Does a court still need to decide whether online rummy is a game of skill before it can shut the game down? Not any more, and that single shift explains Junglee Games’ latest defeat better than any argument about probability. Junglee lost its appeal before India’s Supreme Court, a national-level ruling upholding state powers to ban real-money gaming, as reported by GamblingNews.com. Flutter’s subsequent withdrawal of Junglee from India was reported by SBC News. The loss matters less for what it says about rummy than for what it says about real money gaming regulation sitting with legislatures rather than with game theory.
The popular conclusion in the industry is correct: operators should stop building commercial strategy on the skill-versus-chance distinction. But the usual reason given for that conclusion is wrong. Judges have not quietly decided that rummy is luck. They have decided that the classification question is not the one that settles the case.
The skill defence stopped being load-bearing
For roughly six decades, Indian real-money card operators rested on one doctrine: games that are preponderantly skill fall outside “gambling” and therefore outside prohibition. That doctrine still exists. It just no longer decides whether an operator can take a stake from a player in a given state.
The precedent trail is worth laying out plainly, because it shows a defence that kept winning on the legal point while losing on the practical one.
| Case | Forum and year | Outcome |
|---|---|---|
| State of Andhra Pradesh v. K. Satyanarayana | Supreme Court, 1968 | Rummy held to be preponderantly a game of skill |
| Dr K.R. Lakshmanan v. State of Tamil Nadu | Supreme Court, 1996 | Horse race betting treated as skill-based, not gambling |
| Junglee Games India v. State of Tamil Nadu | Madras High Court, 2021 | Blanket ban on online games played for stakes struck down |
| All India Gaming Federation v. State of Karnataka | Karnataka High Court, 2022 | State amendment restricting online skill games struck down |
| Challenge to the Tamil Nadu Online Gaming Act | Madras High Court, 2023 | Act upheld in general, but held incapable of applying to online rummy and poker, classified as games of skill |
Read that sequence in order. Junglee won the 2021 case as the named petitioner. Tamil Nadu then legislated again, with findings about addiction and financial harm attached, and the second statute survived in substance, although the Madras High Court held in 2023 that it could not reach online rummy and poker, which it classified as games of skill. The skill reasoning kept winning on its own terms. What eventually routed around it was not a better argument about probability but a central prohibition plus a Supreme Court ruling confirming that states may ban real-money gaming.
Jurisdiction, not probability, is doing the work
The cause of the fragmentation is constitutional, not mathematical. Betting and gambling sit in Entry 34 of the State List under India’s Constitution, which hands each state primary power to legislate on the subject. A game can therefore be legal to offer for money in one state and criminal in the next, with the same cards, the same software and the same RNG certification behind it.
That alone would be manageable. What makes it unmanageable is the second layer. The Promotion and Regulation of Online Gaming Act, 2025 introduced a central prohibition on online money games, with criminal exposure for offering them, advertising them, or processing payments for them, while explicitly encouraging e-sports and non-money social gaming. Constitutional challenges to that real money gaming ban are live. So an operator now faces a central statute, a state statute, and litigation over both, any one of which can independently close the business line.
This is why the Supreme Court’s rejection of Junglee’s appeal is not a footnote. The ruling did not settle the fate of the central prohibition. It confirmed that states retain the power to ban real-money gaming outright, which means a single state statute still has the practical force to end a product in a market of tens of millions of people, regardless of how the national question over the 2025 Act resolves.
The cost of the old thesis
Flutter Entertainment bought into Indian real money gaming on a thesis that was reasonable at the time: a large, young, mobile-first player base, a long-standing Supreme Court doctrine protecting games of skill, and a regulatory vacuum that would eventually fill in the industry’s favour. Two of those three held. The third did not.
The error was treating a judicial doctrine as a licence. A licence is granted by a regulator, carries conditions, and is portable across time because it is administrative. A doctrine is a reading of a statute, and statutes get rewritten. Any operator whose India model depended on Satyanarayana holding was depending on a legislature choosing not to legislate.
The practical consequence for anyone holding exposure to a skill gaming law regime, in India or anywhere else: your market access is only as durable as the weakest legislature with power over it. For a federal country, that is a low bar.
Taxonomy or harm, and no clean answer
Here is the trade-off the Junglee ruling forces into the open, and it has no tidy resolution.
Classification-based regulation draws the line at whether outcomes are determined mainly by skill. Its virtue is predictability: an operator can engineer a product to sit on the right side of the line and know where it stands. Its weakness is that the line correlates poorly with player harm. A skilful game played at high speed, for high stakes, funded by instant deposits, can empty an account faster than a slot machine. Classification does not see that, because it looks only at the game and never at the session.
Harm-based regulation looks at stake size, loss velocity, deposit friction, advertising reach and addiction data. Its virtue is that it regulates the thing that actually hurts people. Its weakness is that it offers no safe harbour. Once harm is the test, a genuinely skilful game with a thin margin gets swept in alongside a high-volatility slot, and the operator has no argument left other than pleading for proportionality. Blanket prohibition is the cheapest version of harm-based regulation, and cheap tends to win in legislatures under public pressure.
Neither model is obviously correct. The classification model protected a real distinction and was still used, for years, to market products that behaved like gambling to players who experienced them as gambling. The harm model answers that objection and in doing so erases a distinction that is, mathematically, real. India has moved toward the second. Most of Europe and the regulated US states sit closer to the first, with licensing conditions bolted on. Watching which produces better player outcomes over the next decade is the open question, and nobody has the data yet.
Operating inside rules that do not line up
The workable response is unglamorous, and it starts by abandoning the hope of a single decisive precedent.
- Treat every state, province or sub-national unit as its own market with its own legal opinion, its own geo-fencing rule and its own kill switch. Build the product so a single jurisdiction can be switched off without touching the rest.
- Assume the classification argument will be tested and lose at least once. Price market entry on the assumption that stake-based revenue is temporary, not perpetual.
- Design harm controls before a regulator demands them. Deposit and loss limits, session reminders, cooling-off and self-exclusion are the features that make a proportionality argument credible when prohibition is on the table. An operator that already caps stakes has something to point at; one that does not is arguing from nothing.
- Keep a non-money format viable. The 2025 Indian framework carves out e-sports and social gaming for a reason, and the companies that survived the transition were the ones that already had a product with no stake attached.
- Separate the legal opinion from the commercial forecast. The question “is this lawful today” and the question “will this be lawful in three years” have different answers and should be documented separately.
Players reading this should take the same lesson from the opposite side. A game being classified as skill-based says nothing about whether it is safe to play for money, and nothing about your expected return. Set a deposit limit before you need one, treat stakes as spend rather than investment, and use the self-exclusion tools every licensed operator is obliged to offer.
What the ruling leaves unsettled
The online rummy ruling against Junglee does not establish that skill games are gambling. It establishes something narrower and more awkward: that a legislature can act on harm without first winning the taxonomy argument, and that courts will let it. Whether India’s Supreme Court extends that logic to the central prohibition when those challenges reach it is genuinely unknown, and anyone claiming to know is guessing.
What operators and regulators elsewhere should be watching is not the verdict. It is whether any jurisdiction manages to write a rule that treats skill as real and harm as real at the same time. So far, none has.