Key Takeaways
- Both games are close to 50/50 on the main bet, but they are priced differently: Andar Bahar's published RTP is 97.1% (a 2.9% house edge) and Dragon vs Tiger's is 96.3% (3.7%).
- In Andar Bahar the side that receives the first card wins about 51.5% of rounds in a single-deck game — which is exactly why payouts have to be adjusted, and why "always bet the first side" is not a strategy.
- Dragon vs Tiger's 96.3% RTP falls out of the maths of a 6- or 8-deck shoe when a tie returns half the main bet — 96.30% and 96.27% respectively.
- The Tie bet is the most expensive bet in the game: at a quoted 8× payout it carries an expected loss of roughly 33–40% of every rupee staked, around ten times the main bet's cost.
- Streaks are normal, not signals. In 100 fair 50/50 rounds there is an 81% chance of a run of six or more on the same side, so "roadmap" patterns tell you nothing about the next round.
- On the main bet, Andar Bahar is the cheaper game — about ₹29 expected cost per ₹1,000 wagered versus ₹37 — but the gap is small next to the swings in any single session.
Andar Bahar and Dragon vs Tiger are the two card games most often chosen by players who want something faster than Teen Patti and simpler than Blackjack. You pick a side, one card or one short sequence decides the round, and thirty seconds later you are picking again. On the surface they look interchangeable. Underneath, the rules, the exact probabilities and the cost of each type of bet are noticeably different.
This guide takes both apart. It covers how each game works, the real probabilities behind the deal, how a published RTP figure is built from those probabilities, why side bets are so much more expensive than the main bet, why streak-reading doesn't work, and which game gives better value. As always on this blog, none of it promises a way to beat the house edge — nothing does. It is about understanding what you are paying for. If any of the terms below are unfamiliar, our RTP, house edge and volatility guide explains them from scratch.
Why These Two Games Get Compared
Both games share four traits that make them popular and easy to misjudge. Rounds are quick. The rules take under a minute to learn. The main bet is close to an even-money proposition, which feels like a coin flip and invites the assumption that the game is "fair". And both have a strong culture of pattern-watching — scoreboards, roadmaps and "the last six rounds went Dragon" chatter — that gives the impression of skill where the underlying maths says there is none.
The differences are in the details. One game deals a variable number of cards until a match appears, the other deals exactly two. One pays even money on a slightly lopsided outcome, the other on a symmetrical one with a tie complication. Those details determine the house edge, the pace and, importantly, which bets are worth avoiding altogether.
How Andar Bahar Works
Andar Bahar is a traditional Indian card game played with a standard 52-card deck. The name translates roughly to "inside" and "outside", which are the two positions you can bet on.
A joker card is revealed
One card is placed face-up. Its rank (for example, a 7 or a Queen) is the target for the round — suit doesn't matter.
You choose Andar or Bahar
Before dealing begins you bet on which of the two piles the matching card will land in.
Cards are dealt alternately
The dealer places cards one at a time, alternating between the Andar and Bahar piles, starting with one of them.
The first match ends the round
As soon as a card matching the joker's rank appears, the pile it lands in wins and the round is settled.
Come Game's version is described as completing rounds in under thirty seconds, which fits the game's structure: there is only one decision (which side) and no betting between deals. Which pile receives the first card, and the exact payout for each side, are shown in the game's rules panel and can differ between tables and platforms — so read it before your first round, as that detail matters for the maths that follows.
The Real Odds Behind Andar Bahar
Andar Bahar has a structural feature that most players never notice: the two sides are not equally likely to win. After the joker is removed, 51 cards remain in a single deck, 3 of which match its rank. Cards are dealt alternately, so the side that receives the first card also receives every odd-numbered card (1st, 3rd, 5th and so on), and the other side receives the even-numbered ones.
Working through the probability that the first matching card falls in an odd position gives:
| Setup | Side dealt first wins | Side dealt second wins |
|---|---|---|
| Single 52-card deck | 51.50% | 48.50% |
| 8-deck shoe | 51.94% | 48.06% |
Calculated from the position of the first matching card among the remaining cards. The advantage of the first-dealt side is small but real, and it is a direct consequence of dealing order.
The reason is intuitive once you see it. Whoever is dealt first gets the first chance at every match; the other side only wins if the match arrives at an even position. Over thousands of rounds that head-start adds up to roughly a 3-point gap between the sides.
Why the payouts can't both be 1-to-1
This gap has an important consequence. If both sides paid a straight 1-to-1 and the first-dealt side won 51.5% of the time, that side would have a positive expected return of about 3% — a bet the house would lose on. No real game offers that, so platforms adjust something: commonly the payout on the first-dealt side is reduced (often quoted as slightly under even money), or other round rules are tweaked. The precise mechanism varies between tables, which is one reason to check the rules panel rather than trusting a rule of thumb.
Practically, this means "always bet the side that gets the first card" is not a winning system — the price has already been adjusted for that advantage. Against the published 97.1% RTP, the average cost of playing Andar Bahar is 2.9 rupees for every 100 wagered, whichever side you choose.
How long does a round last?
Because the dealing continues until a match appears, round length varies — which affects both the pace and how the game feels. With 51 cards and 3 matches, the average round runs to 13 cards and the median to 11:
| By card number | Chance the round is over |
|---|---|
| 1st card | 5.9% |
| 5th card | 27.1% |
| 10th card | 48.8% |
| 15th card | 65.7% |
| 20th card | 78.4% |
| 30th card | 93.6% |
Single-deck calculation. Roughly one round in sixteen runs beyond 30 cards — most are short, a few are long, and there is no way to predict which is which.
How Dragon vs Tiger Works
Dragon vs Tiger, sometimes described as the simplest card game in the casino, dispenses with dealing sequences altogether.
You place your bet
Choose Dragon, Tiger, or one of the side bets — Tie, Dragon Pair or Tiger Pair, according to Come Game's description of the game.
One card each is dealt
The Dragon side and the Tiger side receive one card apiece, face up.
The higher card wins
In most versions the Ace is the lowest card and the King is the highest, and suits are ignored. Confirm the ranking in the rules panel.
Payout
A winning main bet is described on the games page as paying 2× your stake (your stake back plus an equal amount). A tie is the special case covered below.
The Maths Behind Dragon vs Tiger
Because only two cards are dealt, the maths is compact enough to check by hand. The key number is the probability of a tie — both sides receiving the same rank — which depends on how many decks are in the shoe.
| Decks in shoe | Chance of a tie | Dragon wins | Main-bet house edge* | Implied RTP* |
|---|---|---|---|---|
| 1 | 5.88% | 47.06% | 2.94% | 97.06% |
| 6 | 7.40% | 46.30% | 3.70% | 96.30% |
| 8 | 7.47% | 46.27% | 3.73% | 96.27% |
*Assumes a winning main bet pays 1-to-1 and a tie returns half the main-bet stake, a common convention in Dragon Tiger. Dragon and Tiger win with equal probability, so the result applies to either side.
Look at the last column. Under the half-back-on-tie convention, a 6-deck or 8-deck shoe gives an RTP of 96.30% or 96.27% — which rounds to the 96.3% that Come Game publishes for Dragon vs Tiger. That is consistent with the platform using a multi-deck shoe and that tie rule. It isn't proof — the rules panel is the authority — but it shows how a published RTP figure is not arbitrary: it is the arithmetic outcome of specific rules. If a tie instead lost the whole main bet, the edge would be about 7.4%, roughly double the published figure, which is another reason the half-back convention is the likelier reading.
About 7.5% of rounds are ties, and on each tie the main bet loses half its stake. Half of 7.5% is roughly 3.7% — that is the entire house edge on the main bet. It comes from the tie rule, not from any imbalance between Dragon and Tiger.
Why the Tie and Pair Side Bets Cost So Much More
The Tie bet looks generous. Come Game's description quotes an 8× payout on a tie, and a payout that big on an event that "just needs both cards to match" is attractive. The way to test any bet is the fair-odds test: a bet quoted at an N× total return only breaks even if it wins at least 1 time in N. An 8× payout would need ties to occur 12.5% of the time. They occur about 7.5% of the time in a multi-deck shoe.
| Bet | Wins how often | Expected loss per ₹100 staked | Cost per ₹1,000 wagered |
|---|---|---|---|
| Main bet (Dragon or Tiger), 8 decks | 46.27% | ₹3.73 | ₹37 |
| Tie, if "8×" = 8-to-1 profit | 7.47% | ₹32.77 | ₹328 |
| Tie, if "8×" = total return (7-to-1 profit) | 7.47% | ₹40.24 | ₹402 |
8-deck shoe. Games quote payouts in different conventions, so both readings are shown; the rules panel states which applies. Either way, the tie bet is roughly nine to eleven times as expensive as the main bet.
Put in session terms: twenty ₹100 tie bets is ₹2,000 wagered, with an expected loss of about ₹655 to ₹805 — while ₹2,000 wagered on the main bet has an expected loss of about ₹75. And the Tie only wins about once in every thirteen rounds, so the stretches between wins are long.
The Dragon Pair and Tiger Pair bets, quoted at 11×, follow the same logic: a bet at an 11× total return needs to win about 9.1% of the time to be fair, so the pair bet is only reasonable value if its true probability sits at or above that level. Before ever placing a side bet, compute 1 ÷ payout and compare it with how often the outcome really happens. High payouts on rare events are how the house edge is concentrated — not a bonus for the player.
🎲 Try Andar Bahar or Dragon vs Tiger on Come Game
Check each game's rules panel for the exact payouts and ranking first. New members get a ₹10,000 welcome bonus on their first UPI deposit — read the terms before claiming.
Streaks, Roadmaps and the Pattern Trap
Both games come with a visual history — a scoreboard of recent results, often drawn as coloured dots in a grid that players call a "roadmap". Some players read these as signals: a long run on one side means "it's due to switch", or "it's on fire, follow it". Both readings are versions of the same mistake.
If each round is independent — as it is when the deck is reshuffled or generated by a random number generator each round — the previous result carries no information about the next. And streaks are far more common than intuition suggests. In a fair 50/50 sequence:
| Streak length | In 50 rounds | In 100 rounds | In 200 rounds |
|---|---|---|---|
| At least 5 in a row | 82.1% | 97.2% | 99.9% |
| At least 6 in a row | 54.4% | 80.7% | 96.5% |
| At least 7 in a row | 30.9% | 54.2% | 79.9% |
| At least 8 in a row | 16.2% | 31.5% | 54.2% |
Probability of at least one run of that length on the same side, in a sequence of independent 50/50 outcomes. A streak of six is not unusual — it is close to guaranteed over a session of a hundred rounds.
A pattern-follower who sees seven Dragon results in a row and concludes something has changed is reacting to an event that happens more than half the time in a hundred-round session. The game has no memory. Backing "the streak" and backing "the reversal" have the same expected return, which is the house edge in both cases. The only reliable effect of watching roadmaps is that it encourages more bets, faster, on a game that already runs quickly.
Andar Bahar vs Dragon vs Tiger, Side by Side
| Andar Bahar | Dragon vs Tiger | |
|---|---|---|
| Published RTP | 97.1% | 96.3% |
| House edge (main bet) | 2.9% | 3.7% |
| Expected cost per ₹1,000 wagered | ₹29 | ₹37 |
| Cards per round | Variable — average 13, median 11 | Exactly 2 |
| Round length | Varies; under 30 seconds per Come Game's description | Very short and consistent |
| Decisions per round | One (which side) | One main choice, plus optional side bets |
| Where the edge comes from | Payout and round-rule adjustments to offset the first-dealt side's advantage | Ties, which lose half the main bet |
| Side bets | Check the rules panel | Tie, Dragon Pair, Tiger Pair — all far more expensive than the main bet |
| Best for | Players who like a little suspense as cards build to a match | Players who want maximum simplicity and speed |
What a Session Actually Costs
Consider a player who bets ₹50 on the main bet for 100 rounds. That is ₹5,000 wagered. The expected loss is the wagered amount times the edge:
- Andar Bahar (2.9%): about ₹145 expected loss.
- Dragon vs Tiger (3.7%): about ₹185 expected loss.
The ₹40 difference between the two is real, but look at it beside the variance. The typical swing in the result of 100 even-money ₹50 bets is around ₹500 either way (the stake times the square root of the number of rounds). So in any single session the result is dominated by luck — you might be up ₹350 or down ₹650 — and the edge only becomes visible across many sessions. That is why the cheaper game feels no different in the short run, and why the discipline that matters most is choosing your session size in advance, as covered in our bankroll management guide.
Betting Approaches: What Helps and What Doesn't
- Flat betting. Staking the same amount every round is the simplest and most predictable approach. It does not change the expected loss, but it keeps the size of a bad session knowable in advance.
- Skipping side bets. This is the single easiest saving available in Dragon vs Tiger. Leaving the Tie and Pair bets alone cuts the cost per rupee wagered by roughly a factor of ten.
- Progression systems. Doubling after a loss (Martingale) feels safe in a near-50/50 game, since a win is always close. The numbers say otherwise: seven straight losses, which is far from rare, would turn a ₹100 opening bet into a ₹12,800 stake. The full arithmetic is in the bankroll guide linked above.
- Choosing sides on "form". Following streaks or betting on reversals makes no difference to expected return. If it makes the game more enjoyable, it is harmless as long as the stake stays constant — but it isn't skill.
- Setting a round or time cap. Fast games reward a hard stop. Decide the number of rounds or minutes before you begin, not after the first swing.
So Which One Is Better Value?
On the numbers alone, Andar Bahar's main bet is the better value: a 2.9% edge against 3.7%, or ₹29 against ₹37 per ₹1,000 wagered. The gap is modest, and it disappears if you add Dragon vs Tiger's side bets, which push the effective cost of those wagers far higher. Its only real weakness is that the round-to-round rhythm varies, with an occasional very long round.
Dragon vs Tiger's advantages are consistency and simplicity: every round is one card each, the pace is steady, and once you skip the side bets there is nothing to decode. If you prefer that rhythm, the extra 0.8 percentage points of edge is the price of it — comparable to a small entry fee, not a reason to avoid the game.
The honest summary is that neither is a good way to make money, and both are reasonable ways to spend a set budget on entertainment. If cost per rupee is your priority, choose Andar Bahar and ignore any side bets; if you want speed and clarity, choose Dragon vs Tiger and ignore the side bets there too. To see how they compare with the rest of the catalogue, our guide to choosing between all 14 Come Game titles lays them out side by side.
Playing Responsibly
The features that make these games appealing — near-even outcomes, short rounds, an instant next decision — are the same ones that make it easy to lose track of how much has been wagered. Ten rounds feel like a warm-up; a hundred is a real session. Set a session budget and a time limit before you start, decide in advance that side bets are off the table, and step away when either limit is reached rather than after one more round. If you find yourself raising stakes after a loss to get back to even, that is the moment to stop for the day. The responsible-gaming setup guide shows how to lock a deposit limit and session reminder in about ten minutes.
Real-money gaming is restricted in Andhra Pradesh, Assam, Odisha, Telangana, Nagaland and Sikkim, and is strictly for adults 18 and over everywhere else. For support tools and helplines in India, see the responsible gaming page.
The Bottom Line
Andar Bahar and Dragon vs Tiger both give you a near-even choice every half-minute, but they aren't the same product. Andar Bahar's structure gives the first-dealt side a real head-start that the payouts have to price in, and it costs about 2.9% per rupee wagered. Dragon vs Tiger is priced at roughly 3.7% through its tie rule, and its Tie and Pair side bets are the most expensive options on the table by a wide margin.
Everything else people say about these games — patterns, streaks, "due" outcomes, systems — falls apart against the arithmetic. What is left is a straightforward choice: which pace you like, which price you are comfortable paying, and how much you have decided in advance to spend.
