Kalshi and Polymarket Now Agree on the GTA 6 Date

On the identical question, will Grand Theft Auto VI be released by 30 November 2026, the two large prediction venues now agree closely. Kalshi's No side is 0.0800 bid against 0.0900 ask and Polymarket's No is 0.075. That is a Kalshi midpoint of 8.5 per cent against Polymarket's 7.5 per cent, a gap of 1.0 percentage point, which is this desk's own arithmetic. This article exists because this site got the comparison wrong yesterday. The correction is set out in full below, along with the matched figures that are now quotable and the reasons neither number should be read as a forecast. The correction On the 24 September cycle this site published gta-6-delay-odds-polymarket-6-percent-kalshi-8-5-percent-impossible-pair. It set Polymarket's odds of a slip past 19 November against Kalshi's odds of a slip past 30 November and called the 2.5 point difference impossible in that direction. Those were two different deadlines. A market on a slip past 19 November and a market on a slip past 30 November are not the same question, and their prices are not required to line up in any particular order. The earlier piece's arithmetic was right about the numbers it used, and it was wrong to treat those numbers as comparable. The word "impossible" in that headline described a mismatch of this desk's own making. The fix is a matched pair, and one is now available on the same deadline at both venues. That is what the rest of this article reports. Kalshi, the 30 November market CONFIRMED. This desk's editor fetched https://api.elections.kalshi.com/trade-api/v2/markets/KXGTA6-26NOV30 and received HTTP 200 at 1,793 bytes at 2026-09-25T03:18:19Z. Fields as served: FieldValue titleWIll GTA6 be released by Nov 30, 2026? yes_bid_dollars0.9100 yes_ask_dollars0.9200 no_bid_dollars0.0800 no_ask_dollars0.0900 last_price_dollars0.9100 volume_fp753,611.29 volume_24h_fp4,958.56 open_interest_fp306,321.18 updated_time2026-09-24T22:45:39.456992Z liquidity_dollars0.0000 statusactive The capitalisation error in "WIll" is Kalshi's own and is quoted here as found rather than tidied up. The market's resolution rule, verbatim: If GTA 6 (Grand Theft Auto VI) has been released on a major platform (PS5, Xbox Series X/S, Windows) by Nov 30, 2026, then the market resolves to Yes. Two further markets in the same series, from the same fetch round. KXGTA6-26OCT31, "WIll GTA6 be released by Oct 31, 2026?", yes_bid 0.0100, yes_ask 0.0300, last 0.0300, volume_fp 828,690.87, open interest 312,163.49. And KXGTA6-27JUL01, "WIll GTA6 be released by Jul 1, 2027?", yes_bid 0.9700, yes_ask 0.9800, last 0.9700, volume_fp 127,126.69, volume_24h_fp 74.00, open interest 44,825.61. Polymarket, the same deadline CONFIRMED. This desk's editor fetched https://gamma-api.polymarket.com/events?slug=gta-vi-released-before-november-2026 and received HTTP 200 at 12,551 bytes at 2026-09-25T03:18:03Z. The event is titled "GTA VI released by...?" and carries three markets: MarketOutcome prices, Yes then No24 hour volumeLifetime volume GTA VI released by November 30?0.925 and 0.0751,138.7947,987.69 GTA VI released by October 31?0.0085 and 0.9915861.6087,562.22 GTA VI released by September 30?0.0005 and 0.99951.0133,664.14 The matched comparison QuestionKalshiPolymarketGap, this desk's own arithmetic Released by 30 November 2026, No side0.0800 bid, 0.0900 ask, midpoint 8.5 per cent0.075, or 7.5 per cent1.0 percentage point Released by 31 October 2026, Yes side0.0100 bid, 0.0300 ask, midpoint 2.0 per cent0.0085, or 0.85 per cent1.15 percentage points A 1.0 percentage point difference between two venues on a market whose Kalshi side is quoted with a one cent bid-ask spread is an ordinary spread between two order books. It is not a contradiction and this desk is not calling it one. On the earlier release question, both books put a pre-November release at low single digits, Kalshi at a 2.0 per cent midpoint and Polymarket at 0.85 per cent, both of which are this desk's own arithmetic off the figures above. Where the priced delay risk sits This desk's own arithmetic, from Kalshi's own two markets. The 30 November No midpoint is 8.5 per cent. The 1 July 2027 No midpoint, derived from yes_bid 0.9700 and yes_ask 0.9800, is 2.5 per cent. So roughly 29 per cent of the delay risk priced at the 30 November bar sits beyond mid-2027 rather than in the seven months in between. Read plainly, that says the market's own structure treats a short slip into December or early 2027 as the more likely failure mode, with a smaller tail for a long delay. That is a description of prices and not a prediction by this desk. Limits, and what these numbers are not Kalshi's own liquidity_dollars field on the 30 November market reads 0.0000. Its 24 hour volume is 4,958.56 against a lifetime 753,611.29, so today's quoted price rests on thin recent trading against a very large historical book. Polymarket's 30 November market turned 1,138.79 in 24 hours. Neither is a deep market today. This desk is not calling the venue gap an arbitrage. The two venues settle under different rules, in different regulatory regimes, with different fee and custody arrangements, and their resolution wording is close rather than identical. A one point difference across that boundary is not a free trade and nothing above should be read as suggesting one. Kalshi's expected_expiration_time field reads 2026-11-19T11:00:00Z. That is the venue's own expectation about when its own market resolves. It is not a Rockstar statement about a release time and should never be cited as one. Prediction market prices are prices. They aggregate the opinions of whoever is willing to post an order, and they have no access to Rockstar's schedule. This desk quotes them as a measure of sentiment with money attached and not as information about the game. With 55 days to 19 November 2026, which is this desk's own arithmetic, the official position remains the announced date and nothing on either venue changes that. Finally, the Polymarket figures are outcome prices from one API response and the Kalshi figures are bid, ask and last from another. The two fetches are sixteen seconds apart on this desk's own arithmetic from the timestamps above, which is as close to simultaneous as this desk could manage, but they remain two separate samples and the gap above carries that imprecision.