but the result slips) × payout

the book pays you in the extra spots where you would normally lose, like your team goes 2 up then draws. no downside, and it just happens on every qualifying bet, so i rate these higher than the small per-bet number makes them look.

insurance and refunds

EV = P(trigger) × R

R is whatever gets refunded (bore draw, split decision, final-set tiebreak, walk-off, late loss). you can't lose anything extra, the trigger just decides how often you collect. more common trigger, more EV.

cashback

EV = P(you finish down) × rate × average loss

some of your losses come back. no timing, no mechanic to catch, just reliable.

wager races and leaderboards

EV = Σ P(you finish rank i) × prize_i − extra vig from the volume

for a normal bettor your slice of the pool is tiny, so i weigh the pool against the field and how many spots pay. nice if you were already betting, rarely worth grinding volume for.

paid insurance and features

stuff like stake shield is priced about fairly, you give up payout for protection so the EV lands near zero. pure features like streams, extra markets, xp and same-game-multi builders don't pay anything, so they sit around 4.

turning EV into a score

i map the EV to a number, then bump it a little for how easy it is to actually get.

+20% and up9 – 10basically free money
+8% to +20%8strong edge on the bet
+4% to +8%7solid, dependable value
+2% to +4%6worth taking
+0.5% to +2%5pretty marginal
around 0%4a feature, not value

bumps up if it applies to every bet with no cap and no downside. bumps down if it is capped, one-off, needs an opt-in, or you have to bet a ton to see any of it.

these are estimates, not promises. i am approximating the true odds from the market and your actual results depend on what you bet and how. not financial advice. promos change and expire so always read the book's terms.