Last updated: 7 October 2026 · By Power.Win Editorial Team
Quick Answer: Blackjack side bets cost far more than the main hand: 21+3 pays 5:1 to 100:1 (3.70% edge), Perfect Pairs 6:1 to 25:1 (4.10%). Wizard of Odds lists these eight-deck tables for live-dealer studios, and the edges are about nine times the 0.43% main-hand edge; other tables differ. Power.Win’s Blackjack Original documentation lists insurance, not 21+3 or Perfect Pairs.
A small circle next to your bet offers 21+3 or Perfect Pairs for a dollar, and the sign above it shows 25 to 1 or 100 to 1. Nobody tells you what the circle costs.
Blackjack side bets are separate wagers placed beside your main hand, each with its own pay table and its own house edge. One bet, 21+3, costs 3.70% on the 100-to-1 pay table and 13.39% on the “Xtreme” one, depending on the pay line printed at the table.
The figures below come from Wizard of Odds. Each side-bet figure is tied to the pay table and deck count it was calculated on. The 0.43% main-hand figure is Wizard’s number for Atlantic City rules, and the page it comes from gives no deck count.
| Bet | Pay table priced | Decks | House edge | Cost per $100 staked |
|---|---|---|---|---|
| Main hand, basic strategy | Atlantic City rules, dealer stands on soft 17 | not stated | 0.43% | $0.43 |
| 21+3 | 5 / 10 / 30 / 40 / 100 to 1 | 8 | 3.70% | $3.70 |
| Perfect Pairs | 6 / 12 / 25 to 1 | 8 | 4.10% | $4.10 |
| Insurance | 2 to 1 | 6 | 7.4% | $7.40 |
House edge is the share of each wager the casino keeps on average over a very long run, and “cost per $100” is the same number on $100 staked. It’s an average, not a prediction for any one session.
Are blackjack side bets worth playing?
Not for value. On the pay tables priced here, side bets cost between 2.17% and 13.39% per dollar against 0.43% on the main hand, so each side-bet dollar costs about 5 to 31 times as much.
Big payout lines are the hook: 25 to 1 and 100 to 1 look generous, but they pay on rare hands. A Perfect Pairs 25 to 1 lands about once in 59 rounds.
On the 21+3 pay table priced here, any winning poker hand lands about 9.7% of the time, roughly once in ten rounds, and the 100 to 1 line lands about once in 4,100 rounds. The frequent, small wins keep the bet feeling fair while the average cost stays well above the main hand’s.
Worked example: 100 hands, $5 main bet plus $1 on 21+3
Say you play 100 hands at $5, ignoring doubles and splits, and add $1 on 21+3 to each one.
- Main hand: 100 × $5 = $500 staked × 0.43% = $2.15 expected cost.
- 21+3 side bet: 100 × $1 = $100 staked × 3.70% = $3.70 expected cost.
- Together: $600 staked, $5.85 expected cost, which is 0.975% of everything you put down.
The side bet is one-sixth of the money staked and about 63% of the expected cost ($3.70 ÷ $5.85). Here’s how the same $1-a-hand habit looks on other pay tables:
| $1 side bet on every hand | Side-bet stake | Side-bet cost | Main-hand cost | Total expected cost | Side bet’s share of cost |
|---|---|---|---|---|---|
| 21+3, 5 / 10 / 30 / 40 / 100 to 1, 8 decks | $100 | $3.70 | $2.15 | $5.85 | 63% |
| Perfect Pairs, 6 / 12 / 25 to 1, 8 decks | $100 | $4.10 | $2.15 | $6.25 | 66% |
| 21+3 “Xtreme”, 5 / 10 / 20 / 30 to 1, 6 decks | $100 | $13.39 | $2.15 | $15.54 | 86% |
Those are averages, and a hundred hands is a short run. Using the Wizard of Odds combination counts for 21+3 at eight decks, the standard deviation of 100 one-dollar bets is about $39.83, roughly eleven times the $3.70 expected cost.
About 40% of 100-hand runs finish ahead on the side bet alone, about 1% finish exactly level and about 59% finish behind. A run of good luck doesn’t change the price, it only hides it.
What are the odds and payouts on 21+3?
21+3 pays on a poker hand made from your first two cards and the dealer’s upcard, from 5 to 1 for a flush to 100 to 1 for suited three of a kind, at a 3.70% house edge with eight decks.
This is the pay table and return table Wizard of Odds lists for eight decks as Version 7. The page says it hears live-dealer studios use this table, so treat that attribution as secondhand:
| Hand | Pays | Probability | Return per $1 |
|---|---|---|---|
| Suited three of a kind | 100 to 1 | 0.0244% | $0.0244 |
| Straight flush | 40 to 1 | 0.2063% | $0.0825 |
| Three of a kind | 30 to 1 | 0.5169% | $0.1551 |
| Straight | 10 to 1 | 3.0947% | $0.3095 |
| Flush | 5 to 1 | 5.8841% | $0.2942 |
| No qualifying hand | Lose $1 | 90.2736% | −$0.9027 |
Add the five winning returns ($0.8657) to the loss ($0.9027) and you get −$0.0370 per dollar, the 3.70% edge. About 70% of the return comes from the flush and the straight, the two hands that pay least.
The pay table is the whole story. The same bet with a 5 / 10 / 20 / 30 to 1 table, known as 21+3 “Xtreme”, carries a 13.39% edge at six decks. Deck count matters too: on that Xtreme table the edge is 22.73% with one deck and 12.89% with eight.
Wizard of Odds lists ten pay-table versions, so a 21+3 circle with an unfamiliar pay line can’t be priced by name alone.
What are the odds and payouts on Perfect Pairs?
Perfect Pairs pays when your first two cards are a pair: 25 to 1 for the same rank and suit, 12 to 1 for the same colour, 6 to 1 for mixed colours, a 4.10% house edge at eight decks.
Wizard of Odds prices four Perfect Pairs pay tables, all at eight decks, and ties table A to live-dealer studios:
| Pay table | Perfect pair | Same colour | Different colour | House edge |
|---|---|---|---|---|
| A | 25 to 1 | 12 to 1 | 6 to 1 | 4.10% |
| B | 30 to 1 | 10 to 1 | 5 to 1 | 3.37% |
| C | 25 to 1 | 12 to 1 | 5 to 1 | 7.95% |
| D | 25 to 1 | 15 to 1 | 5 to 1 | 2.17% |
Look at tables A and C. They differ by one point on a single line, 6 to 1 against 5 to 1 for a pair in different colours, and the edge nearly doubles from 4.10% to 7.95%. Mixed-colour pairs land 3.86% of the time at eight decks, so one point on that line is worth 3.86 points of edge.
The hit rates, from the same page, are 1.69% for a perfect pair (about 1 in 59), 1.93% for a same-colour pair and 3.86% for a different-colour pair. Any pair, which is what a main-hand split would need, lands 7.47% of the time.
A different bet shares the name. The singular Perfect Pair, found on some online live tables, looks at the player’s and the dealer’s first two cards and pays 25 to 1 for one perfect pair and 200 to 1 for two.
Wizard of Odds puts it at an 8.05% edge at eight decks, and 13.03% if both outcomes pay 25 to 1.
How much does blackjack insurance cost, and is even money the same bet?
Insurance is a side bet of up to half your wager that pays 2 to 1 if the dealer’s hole card is a ten. At six decks it costs 7.4% per dollar, and taking even money on a natural is the same bet.
Insurance isn’t protection for your hand. It’s a separate wager on one question, whether the dealer’s hole card is a ten-value card. Paying 2 to 1 means the bet needs a one-in-three chance of a ten to break even. A six-deck shoe has fewer, and Wizard of Odds works the loss out as 7.4 cents per dollar insured.
The maths is short. After a dealer ace is showing, 96 of the 311 unseen cards in a six-deck shoe are tens, or 30.87%. Three times that, minus one, is −0.074, the 7.4% edge. On a $10 hand, the largest insurance bet is $5, which costs about $0.37 on average.
Even money is the offer when you hold a natural against a dealer ace: a fixed 1 to 1 payout, settled at once, instead of waiting for the dealer’s hole card. It is the same bet as insurance, because insurance on a natural also nets one unit either way.
The cost shows up differently, though. On the same Wizard of Odds page, declining even money returns about 103.88% of your main bet on average at six decks, against a flat 100% for taking it. That’s about 3.88% of the main wager given up, a different base from the 7.4% above, which is per dollar of insurance.
Read More: How to play blackjack online: rules and when insurance is offered
Which blackjack side bets cost the most?
Ranked by house edge on the pay tables priced here, the side bets run from 2.17% for Perfect Pairs table D to 13.39% for 21+3 Xtreme, and every one costs at least five times the main hand’s 0.43%.
| Side bet | Pay table priced | Decks | House edge | Cost per $100 | Versus main hand |
|---|---|---|---|---|---|
| Main hand (reference) | Atlantic City rules, stands on soft 17 | not stated | 0.43% | $0.43 | 1× |
| Perfect Pairs, table D | 25 / 15 / 5 to 1 | 8 | 2.17% | $2.17 | 5× |
| Perfect Pairs, table B | 30 / 10 / 5 to 1 | 8 | 3.37% | $3.37 | 8× |
| 21+3 | 5 / 10 / 30 / 40 / 100 to 1 | 8 | 3.70% | $3.70 | 9× |
| Perfect Pairs, table A | 25 / 12 / 6 to 1 | 8 | 4.10% | $4.10 | 10× |
| Insurance | 2 to 1 | 6 | 7.4% | $7.40 | 17× |
| Perfect Pairs, table C | 25 / 12 / 5 to 1 | 8 | 7.95% | $7.95 | 18× |
| Perfect Pair (singular) | 200 / 25 to 1 | 8 | 8.05% | $8.05 | 19× |
| 21+3 “Xtreme” | 30 / 20 / 10 / 5 to 1 | 6 | 13.39% | $13.39 | 31× |
The “versus main hand” column divides each edge by 0.43% and rounds to a whole number. Wizard of Odds gives the 0.43% baseline for Atlantic City rules, where the dealer stands on soft 17 and doubling after a split is allowed. The page states it without a deck count.
Two things follow. A pay table decides the ranking more than the name of the bet does, since Perfect Pairs is both the cheapest side bet in the table and one of the dearest.
And no row gets as cheap as the main hand, which is why the choice is never between a “good” and a “bad” side bet, only between cheaper and dearer ones.
Why do casinos offer blackjack side bets?
Side bets raise the table’s hold: a dollar on 21+3 gives the house about 3.7 cents on average where a dollar on the main hand gives 0.43 cents, so a full set of side-bet circles brings in several times more per round.
That’s the honest answer, and it’s worth knowing before you place one. The operator chooses the pay table, so it chooses the edge: the same 21+3 circle can cost 3.70% or 13.39% depending on which pay line was set.
A top payout of 100 to 1 or 200 to 1 also draws attention, because a rare big win is easier to picture than a steady 3.7-cent cost.
None of that makes a side bet dishonest. A bet with a published pay table and a calculable edge is a fair product at a price. The problem is only when the price isn’t read.
When is a blackjack side bet defensible?
Only as paid entertainment, with a stake you’ve capped in advance and counted as a cost rather than a return. At 3.70%, a $1 21+3 bet costs about 3.7 cents a round on average, while its results swing far more than that.
That isn’t a recommendation. It’s the condition under which the maths stops being a surprise. Three checks keep the cost visible:
- Read the pay table first. If you can’t read every pay line before you bet, you can’t price the bet, so skip it.
- Set the side-bet stake before the session. A $1 side bet over 100 hands is $100 staked, and the average cost is $3.70, but the swing is about eleven times that.
- Keep the main hand right. Basic strategy lowers the cost of your decisions on the 0.43% bet, and our blackjack basic strategy guide covers it. A side bet can’t make up for a misplayed hand, because it’s a separate wager with its own edge.
Side bets don’t change the main hand’s edge, only your total cost.
Which blackjack side bets does Power.Win list?
Power.Win’s documentation for its PWR Originals Blackjack lists insurance among the actions, with 3:2 naturals and a dealer who stands on 17, and names no 21+3 or Perfect Pairs. Live tables set their own side bets.
The Blackjack documentation lists hit, stand, double, split and insurance, lets you play one, two or three hands per round, and states that a natural pays 3:2 and the dealer stands on 17. It describes a “standard 52-card deck” and gives no shoe size.
The docs FAQ adds that the classic blackjack dealer stands on soft 17, which matches the 0.43% baseline above. Power.Win doesn’t publish a main-hand edge, so treat 0.43% as a reference for those rules, not as Power.Win’s own figure.
The Blackjack Original is a provably fair PWR Original: you see a hashed server seed before the hand, set your own client seed, and can check the deal afterwards.
Live dealer blackjack in the Power.Win blackjack lobby comes from certified third party providers. Each live table sets its own rules, and this page can’t confirm which of them carry 21+3 or Perfect Pairs, so the pay table on that table’s own rules panel is the figure to read before any side bet.
The operator behind Power.Win is 3-102-968317 SRL of Costa Rica, which holds licence 0000209 from the Tobique Gaming Commission.
Power.Win isn’t available to players in the United States: section 5.2 of its Terms and Conditions lists the United States of America among the countries whose players aren’t eligible to deposit or play real-money games. Read this page as an explanation of how this works, not as an invitation to sign up.
Checked against Power.Win’s documentation and live pages on 7 October 2026.
Frequently Asked Questions
When does insurance break even?
Insurance breaks even when one in three unseen cards is a ten, because it pays 2 to 1 and loses 1. After a dealer ace in a six-deck shoe, 96 of 311 unseen cards are tens (30.87%), so the bet loses 7.4% per dollar.
How often does a perfect pair actually land?
At eight decks, your first two cards are an identical rank and suit 1.69% of the time, about 1 round in 59. A pair of any kind lands 7.47% of the time. These are the Wizard of Odds figures for the Perfect Pairs bet.
Do side bets change the house edge on your main hand?
No. A side bet is a separate wager, so the main hand keeps its own edge. Your blended cost is the stake-weighted average: $5 on the main hand at 0.43% plus $1 on 21+3 at 3.70% comes to 0.975% of the $6 staked.
Does card counting make side bets worth it?
That’s a judgement call, and nothing here is advice on it. Counting tracks which cards remain, and our guide to how card counting works covers the method. Power.Win’s documentation says each Blackjack Original hand’s cards derive from a server seed, client seed and nonce, so there is no shoe to track, and the counting guide explains why.
Sources (Citations)
- Wizard of Odds, 21+3: 100 / 40 / 30 / 10 / 5 to 1 pay table (Version 7), 8 decks, 3.70% edge; “Xtreme” 30 / 20 / 10 / 5 to 1 table, 13.39% at 6 decks, 22.73% at 1 deck, 12.89% at 8. The live-dealer attribution for Version 7 is the page’s own “I hear”. Read 7 October 2026.
- Wizard of Odds, Perfect Pairs: pay tables A to D at 8 decks (4.10%, 3.37%, 7.95%, 2.17%); hit rates; table A tied to live-dealer studios; singular Perfect Pair at 8.05%. Read 7 October 2026.
- Wizard of Odds, Should You Take Insurance in Blackjack?: 2 to 1 payout, 7.4% edge at six decks, even money is the same bet as insurance; declining it returns 103.88% against 100% for taking it (six decks, 3:2 natural). Read 7 October 2026.
- Wizard of Odds, Blackjack basics: 0.43% house edge under Atlantic City rules, stated with no deck count; insurance up to half the original wager. Read 7 October 2026.
- Power.Win Docs, Blackjack: actions, 3:2 natural, dealer stands on 17, up to three hands, “standard 52-card deck”, card draws derived from server seed, client seed and nonce. Read 7 October 2026.
- Power.Win Docs, Games and Fairness FAQ (docs.power.win/documentation/faq/games-fairness): classic blackjack, “dealer stands on soft 17, double on any two cards, split up to three times, blackjack pays 3:2”. Read 8 October 2026.
Take the Lead, Gamble Responsibly
A side bet is a cost you choose to add, and the house edge sits on every one. Decide what a session is worth to you before you start, count the side bets in that figure, and stop when you reach it.
Power.Win’s responsible gambling tools let you set deposit and wagering limits, take a self-exclusion break, and review your bets under Track your Activity. That page also lists independent support services for anyone whose gambling has stopped feeling like a choice.
By the Power.Win Editorial Team.