Blackjack Side Bets: Perfect Pairs & 21+3, Computed Exactly
Side bets cost 5-10 times more than the main game: Perfect Pairs runs about a 6.11% house edge and 21+3 about 3.24-4.62%, versus ~0.6% for basic strategy. Treat them as fun money only — they are never part of a winning strategy.
Every side-bet circle on the felt is a tiny, brightly painted toll booth. The numbers below are computed combinatorially for a 6-deck shoe — every possible two- and three-card combination counted exactly, no simulation, no rounding to whatever sounds good. That matters because competitors publish contradictory side-bet odds without saying what game they assume; ours state their assumptions: 6 decks, the paytables named in each table.
Perfect Pairs: what it is and what it pays
Perfect Pairs wagers that your first two cards are a pair. Three tiers, paid by how closely they match. Out of every possible 6-deck starting hand:
| Outcome | Definition | Probability | Pays (25/12/6) | Pays (30/10/5) |
|---|---|---|---|---|
| Perfect pair | Identical card (same rank & suit) | 1.61% | 25:1 | 30:1 |
| Colored pair | Same rank, same color, different suit | 1.93% | 12:1 | 10:1 |
| Mixed pair | Same rank, different color | 3.86% | 6:1 | 5:1 |
| Any pair | All of the above | 7.40% | edge 6.11% | edge 5.79% |
A pair lands about once every 13.5 hands, and 92.60% of the time the bet simply vanishes. At the common 25/12/6 paytable the house keeps 6.11% of every dollar wagered — you lose roughly 6¢ per $1, ten times the ~0.6% of the main game. The 30/10/5 version is fractionally kinder at 5.79%, which is still terrible.
21+3: what it is and what it pays
21+3 combines your two cards plus the dealer's up-card into a three-card poker hand. In a 6-deck game:
| Hand | Probability | Classic flat | Tiered (5/10/30/40/100) |
|---|---|---|---|
| Flush | 5.84% | 9:1 | 5:1 |
| Straight | 3.10% | 9:1 | 10:1 |
| Three of a kind | 0.50% | 9:1 | 30:1 |
| Straight flush | 0.21% | 9:1 | 40:1 |
| Suited three of a kind | 0.02% | 9:1 | 100:1 |
| Any qualifying hand | 9.68% | edge 3.24% | edge 4.62% |
Note the counter-intuitive bottom line: the flashy tiered paytable with the 100:1 jackpot costs more (4.62%) than the boring flat 9:1 version (3.24%). The big numbers sit on hands you'll almost never see — suited three of a kind arrives once in roughly 5,000 hands — while the flush, which does the real work at 5.84%, gets cut from 9:1 to 5:1 to pay for the fireworks.
Why the math can't be beaten
The main game's edge is low because you make decisions — every hit, stand, 3:2 blackjack payout, and double is a lever. Side bets settle on the deal. No decision exists, so no skill exists, and the edge baked into the paytable is the edge you pay, forever. It's the same trap as insurance — which is itself just a side bet on the dealer's hole card carrying a 7.69% edge. You can verify any of these against your own table's rules with the house edge calculator.
Why they feel better than they are
21+3 hits about 1 in 10 hands — frequent enough that a session reliably includes a few wins, each paid in a satisfying stack of chips. That's the design: intermittent rewards on a fast cycle are the most habit-forming schedule there is. Your memory keeps the 9:1 flush; the ledger keeps the nine losing dollars around it. Perfect Pairs is harsher — one hit per 13.5 hands — but a 25:1 payout buys a lot of selective memory.
If you must
- Budget it as entertainment, like a movie ticket — money you expect to spend, not stake. A $5 21+3 bet every hand costs about $16-23 per 100 hands in expectation; decide whether that's worth the sweat to you.
- Avoid the tiered-paytable trap. If you can choose, the dull flat 9:1 21+3 at 3.24% is the cheapest seat in the side-bet house. Big headline numbers on the felt almost always mean a worse bet underneath.
- Never let it touch your main-hand play. The side bet is settled; the hand in front of you still deserves perfect basic strategy.
Frequently asked questions
Is 21+3 ever a positive-EV bet?
Only with specialized counting systems that track suit imbalances — because most of 21+3's value sits in flushes, a shoe rich in one suit can briefly tip the bet positive. In practice the systems are impractical: the edge windows are rare and thin, and casinos cap side-bet stakes precisely because of them. For any human at a real table, 21+3 loses.
Do side bets affect the main hand?
No. They're resolved off the initial cards before you make any decision, and they change nothing about correct basic strategy. Win or lose the side bet, you play 16 vs 10 exactly the same way.
Which side bet is least bad?
Of the common ones, 21+3 with the classic flat 9:1 paytable at about 3.24% in a 6-deck game. That's still five times worse than the ~0.6% main game — "least bad" is doing heavy lifting.