NBA PrizePicks Math: Break-Even Rates and What to Avoid

On PrizePicks standard projections, each pick needs roughly a 54 to 62 percent hit rate for an entry to break even, depending on entry type. Two-pick entries carry the steepest requirement: 57.7 percent for Power and 61.8 percent for Flex. Avoid entries built from near coin-flip picks, Goblin-heavy lineups, and legs that work against each other.

What pick'em contests actually pay

PrizePicks is a pick'em daily fantasy format: an entry picks more or less on two to six player projections, and the payout is a fixed multiplier of the entry fee set by how many picks hit. That makes it a math problem with a known payout table, which is easier to analyse than most betting products.

PrizePicks' "How to Pick" page, checked on 2026-11-05, lists these multipliers for lineups on standard projections. Power Play, where every pick must hit: 2 picks 3x, 3 picks 6x, 4 picks 10x, 5 picks 20x, 6 picks 37.5x. Flex Play, with partial payouts: 2 picks pay 2x for 2 of 2 and 0.5x for 1 of 2; 3 picks pay 3x and 1x for 2 of 3; 4 picks pay 6x and 1.5x for 3 of 4; 5 picks pay 10x, 2x for 4 of 5 and 0.4x for 3 of 5; 6 picks pay 25x, 2x for 5 of 6 and 0.4x for 4 of 6. The page notes that multipliers are subject to change, that Goblin projections lower multipliers, that Demon projections raise them, and that team picks are priced from live markets. Always confirm the payout on the entry screen.

The break-even hit rate for each entry

Assume every pick in an entry hits with the same probability p, independently of the others. The break-even p is the value at which expected payout equals the entry fee. For Power Play it is the n-th root of 1 divided by the multiplier.

Power Play break-even per pick: 2 picks 57.74 percent, 3 picks 55.03 percent, 4 picks 56.23 percent, 5 picks 54.93 percent, 6 picks 54.66 percent.

Flex Play break-even per pick, solved numerically from the binomial distribution: 2 picks 61.80 percent, 3 picks 57.74 percent, 4 picks 55.03 percent, 5 picks 54.25 percent, 6 picks 54.21 percent.

For scale, a standard -110 sportsbook bet breaks even at 52.38 percent. A 57.74 percent requirement is the same as needing every pick to be worth about -137 in American odds. Even the most favourable entries need picks worth roughly -118 to -122.

Why coin-flip picks are the first thing to avoid

Projections are set near the median outcome, so a pick made without real information hits about half the time. At p of 0.50 every entry type loses heavily: 25 percent of the fee for 2-pick and 3-pick Power, 37.5 percent for 4-pick and 5-pick Power, 41.4 percent for 6-pick Power, and 25 to 32.8 percent for Flex entries.

At a more ambitious 55 percent per pick, the picture splits. A 2-pick Power Play still loses 9.2 percent and a 3-pick Power Play is roughly break-even, while a 6-pick Power Play shows plus 3.8 percent and a 6-pick Flex plus 7.5 percent. Those figures are illustrative and depend entirely on the 55 percent being real.

The entries with the steepest hurdle

The 2-pick entries are the hardest to beat on standard projections. Power needs 57.74 percent per pick and Flex needs 61.80 percent, the highest requirement in the table. The 4-pick Power Play is also worse than both its neighbours, at 56.23 percent against 55.03 percent for 3 picks and 54.93 percent for 5.

Goblin projections deserve the same caution. They are easier to hit but lower the multiplier, and the break-even rises with them. An easier pick is only an improvement if its hit probability rises by more than the payout falls, which has to be measured, not assumed.

Longer entries trade a lower hurdle for variance

The 6-pick entries have the lowest break-even rates, but they rarely pay in full. At 55 percent per pick, all six hit only 2.8 percent of the time, roughly once in 36 entries. A bankroll that cannot absorb long losing runs will not survive long enough for a small edge to show. This is the same trade off as in sportsbook parlays: more legs can mean a lower per-pick hurdle, but far more variance.

Correlation changes the maths

The break-even figures assume independent picks. Real picks are often correlated. Two picks from the same NBA game can move together: a fast, high scoring game lifts several players' points, and an injury to one player shifts usage to teammates.

Take two picks each at 55 percent. Independent, both hit 30.25 percent of the time and a 2-pick Power Play loses 9.2 percent. With a correlation of 0.2 between them, both hit 35.2 percent of the time and the same entry shows plus 5.6 percent. With negative correlation, such as overs on two players competing for the same shots, the joint probability falls and the entry gets worse. Correlated combinations are well known, so it is reasonable to expect the operator to account for the obvious ones in its projections and rules. Measure correlation from historical box scores rather than guessing.

NBA-specific traps

NBA player outcomes depend heavily on minutes. Blowouts cut starters' minutes in the fourth quarter. Rest days and late scratches change rotations close to tip-off. A projection set before injury news may be stale, and a pick that looked like 58 percent at noon may be a coin flip by evening. Check how the operator handles a player who does not play, because it changes the effective entry.

Estimating a pick's probability

The usual starting point is the sportsbook market for the same player and line. Remove the margin from the over and under prices to get a fair probability, then compare it with the break-even rate for the entry type. If the sportsbook line differs from the projection, adjust using the player's historical distribution. Track every pick's estimated probability and result, and check calibration over hundreds of picks before trusting the estimates. This page describes contest mathematics, not picks, and eligibility for these contests varies by state.

Frequently asked questions

What hit rate do you need to break even on PrizePicks?
On standard projections as listed on 2026-11-05, about 54.2 to 61.8 percent per pick, assuming independent picks. The 6-pick Flex needs about 54.2 percent and the 6-pick Power about 54.7 percent. The 2-pick Power needs 57.7 percent and the 2-pick Flex 61.8 percent. Multipliers change, so recompute from the payout shown on the entry screen.
Is Power Play or Flex Play better?
It depends on entry size. At 2 picks Power has the lower break-even, 57.7 against 61.8 percent. At 3 picks Power is lower, 55.0 against 57.7; at 4 picks Flex is lower. At 5 and 6 picks Flex needs about 54.2 percent against 54.7 to 54.9 percent for Power, and pays partial prizes, so it has less variance. Both lose heavily if picks are near 50 percent.
Do correlated picks help on PrizePicks?
Positively correlated picks raise the probability that all of them hit, which helps entries that need every pick. In an illustrative 2-pick entry at 55 percent per pick, a correlation of 0.2 turns an expected loss of 9.2 percent into a gain of 5.6 percent. Negatively correlated picks do the reverse. Measure correlation from historical data rather than intuition.
Are Goblin picks worth it?
Only if the increase in hit probability outweighs the lower multiplier. PrizePicks states that Goblins are easier to win but decrease payout multipliers, so the break-even hit rate per pick rises. Compute the new break-even from the payout shown on the entry, estimate the Goblin line's true probability from player data, and compare the two.