Two Markets Price the GTA 6 Date, and One Must Be Wrong

Two prediction markets are quoting the Grand Theft Auto VI release date right now, and their prices cannot both be right. Polymarket puts the chance of a slip past 19 November at about 6 per cent. Kalshi puts the chance of a slip past 30 November, an easier bar to clear by eleven days, at about 8.5 per cent. A later deadline cannot be harder to hit. This desk queried both venues directly through their public APIs, twice each, ten minutes apart. Here is what they say, what the gap means, and why the number you should attach most caution to is the trading volume rather than the price. Will GTA 6 be delayed again? What Polymarket prices CONFIRMED. The Polymarket event API answered HTTP 200 at 7,413 bytes at 2026-09-24T03:02:22Z, re-read at 03:12:54Z. The market is titled "GTA 6 launch postponed again?" and its resolution text reads, verbatim: This market will resolve to "Yes" if by November 19, 2026, 11:59 PM ET, Rockstar Games or its parent company, Take-Two Interactive, officially announces that Grand Theft Auto VI will not be released by November 19, 2026, or if Grand Theft Auto VI is otherwise not released in the US by November 19, 2026, 11:59 PM ET. It adds, verbatim: "If the release is only for certain consoles (e.g. Xbox Series X/S) it will count." At the first read the book showed outcome prices of 0.055 Yes and 0.945 No, a best bid of 0.04, a best ask of 0.07 and a spread of 0.03. At the second read, ten minutes later, outcome prices were 0.06 and 0.94 with a best bid of 0.05 and the same 0.07 ask. Taking the midpoint of the executable book, that is an implied 6.0 per cent chance of a slip, and that midpoint is this desk's own arithmetic. What Kalshi prices CONFIRMED. The Kalshi markets API answered HTTP 200 at 10,681 bytes at 2026-09-24T03:02:55Z, with the 30 November contract re-read at 03:12:54Z. Six markets were open in the series, every one carrying a resolution rule of the form "If GTA 6 (Grand Theft Auto VI) has been released on a major platform (PS5, Xbox Series X/S, Windows) by date, then the market resolves to Yes." ContractReleased byYes bid / askNo bid / askLast KXGTA6-26SEP3030 Sep 20260.0000 / 0.01000.9900 / 1.00000.0100 KXGTA6-26OCT3131 Oct 20260.0100 / 0.02000.9800 / 0.99000.0200 KXGTA6-26NOV3030 Nov 20260.9100 / 0.92000.0800 / 0.09000.9200 GTA6-26DEC3131 Dec 20260.9200 / 0.93000.0700 / 0.08000.9200 KXGTA6-27JUL011 Jul 20270.9700 / 0.98000.0200 / 0.03000.9800 KXGTA6-28JAN011 Jan 20280.9700 / 0.98000.0200 / 0.03000.9800 The No side of the 30 November contract is the one to read against Polymarket. Its midpoint is 8.5 per cent, which is this desk's own arithmetic on the 0.08 bid and 0.09 ask. Why the two prices cannot both be right 30 November is eleven days after 19 November. The probability of a game not being out by 30 November can never be higher than the probability of it not being out by 19 November, because every world where it misses the later date also misses the earlier one. The two books say 8.5 per cent against 6.0 per cent, which is a 2.5 percentage point violation of that ordering. Taking the most charitable executable pair instead, Kalshi's No bid at 0.08 against Polymarket's Yes ask at 0.07, the gap is still 1.0 percentage point the wrong way. Both of those comparisons are this desk's own arithmetic and they are the finding. This is not an arbitrage and this article does not call it one. These are two venues in two regulatory and settlement regimes. Kalshi is a designated contract market settling in dollars. Polymarket settles in stablecoin on a blockchain. Nobody can cheaply hold one leg against the other, which is precisely why a gap like this can sit there. What it measures is how thin and how disconnected the two books are. The questions are also close rather than identical, and the difference is worth stating. Polymarket resolves on release in the United States and explicitly counts a single-console release. Kalshi requires release on a major platform, naming PS5, Xbox Series X/S and Windows. Close, not the same. The number that should worry you most is the volume CONFIRMED. Polymarket's own API reports lifetime volume on that market of 812,838 dollars and open interest of 252,420. Its 24 hour volume is 28.57 dollars. Its one-week volume is 6,254 dollars. Twenty-eight dollars and change, on a market with three quarters of a million dollars of lifetime turnover. Today's price on that book is being set by almost nobody. A midpoint taken from a three-cent spread on a near-dormant order book carries error bars wide enough to swallow the entire 2.5 point discrepancy above, and anybody quoting "the market says 6 per cent" without that number beside it is quoting a price with no trading behind it. Kalshi's response shape did not expose volume fields at all on this query, so no Kalshi volume is printed here. Open interest on the 30 November contract came back as 303,876 in the units the API returns, and since the payload does not label that unit, this desk is not going to assume it is dollars. What the price history shows CONFIRMED. The Polymarket daily closes endpoint answered HTTP 200 at 870 bytes at 2026-09-24T03:02:40Z for the recent window, and at 7,784 bytes with 292 points for the full history. Inside this week the Yes token closed at 0.065 on 18 September, 0.06 on 19 September, 0.055 on 20 and 21 September, and 0.06 on 22, 23 and 24 September. The full series starts on 8 November 2025 at 0.275. Its all-time high is 0.545 on 10 January 2026 and its all-time low is 0.055 on 3 September 2026. Month openings run 0.315 in December 2025, 0.335 in January, 0.415 in February, 0.225 in May, 0.175 in June, 0.105 in July, 0.11 in August and 0.065 in September. From that January peak to today is a fall of 88.99 per cent, and across this week the price moved 0.5 percentage points. Both are this desk's own arithmetic on the closes above. One more thing the Kalshi board says Read the 30 November and 31 December contracts side by side. Yes bids of 0.91 and 0.92, Yes asks of 0.92 and 0.93. The implied probability of the scenario "GTA 6 misses 30 November but ships by 31 December" is therefore somewhere between zero and one cent, call it under 1 per cent. That is this desk's own subtraction across the two books. Whatever else those traders believe, they do not believe in a short slip. In their pricing, if the November date goes, December goes with it. The October contract carries the same story from the other end: Yes at 0.01 to 0.02 means roughly 98.5 per cent confidence the game is not out by 31 October, which is just the announced date being taken seriously. What prediction markets can and cannot tell you They cannot tell you anything about Rockstar's internal schedule. Nobody trading these contracts has seen a build. What they measure is what a small number of people will risk money on, and on this particular market that number of people is very small indeed. The document that actually binds is a filing, not a book. Take-Two has put the 19 November date in front of investors, and our reading of that is at the SEC filing check. The full chronology of how this date arrived where it is sits at the delay timeline, and the standing questions readers keep asking are handled at the delay questions piece. Today is 56 days before 19 November 2026, which is this desk's own arithmetic. The honest summary of the betting is that two thin books disagree with each other by more than either of them is moving, one of them traded twenty-eight dollars yesterday, and both of them are still pricing the announced date as overwhelmingly likely to hold.