Live Dealer Side Bets: What 21+3, Perfect Pairs and Insurance Actually Cost
Published · Updated: · Gambling Websites Editorial Team
Side bets are sold as an accessory to blackjack. They are not. Each one is a separate wager with its own paytable and its own house edge, settled independently of how your hand plays out, and the edge on most of them is several times the edge on the game they sit next to.
Published estimates put the main blackjack game under 1% with reasonable rules and play. Across the side-bet guides the operators and gaming publications maintain, most blackjack side bets fall in the 5–15% range, and individual figures run from about 2–3% at the friendliest paytables to above 11% at the worst.
That is the comparison worth holding: a sub-1% game with a 5-to-15% bet attached to it, on the same felt, at the same moment.
Insurance: the familiar one, and not a hedge
Insurance is offered when the dealer shows an Ace. You stake half your original bet, and it pays 2:1 if the dealer has blackjack.
The structure looks like protection, which is why it is taken so often. Arithmetically it is a standalone bet on one proposition — that the dealer’s hole card is a ten-value card — and published estimates put its house edge at around 7.4–7.5%. It does not reduce the variance of your main hand; it adds a second, worse bet alongside it.
The phrasing does the damage here. “Insurance” describes what the bet feels like, not what it is.
21+3 and Perfect Pairs: the numbers disagree for a reason
21+3 combines your two cards with the dealer’s up card and pays if the three make a poker hand. Top payouts commonly reach 100:1 for a suited three of a kind.
Perfect Pairs pays if your first two cards are a pair, at escalating rates: a mixed pair lowest, a coloured pair higher, a suited “perfect” pair highest. Published top payouts range from 25:1 to 30:1 for the perfect pair, with coloured pairs commonly around 12:1.
Now the awkward part. Published house-edge figures for these two bets do not agree with each other, and the disagreement is not sloppiness.
One widely cited series gives 21+3 by deck count: about 8.78% with four decks, 7.81% with five, 7.14% with six and 6.29% with seven. Another guide puts the typical 21+3 edge at 3–4% on standard paytables, with worse variants exceeding 13%. A third gives a range of roughly 3.24% to 7.14%.
Perfect Pairs is spread just as wide: one guide gives a typical 5–7% depending on paytable and deck count, another reports figures as far apart as 4.14% and 10.7%, and a third puts the overall range at 2% to 11%.
Those are not competing measurements of one bet. They are measurements of different bets carrying the same name. The house edge on a side bet is a product of its paytable and the number of decks in the shoe, and both vary from table to table. Change the payout for a suited three of a kind from 100:1 to 40:1 and the edge moves by several points without a single word on the table changing.
Which is why a remembered number is worthless at a live table. “21+3 is about 3%” is true somewhere and badly wrong elsewhere.
The one rule the deck count gives you
Within a single paytable, the direction of travel is consistent for 21+3: more decks, lower edge. The cited series moves from 8.78% at four decks down to 6.29% at seven, because a larger shoe makes the flushes the bet pays for slightly more likely.
That is the opposite of the instinct carried over from the main game, where fewer decks favour the player. Deck count works one way for blackjack and the other way for 21+3, and a table chosen on main-game rules is not chosen for its side bets.
Perfect Pairs runs the other direction — pairs are harder to make in a bigger shoe — which is why the published numbers for it move around so unpredictably once paytables differ as well.
What it costs over a session
The edges only become concrete at the volume a live table actually deals.
Take 100 hands, a $10 main bet and a flat 0.5% edge on well-played basic strategy: expected cost around $5 across the session.
Add a $5 side bet on every hand at a 6% edge — a mid-range figure for either 21+3 or Perfect Pairs — and that is $500 of extra turnover at 6%, around $30 expected. The side bet is half the stake size and six times the cost.
Take the same $5 side bet at an 11% paytable and it is about $55, on top of a $5 main game. The accessory is now ten times the cost of the game.
Nothing in that arithmetic is a prediction of one session; the variance on side bets is high, which is exactly what makes them appealing. It is the long-run price of leaving the side-bet circle covered, and the reason a $5 optional wager is not a small decision.
How to read a side bet before you take it
Three checks, all available at the table before you place a chip.
Open the paytable. Live dealer interfaces publish it. The top payout for the rarest hand is the single most informative number, because that is where operators cut to change the edge.
Note the deck count. It is in the game rules, and for 21+3 it moves the edge by two or three points across the common range.
Price the turnover, not the bet. A $5 side bet on 100 hands is $500 of exposure at side-bet rates. Compare that to what the same $500 would cost you in the base game and the decision makes itself.
None of this makes side bets a mistake to take. They are a genuinely different product — short, high-variance, and priced accordingly — and they are worth the money to some players for exactly that reason. What they are not is an extension of blackjack, and the published numbers, contradictory as they are, all point the same way about the price.