The baccarat Tie bet is a useful lesson in how casino prices work. In a standard eight-deck game, a Tie occurs on about 9.5156% of completed coups. That is frequent enough to produce memorable wins, but not frequent enough to justify the common 8-to-1 payout. At 8:1, the house edge is about 14.36%. At 9:1, it falls to about 4.84%.
This page is about doing the calculation correctly. For the basic rule and settlement procedure, use Tie Bet Explained. For a direct payout comparison, see Baccarat Tie Bet House Edge.
Start with net profit, not total return
Casino odds such as 8:1 normally describe net profit. A winning $10 Tie bet at 8:1 produces:
- $80 profit;
- the original $10 stake returned;
- $90 total back to the player.
That distinction matters in expected-value work. The win outcome is +8 units, not +9. The loss outcome is −1 unit.
Let:
- (p) = probability of a Tie;
- (b) = net payout in units for each unit staked;
- (EV) = expected net result per unit wagered.
The calculation is:
[ EV = p(b) - (1-p) ]
House edge is the negative of expected value when EV is below zero:
[ \text{House edge} = -EV ]
The 8-to-1 calculation
Using the standard eight-deck Tie probability (p \approx 0.095155968):
[ EV = (0.095155968 \times 8) - (0.904844032 \times 1) ]
[ EV \approx -0.143596288 ]
The expected loss is therefore about 0.1436 units per unit wagered, or a 14.36% house edge.
A player who puts $5 on Tie for 40 coups creates $200 of Tie action. The long-run expected loss is:
[ $200 \times 0.1436 \approx $28.72 ]
The player will not necessarily lose $28.72 in that session. One Tie produces a large jump, while a long run without one produces a string of complete losses. Expected value prices the wager; it does not forecast the next shoe.
Why 9 to 1 changes the price so much
At 9:1, only the payout changes. The event probability is the same:
[ EV = (0.095155968 \times 9) - 0.904844032 ]
[ EV \approx -0.04844032 ]
That is a house edge of about 4.84%. One extra unit of payout removes roughly two-thirds of the casino advantage found at 8:1.
| Posted net payout | Expected return per $1 | House edge | Expected loss per $1,000 wagered |
|---|---|---|---|
| 8:1 | About $0.8564 | About 14.36% | About $143.60 |
| 9:1 | About $0.9516 | About 4.84% | About $48.44 |
The 9:1 version is clearly better. It is still more expensive than the normal Banker and Player bets.
The fair payout is not 9 to 1
A common shortcut says, “Tie happens about one time in ten, so 9:1 must be fair.” The shortcut is too rough.
The fair net payout is:
[ \text{Fair payout} = \frac{1-p}{p} ]
For (p = 0.095155968):
[ \frac{0.904844032}{0.095155968} \approx 9.509:1 ]
A zero-edge Tie wager would therefore need to pay about 9.51 to 1, before considering rounding or any different deck and rule configuration. A 9:1 payout is closer to fair than 8:1, but it still underpays the event.
The exact Tie frequency is not obtained by treating final totals as ten equally likely numbers. Baccarat hands are produced by card ranks, card removal, natural totals, and the fixed third-card rules. Those mechanics make some final-total pairings more common than others. The probability comes from applying the correct dealing rules to the possible card sequences.
A payout-sensitivity shortcut
Once the probability has been established for the exact game, any proposed net payout can be priced with one compact expression:
[ ext{House edge}=1-p(b+1) ]
The term (b+1) includes the net win plus the returned stake. Using the eight-deck probability:
| Net payout | Approximate house edge |
|---|---|
| 8:1 | 14.3596% |
| 8.5:1 | 9.6018% |
| 9:1 | 4.8440% |
| 9.5:1 | 0.0862% |
| About 9.509:1 | 0% before rounding |
This table shows why half-unit changes matter so much on a bet that wins less than one coup in ten. It also shows why a promotional label such as “enhanced Tie” is incomplete without the actual posted odds.
Deck count and approved game variant are inputs, not footnotes. A probability calculated for a standard eight-deck punto banco game should not be copied automatically to a six-deck game, a side bet that uses only the first four cards, or a variant with altered drawing rules. The calculation method remains the same, but (p) must come from the correct rule set.
Average waiting time is not a schedule
The reciprocal of the Tie probability is about 10.51 coups. That is an average spacing over a very large sample, not a promise that a Tie should appear every tenth or eleventh hand.
The probability of seeing at least one Tie in 20 independent coups is approximately:
[ 1-(1-p)^{20} \approx 86.46% ]
That still leaves about a 13.54% chance of no Tie during those 20 coups. More importantly, after 20 misses, the next coup is not “owed” a Tie. The new hand is priced by the cards remaining in the shoe, not by a requirement to repair the scoreboard.
Baccarat roadmaps record outcomes. They do not add a missing Tie to the deck or alter the posted payout. For the broader misconception, see Roadmap Prediction Myth.
The regulator-approved rules describe when each hand draws and how a Tie wager settles. The Massachusetts baccarat rules provide a useful primary reference for the drawing procedure and require a winning Tie wager to pay at least 8:1. The table’s actual posted payout still determines the mathematical price.
A Tie result affects three wagers differently
Suppose a player has:
- $50 on Banker;
- $10 on Tie;
- no other side bets.
If the final totals are equal, the normal Banker wager pushes and the Tie wager wins. At 8:1, the player receives the $50 Banker stake back and earns $80 profit on Tie. If Banker or Player wins instead, the $10 Tie wager loses.
The push on the main bet does not make the Tie wager free. Each wager has its own resolution rule and expected value.
Errors that distort the calculation
Using total return as net profit. Treating 8:1 as +9 units produces the wrong expected value.
Rounding the probability to 10%. That makes 9:1 appear fair when it is not.
Comparing payout size without frequency. A high award can coexist with a high house edge.
Using hands played instead of dollars wagered. Session cost depends on total Tie action. Forty $1 bets and forty $25 bets have the same edge but very different expected dollar loss.
Treating a hit as proof of value. A single winning Tie says nothing about whether the posted price was fair.
Combining unlike versions. Six-deck and eight-deck games, alternative drawing rules, or different Tie payouts require their own inputs. Never copy one edge into a different version without checking the rules.
Use the number as a price tag
The useful decision is simple: verify the posted payout before placing the wager. Moving from 8:1 to 9:1 substantially improves the price, but neither payout turns Tie into a low-cost substitute for Banker or Player.
For the event probabilities, continue to Baccarat Odds. To compare posted awards with fair awards, read Baccarat Payouts vs True Odds. The expected-loss calculator can convert an edge and planned total action into a dollar estimate.