Super Andar Bahar RTP & Odds: Andar, Bahar and Card Count House Edge (2026)
By LiveCasinoData editorial team · Updated 2026-08-28 · 12 min read
Super Andar Bahar returns 97.85% on Andar and 97.00% on Bahar, a rare case where the even-money side (Bahar) carries the lower theoretical RTP. The ten Card Count side bets run roughly 95% to 97% depending on the bucket, with Evolution's Lightning feature boosting one to five buckets by 20x to 200x or more each round. Our tracker has logged 1,371 rounds since 2026-08-27, an early sample that already lines up closely with the published Andar/Bahar split.
Key takeaways
- Andar pays 0.9:1 with a theoretical RTP of about 97.85%, and Bahar pays even money (1:1) with a theoretical RTP of about 97.00%.
- Andar wins more often than Bahar (51.5% versus 48.5% in our configuration), which is why the shorter payout still produces the higher RTP of the two.
- The ten Card Count side bets pay from 2:1 on 1-5 cards up to 800:1 on 46-49 cards, with published RTPs running roughly 95% to 97% per bucket (verify exact figures).
- Evolution's Lightning feature boosted 249 of our first 1,371 tracked rounds (18.16%), multiplying the winning Card Count bucket's payout for that round only.
- Third-party sources cite a 4,000x maximum win on the 46-49 Card Count bet (verify), while our own feed's largest recorded payout so far is 2,500x on a single occurrence.
Super Andar Bahar pays Andar at 0.9:1 for a theoretical RTP near 97.85%, and pays Bahar at even money for a theoretical RTP near 97.00%. The ten Card Count side bets run roughly 95% to 97% depending on the bucket, boosted round by round when Evolution's Lightning feature strikes. Our live tracker has logged 1,371 rounds since this game joined our coverage on 2026-08-27, and this page checks that early sample against the published numbers.
What is the RTP of Super Andar Bahar?
Andar returns about 97.85% in theory and Bahar returns about 97.00%, both figures published in Evolution's game rules for the main bet. That puts the house edge at roughly 2.15% on Andar and 3.00% on Bahar, a gap of under a full point between the two sides of the same deal.
The ten Card Count side bets sit below the main bets, in a broad 95% to 97% band that varies by bucket (exact per-bucket figures are unconfirmed here, treat as verify). Those RTPs already assume the Lightning multiplier is part of the design, since Evolution builds the boost into the bet's published return rather than treating it as a bonus layered on top.
None of these numbers move during a round. The dealer's shuffle and the card-matching mechanic are fixed by the rules of Andar Bahar; what changes from spin to spin is only which side happens to catch the match and, on the side bet, whether a Lightning strike lands on the bucket that wins.
What are the true payouts and win rates for Andar and Bahar?
Andar and Bahar are not mirror images of each other, because one deals first. The dealer places the Joker face up, then deals cards alternately starting with Andar, and the round ends the instant a card of the same rank as the Joker appears on either side.
| Side | Payout | Theoretical RTP | Deals first |
|---|---|---|---|
| Andar | 0.9 : 1 | ~97.85% | Yes |
| Bahar | 1 : 1 (even money) | ~97.00% | No |
Because Andar gets the first card of every round, it has a small structural edge in how often it catches the match, which our configuration puts at 51.5% against Bahar's 48.5%. Andar's payout is shaved down to 0.9:1 to compensate for that edge, and the shave is calibrated to leave Andar with the higher RTP anyway. Bahar pays a clean 1:1 but wins less often, so its return lands a shade lower.
Why does Bahar carry a lower RTP than Andar despite paying even money?
Because win probability and payout pull in opposite directions on this game, and the payout structure is what separates the two RTPs, not the raw odds. In most banker-style live games, the side with worse win odds also has the worse payout table, so the gap is easy to predict. Super Andar Bahar inverts part of that logic.
Andar wins more often (51.5%) but is paid less per win (0.9:1). Bahar wins less often (48.5%) but is paid the full unit per win (1:1). Run the arithmetic and Andar's frequent-but-smaller wins add up to a slightly larger long-run return than Bahar's rarer-but-full-sized wins. Neither side is a mistake to bet; the difference between 97.85% and 97.00% amounts to less than one unit of edge per hundred wagered, smaller than the gap between many roulette bets and their neighbors.
What this means at the table is straightforward. Choosing Andar over Bahar shaves a little off the house's cut, but it will not turn a losing session into a winning one, and it says nothing about which side is due on the next hand. Our strategy page walks through how to size stakes around that small difference without overstating what it buys you.
What is the Card Count side bet and how does it pay?
The Card Count side bet predicts how many total cards will be dealt, across both hands, before one matches the Joker's rank. Ten overlapping buckets cover the full range from a fast 1-5 card match to a marathon 46-49 card match, and only one bucket wins per round.
| Card range | Base payout |
|---|---|
| 1-5 | 2:1 |
| 6-10 | 3:1 |
| 11-15 | 4:1 |
| 16-20 | 5:1 |
| 21-25 | 8:1 |
| 26-30 | 12:1 |
| 31-35 | 20:1 |
| 36-40 | 40:1 |
| 41-45 | 110:1 |
| 46-49 | 800:1 |
The paytable climbs steeply because the buckets are not equally likely. As more cards get dealt without a match, the deck thins and the odds of the next card matching the Joker's rank actually rise, so long counts are the rarer outcome and the low-card buckets dominate. Our first 1,371 tracked rounds landed 27.7% in the 1-5 bucket and only 0.1% (2 rounds) in 46-49, a shape consistent with that logic even at this early sample size.
Base payout alone does not tell you the full return on a bucket; the Lightning multiplier is baked into the published 95% to 97% RTP range, and the next section shows what that boost looks like in our own data.
How much of the Card Count payout comes from the Lightning multiplier?
A meaningful share. Before each deal, Evolution's RNG selects between one and five of the ten Card Count options and multiplies that bucket's payout for the round only, and our collector flags this as isLightningHand. Across our first 1,371 rounds, 249 (18.16%) were flagged, split almost evenly between Andar-won and Bahar-won rounds (122 versus 127), which is expected since the side bet outcome has nothing to do with which main side wins.
If we treat our early bucket shares as a rough stand-in for true probability, the base paytable alone (no Lightning) returns somewhere in the 70s-to-low-90s percent range depending on the bucket, well under the 95-97% published figure. That gap is exactly what the multiplier exists to close, the same relationship you would find on any live game that pairs a conservative base paytable with a random boost feature. The rare buckets in our sample, 41-45 and 46-49, had only six and two rounds respectively, so any base-return estimate on them is closer to noise than to a stable figure and should not be read as precise.
| Card range | Base payout | Observed boosted range | Sample size note |
|---|---|---|---|
| 1-5 | 2:1 | 5x-10x | 27.7% of rounds, largest bucket |
| 6-10 | 3:1 | 6x-10x | 21.1% of rounds |
| 11-15 | 4:1 | 8x-20x | 16.4% of rounds |
| 16-20 | 5:1 | 10x-25x | 12.7% of rounds |
| 21-25 | 8:1 | 15x-20x | 9.3% of rounds |
| 26-30 | 12:1 | 25x-60x | 6.1% of rounds |
| 31-35 | 20:1 | 40x-60x | 4.4% of rounds |
| 36-40 | 40:1 | 80x-100x | 1.8% of rounds |
| 41-45 | 110:1 | 200x (1 sample) | 0.4% of rounds, tiny sample |
| 46-49 | 800:1 | 2,500x (1 sample) | 0.1% of rounds, single occurrence |
Across the 249 lightning-flagged rounds in our sample, the average boosted multiplier was 25.99x, the median was 8x, and the 90th percentile was 36x. The single 2,500x round on the 46-49 bucket sits far above that typical range, which is exactly what you would expect from one rare event rather than a representative outcome.
How volatile is a bet on Andar or Bahar compared with Card Count?
Very different. Andar and Bahar are near-coin-flip bets with a hit roughly every other round, so a session on either side moves in small, frequent steps. The Card Count buckets range from almost as steady (1-5 hits close to once every four rounds) to extremely spiky (46-49 hit twice in our entire 1,371-round sample so far).
That spread in hit frequency is also a spread in bankroll behavior. A chip on Andar or Bahar rarely produces a long losing run relative to its win rate, so swings stay contained. A chip on a high Card Count bucket can go dozens or hundreds of rounds without paying, and when it does pay, the base 800:1 (or a Lightning-boosted multiple of it) can swing a session in a single hand. Neither approach is safer in the sense of expected return; they differ in how the same house edge gets delivered, in steady sips on the main bets or in rare, large jumps on the rare Card Count buckets.
Our early gap data adds one more data point: the average wait between Lightning-boosted rounds (any bucket) was 5.52 rounds, with the longest observed gap so far at 31 rounds. Using our observed 18.16% Lightning rate as a rough estimate, the chance of a 10-round stretch with no Lightning hand at all works out to about 0.818 raised to the 10th power, near 13%, a descriptive figure from a small sample rather than a certified probability.
What is the maximum win in Super Andar Bahar?
Third-party sources report a ceiling around 4,000x stake, reachable by landing a Lightning-boosted 46-49 Card Count bet at the high end of its multiplier range (verify against Evolution's own in-client help, as we have not independently confirmed the exact 4,000x figure or whether it is quoted as 3,999x plus stake). We treat this as unverified until we can check it against the operator's own rules panel.
What we can confirm from our own feed is smaller: the largest multiplier recorded across our first 1,371 rounds is 2,500x, from a single 46-49 Card Count round. That single data point does not prove or disprove the 4,000x ceiling; it simply reflects that our tracker has not yet observed a round anywhere near the reported top end, which is unsurprising given how rare the 46-49 bucket is (0.1% of rounds so far) and how much rarer a maximum-strength Lightning strike on that specific bucket would be. Expect this number to firm up as more rounds accumulate.
What are the table limits?
Third-party reporting puts the minimum bet at roughly 0.50 per spot and the maximum around 10,000 per spot, though we have not independently verified these against the operator (verify). As with any live table, limits set how much you can risk per round; they do not change the underlying RTP of Andar, Bahar, or any Card Count bucket, which is fixed by the payout table and the Lightning distribution regardless of stake size.
Rounds run roughly 30 to 45 seconds of betting time before the deal, dealt continuously around the clock, which keeps the pace brisk relative to slower-paced table games. The how-to-play guide covers the betting window and interface in more detail for players new to the format.
Does our observed data match the published RTP figures?
So far, yes, on the piece we can check directly. Our tracker recorded Andar winning 707 of 1,371 rounds (51.57%) and Bahar winning 664 (48.43%), against a configured theoretical split of 51.5% Andar and 48.5% Bahar. That is a close match at this sample size, and it is the kind of check we can run without needing Evolution's internal RTP model, since main-side win frequency is directly observable from round outcomes.
What we cannot yet verify independently is the exact RTP percentage itself, because that depends on payout math applied consistently over a very large number of rounds, and our history here is short. This is a brand-new addition to our coverage: the first round was captured on 2026-08-27, and as of 2026-08-28 we have just under two days of data, well short of the sample needed to treat any of these figures as a stable long-run baseline. Card Count bucket shares, Lightning strike rates, and multiplier distributions should all be read as an early snapshot that will shift, sometimes noticeably on the rarer buckets, as the dataset grows.
The Joker rank distribution offers one more sanity check: across all 13 ranks over 1,371 rounds, counts ranged from 86 to 131 against an average of about 105 per rank, which is roughly what a fair single-deck draw should produce and shows no obvious skew beyond ordinary small-sample noise.
How should you read our Super Andar Bahar RTP data?
Start with the live stats page, which breaks the Andar/Bahar split, the Card Count bucket distribution, and the Lightning Hand rate down over selectable windows. Because this game is new to our tracker, every window right now is effectively the same short window, so treat any single day's figures as a small sample rather than a settled pattern.
The main-side split is the easiest number to trust early, because it only needs a moderate sample to stabilize and our 51.57%/48.43% reading already sits close to theory. The Card Count bucket shares and the Lightning multiplier statistics need far more rounds before the rare buckets (41-45, 46-49) mean anything, since those currently rest on single-digit counts. Read the common buckets (1-5 through 16-20) with more confidence than the tail.
The drought tracker on the live page counts rounds since a bucket or a Lightning strike last occurred. A long gap says nothing about the next round's odds; the cards have no memory of the last shuffle, and a bucket that has gone quiet is not due to hit. The same caution applies to Andar or Bahar after either side strings together several wins in a row.
One broader point is worth carrying into any session. A 97-98% RTP on the main bets, or a 95-97% RTP on a Card Count bucket, describes an average outcome across many thousands of rounds, not a promise about your next ten hands. No amount of watching the drought tracker changes the probability of the next card. Set a budget before you sit down, treat any streak you see as a description of the past rather than a signal about the future, and use our RTP monitor and responsible gambling observatory if you want to see how this game's early numbers compare with the rest of our coverage as more data comes in. For the mechanics of the Lightning feature and Card Count bet in more depth, see the bonus-rounds guide; for the game overview, the full guide is the place to start.
Frequently asked questions
What is the RTP of Super Andar Bahar?
Andar carries a theoretical RTP of about 97.85% at 0.9:1, and Bahar carries about 97.00% at even money, both figures from Evolution's published game rules. The ten Card Count side bets sit lower, in a roughly 95% to 97% range depending on the bucket (verify exact per-bucket figures), before accounting for the random Lightning multiplier.
Is Andar or Bahar the better bet in Super Andar Bahar?
Andar has the higher published RTP, 97.85% against Bahar's 97.00%, so it returns marginally more per unit staked over the long run. The gap is small, under a full percentage point, and both bets carry the same house-edge logic: no side can be beaten by timing or pattern reading, only by staking less over time.
How does the Card Count side bet pay in Super Andar Bahar?
It pays according to how many total cards are dealt before a card matches the Joker's rank, from 2:1 on a 1-5 card match up to 800:1 base on a 46-49 card match. Low counts are common and pay little; high counts are rare and pay a lot, which is standard structure for a rarity-scaled side bet.
What is the maximum win on Super Andar Bahar?
Third-party reviews cite a ceiling near 4,000x stake on a Lightning-boosted 46-49 Card Count bet, though we have not independently confirmed the exact figure against Evolution's in-client help (verify). Our own feed's largest recorded multiplier so far is 2,500x on one occurrence, from an early and still-small sample.
Does Super Andar Bahar pay out fairly according to tracked data?
Our early sample of 1,371 rounds shows Andar winning 51.57% of the time against a configured theoretical split of 51.5%, and Bahar at 48.43% against 48.5%, a close match. The sample is under two days old as of 2026-08-28, so treat it as an early check rather than a confirmed long-run baseline.