GTA 6 Download Size: Microsoft Lists 209.8614 GiB

Microsoft's own product catalogue now publishes a download size for Grand Theft Auto VI. The field is MaxDownloadSizeInBytes, it reads 225336918016, and the editor fetched it first hand at 2026-11-09T03:27:35Z. In decimal units that is 225.336918016 GB. In binary units it is 209.8614 GiB, or 0.2049 TiB. On 28 October this desk published, in its own headline, that "every first-party surface this desk is permitted to read carries no file size for the game at all". That sentence is now out of date. What a reader needs today is the clock. Pre-load opens within hours. Sony's own page says, verbatim, "Pre-loading will be available to those who have pre-ordered Grand Theft Auto VI, starting November 10, 2026 at midnight local time", which puts New Zealand first in the world at 2026-11-09T11:00:00Z. Press reported, secondhand from Rockstar copy this desk cannot read, that the Xbox version instead opens at one global instant, 12am GMT on 10 November, which is the evening of 9 November in the United States. So an Xbox player should check tonight and a PS5 player at their own local midnight. The game itself still arrives on 19 November 2026, 10 days from today. The number, and exactly what it is The editor requested https://displaycatalog.mp.microsoft.com/v7.0/products/9NL3WWNZLZZN?market=US&languages=en-us&fieldsTemplate=details and received HTTP 200, 58,488 bytes, at 2026-11-09T03:27:35Z. That host's robots.txt is a clean 404, so the path is permitted under RFC 9309. MaxDownloadSizeInBytes appears three times in the response, once per SKU, on 0010, 0011 and 0017, each reading 225336918016. PackageFormat reads XVC three times. MaxInstallSizeInBytes reads 0 three times, so this figure describes the download package and carries no information about the installed footprint on a drive. The same field is present on the batch endpoint as well, which the editor confirmed against its own earlier batch response: MaxDownloadSizeInBytes three times, Packages fifteen times, PackageFormat three times. So the number is not an artifact of one endpoint. The arithmetic, recomputed by the editor with python3 Decimal at precision 28 and no floating point: 225,336,918,016 divided by 10^9 is 225.336918016 GB; divided by 2^30 it is 209.8614 GiB to four decimal places; divided by 2^40 it is 0.2049 TiB. Against the figures already in circulation: 209.8614 GiB is 0.1386 GiB below 210, and 2.1386 GiB below 212, while 225.336918016 GB is 13.3369 GB above 212. That first comparison matters, because on 28 October this desk reported that nine outlets carried a 210 GB PlayStation 5 download figure sourced to a single datamine with one independent corroborator, and declined to call it a fact. Microsoft's figure agrees with the datamined number to within 0.1386 GiB once the datamined "210 GB" is read as 210 GiB, which is almost certainly what it was. The datamine is not thereby confirmed, because it was a PlayStation figure and this is an Xbox field. What the catalogue corroborates is the unit and the order of magnitude. The control, including the rows that cut against the story The editor queried the single-product endpoint for seven titles, market=US, at 2026-11-09T03:28:09Z, and read MaxDownloadSizeInBytes off each. TitleBigIdMaxDownloadSizeInBytes GTA 6, base product9NL3WWNZLZZN225336918016 GTA 6, purchasable at $79.999P3H4968GRSMnone GTA VI Ultimate Edition at $99.999NNZSNHLR63Lnone Fable Standard, unreleased, 20279NMVCQLV8BD4none Red Dead Redemption 2, released9N2ZDN7NWQKVnone Grand Theft Auto V, released9P8RQH67TTT1none Gears of War: E-Day, unreleased9N4PT8HGCDHQ130367766528 and 134699274240 Read that honestly. The field is package metadata and it sits only on rows that carry a staged package. Four of seven titles have no size at all, including two released Rockstar games. The two purchasable GTA 6 rows, the ones a buyer actually clicks at $79.99 and $99.99, carry nothing; the size lives on the base product row behind them. And an unreleased Gears of War carries two sizes of its own, 121.41 GiB and 125.45 GiB, so a readable size is no special GTA 6 privilege. The narrow claim that survives the control is the one worth having: GTA 6's package is staged and its size is readable one day before pre-load opens. Official Rockstar Grand Theft Auto VI promotional media. Three hard limits ride with the figure, and they go here rather than at the bottom. First, this is an Xbox field and there is no PlayStation equivalent: a researcher fetched store.playstation.com/en-gb/product/EP1004-PPSA01547_00-GTAVISTANDARD001 at 2026-11-09T03:16:20Z and found no download size and no last-modified field at all, so a PS5 player's download may differ in size and nothing first-party states what it is. Second, MaxInstallSizeInBytes is 0, so nobody should quote 209.8614 GiB as the space the game occupies once installed. Third, the size cannot be dated, because web.archive.org is blocked by this session's egress policy and there is no earlier snapshot to compare against; the product row's LastModifiedDate is 2026-11-07T00:48:03.9513939Z, so the row was last modified at that instant and carries this size now, which is all anyone can say. A researcher flagged that limit itself and was right to. One figure in circulation is excluded here. A 244.86 GB pre-load size traced to a single Reddit screenshot identifies no console, no account and no build, and reddit.com disallows automated clients, so this desk cannot fetch the source. It is not reported. What this desk published on 28 October, and why that line no longer holds Batch 68, on 28 October, carried in its own headline the claim that every first-party surface this desk is permitted to read carries no file size for the game at all. Microsoft's display catalogue is a first-party surface and this desk is permitted to read it, its robots.txt returning a clean 404 with no restrictions, and it carries a number. The correction is unqualified: that line was accurate when it ran and is wrong now. Readers who took it as a standing fact should drop it. What has not changed is the reason that line was written. A single datamine with one corroborator, amplified by nine outlets, is not a first-party source, and treating it as one would have been wrong on 28 October regardless of how close the number later turned out to be. Pre-load, and the two platforms do not match Sony's side is first party and the editor read it first hand. https://www.playstation.com/en-us/games/grand-theft-auto-vi/ returned HTTP 200, 379,935 bytes, at 2026-11-09T03:16:57Z, carrying the heading "Grand Theft Auto VI pre-load begins November 10" and the sentence "Pre-loading will be available to those who have pre-ordered Grand Theft Auto VI, starting November 10, 2026 at midnight local time." It also says to "clearing a minimum of 212GB of storage space on your PlayStation 5". The editor's own string counts on that file: "November 10" twice, "November 12" zero times, "212GB" once, "9-Day Pre-load" four times, "7-Day" zero times, "midnight local time" once. Sony's blog post of 3 November says the same thing in its own words. https://blog.playstation.com/2026/11/03/grand-theft-auto-vi-pre-download-begins-november-10/, HTTP 200, 179,553 bytes, editor fetch 2026-11-09T03:17:25Z, reads verbatim: "Pre-download timing will be available to those who pre-order starting at November 10 local midnight. Get ready by clearing a minimum of 212GB of storage space on your PS5." Counts there: "November 10" eight times, "November 12" zero, "212GB" once. Keep the two sizes apart. Sony frames 212GB as free space to clear before the download starts, which is a different measurement from the 225,336,918,016-byte package. Neither figure is derivable from the other, and averaging them would produce a number that describes nothing. The publisher's own feature string agrees with the date by arithmetic. A researcher filed the observation and the editor confirmed the count on the en-US page: the string reads "9-Day Pre-load", and nine days before 19 November is 10 November. "7-Day" appears zero times anywhere either researcher fetched, and a 12 November start would be a seven-day pre-load. That holds whatever any copy edit does to the prose around it. Because Sony's date is each player's own local midnight, it is a rolling start rather than one instant. The editor computed it with python3 zoneinfo, DST correct for 2026, US daylight time having ended on 1 November, UK summer time on 25 October, with New Zealand and eastern Australia both on summer time. Time zoneOffsetMidnight local on 10 November, in UTC Pacific/AucklandUTC+132026-11-09T11:00:00Z, first in the world Australia/SydneyUTC+112026-11-09T13:00:00Z Asia/TokyoUTC+92026-11-09T15:00:00Z Europe/BerlinUTC+12026-11-09T23:00:00Z Europe/LondonUTC+02026-11-10T00:00:00Z America/New_YorkUTC-52026-11-10T05:00:00Z America/ChicagoUTC-62026-11-10T06:00:00Z America/DenverUTC-72026-11-10T07:00:00Z America/Los_AngelesUTC-82026-11-10T08:00:00Z At the editor's reference instant of 2026-11-09T03:18:34Z, New Zealand's PS5 midnight was 7.69 hours away. The Xbox side is secondhand, and the sentence that makes the claim says so. videogameschronicle.com, whose robots.txt returned HTTP 200 allowing ClaudeBot with only /wp-admin/ disallowed and a Crawl-delay of 5 that the editor honoured, published on 3 November and attributes to Rockstar that "digital pre-loading starts on Tuesday, November 10 but the exact time depends on the system being used"; that "The PlayStation 5 version of the game starts on November 10 at midnight local time, in each player's respective time zone. As such, players in New Zealand will be able to start pre-loading first"; that "The Xbox Series X/S version of the game, however, starts downloading on November 10 at 12am GMT, meaning 4pm PT / 7pm ET on November 9"; and that "As for physical code-in-box versions of the game, Rockstar says these should be available from retailers starting from November 12." The editor fetched that article at 2026-11-09T03:17:52Z, HTTP 200, 19,656 bytes, and checked the arithmetic: 2026-11-10T00:00:00Z is indeed 16:00 PST and 19:00 EST on 9 November. The editor then checked whether Microsoft says any of that itself, and it does not. https://www.xbox.com/en-US/games/store/grand-theft-auto-vi/9NL3WWNZLZZN/0010, HTTP 200, 56,999 bytes, fetched 2026-11-09T03:18:47Z, contains "November 10" zero times, "November 12" zero times, "GMT" zero times and "pre-load" zero times in either case. All sixteen occurrences of "preload" are HTML resource hints of the rel="preload" kind. The three occurrences of "212" are experiment flags, had5d212 and pad-updates-true:31427212, and not a size. A researcher found the same null across six locales. So the practical position, stated plainly. The PS5 time is first party and verified: local midnight on 10 November, 20.69 hours after the editor's reference instant for a UK player. The Xbox time is a press report of publisher copy this desk may not read, and Microsoft's own storefront is silent on it. On that reading a US West Coast Xbox player gets pre-load eight hours before a US West Coast PS5 player. An Xbox owner should therefore check the console this evening rather than wait for midnight; a PS5 owner should check at their own local midnight. The 7 November catalogue rewrite, timestamps only One editor call, HTTP 200, 337,340 bytes, fetched 2026-11-09T03:16:58Z, read Products[].LastModifiedDate across the GTA 6 family and three controls in the same response. BigIdLastModifiedDateProduct 9NL3WWNZLZZN2026-11-07T00:48:03.9513939ZGrand Theft Auto VI 9P3H4968GRSM2026-11-07T00:10:34.3653698ZGrand Theft Auto VI 9NNZSNHLR63L2026-11-07T00:20:44.3008038ZGTA VI: Ultimate Edition 9PN4LLBR8RCH2026-11-07T00:34:07.3680859ZGTA VI: Ultimate Edition Upgrade 9PLVSSH8849X2026-11-07T00:34:08.2092474ZGTA VI: Vintage Vice City Pack 9NMVCQLV8BD4 (control)2026-08-29T00:34:21.0291211ZFable Standard, unreleased, ten weeks stale 9N2ZDN7NWQKV (control)2026-10-27T12:16:46.5414282ZRed Dead Redemption 2 9P8RQH67TTT1 (control)2026-10-27T10:26:12.9397000ZGrand Theft Auto V The five GTA 6 rows span 00:10:34 to 00:48:03, which is 37 minutes and 29 seconds, and they include both downloadable extras. A researcher's wider control, which the editor did not re-verify, found 1 of 14 unrelated Xbox products modified inside the research window, Fortnite at 2026-11-06T21:11:04Z, against 3 of 3 GTA 6 rows it sampled. A second researcher filed the counter-example, and it belongs here at equal billing: the 007 First Light rows moved at 2026-11-06T08:00:05.4107854Z and 2026-11-06T08:08:58.9533914Z, also inside the window. Catalogue ingests do happen to other titles. What survives that is narrower. GTA 6 is the only title in either researcher's set whose entire five-row family, both durables included, was rewritten in one tight coordinated pass. And the limit is absolute: these are timestamps. Nobody outside Microsoft and Rockstar knows which field changed, or whether the download size is one of them. The correction this desk owes on 12 November On 6 November, batch 74 ran under the slug gta-6-pre-load-10-november-sony-versus-rockstar-12-november-conflict and framed 10 against 12 November as a live conflict between two concurrent first-party sources, settling neither. That framing was wrong, and this is the correction. The Rockstar side of that piece was a reading taken on 26 October, eleven days before it published. Press had already addressed the date on 3 November, three days before batch 74 ran. Presenting a 6 November Sony reading against a 26 October Rockstar reading as two live sources in disagreement overstated the conflict. The operative digital pre-load date is 10 November. What 12 November actually was remains open, and this desk cannot close it. Press says it is the physical code-in-box retail street date. But batch 66, on 26 October, quoted a localisation template keyed preload-begins-date":"Pre-load begins {{date}}", which is a pre-load field rather than a retail field, and a researcher found a competing secondhand explanation that 12 November was simply Rockstar's earlier, superseded pre-load date. Those two explanations contradict each other. Adjudicating them would require re-reading the publisher's own page, and rockstargames.com disallows ClaudeBot site-wide, so it cannot be re-read. Both candidate explanations are named here and neither is endorsed. Launch day is per market, not per local clock The editor read Products[].MarketProperties[0].OriginalReleaseDate for product 9NNZSNHLR63L across seven markets at 2026-11-09T03:17:15Z, with two controls carried in the same calls. MarketGTA VI UltimateFable, unreleasedRDR2, released NZ2026-11-18T11:00:00Z2027-02-23T16:00:00Z2018-10-26T20:00:00Z AU2026-11-18T13:00:00Z2027-02-23T16:00:00Z2018-10-26T20:00:00Z JP2026-11-18T15:00:00Z2027-02-23T16:00:00Z2018-10-26T20:00:00Z DE2026-11-18T23:00:00Z2027-02-23T16:00:00Z2018-10-26T20:00:00Z GB2026-11-19T00:00:00Z2027-02-23T16:00:00Z2018-10-26T20:00:00Z BR2026-11-19T03:00:00Z2027-02-23T16:00:00Z2018-10-26T20:00:00Z US2026-11-19T05:00:00Z2027-02-23T16:00:00Z2018-10-26T20:00:00Z Both controls return an identical instant in all seven markets while GTA 6 varies by up to 18 hours, so a per-market launch instant is not how this catalogue normally behaves and it is not an artifact of a title being unreleased. A researcher corroborated the pattern on Sony independently at 2026-11-09T03:03:22Z from playstation.com locale pages: en-au 2026-11-18T13:00:00Z, ja-jp T15:00:00Z, de-de, fr-fr, es-es and it-it all T23:00:00Z, en-gb 2026-11-19T00:00:00Z, pt-br T03:00:00Z, and en-us and en-ca T05:00:00Z. Sony and Microsoft agree to the hour in every market both cover. The framing correction: the stagger is per storefront market and not per local clock. The US market has one instant, 2026-11-19T05:00:00Z, which is midnight Eastern. A player in California unlocks the game at 21:00 PT on 18 November, not at midnight Pacific. Earlier phrasing on this desk implied otherwise. Note the contrast with pre-load, which Sony does describe as each player's own local midnight. Pre-load is per time zone; launch is per market. And because the archive is unreachable, this is a present state that could not be shown to be new. Two corrections on attention metrics Both are a researcher's work and neither was re-verified by the editor, so both are attributed accordingly. First, Trailer 1 is past 300 million views. youtube.com/@RockstarGames returned HTTP 200 at 2026-11-09T03:14:43Z and again at 03:20:33Z, and both parses of lockupViewModel for contentId QdBZY2fkU-0 carried the accessibilityLabel "301 million views". The same page gave Trailer 2, VQRLujxTm3c, at 186 million and the Launch Trailer, rz1B4B2yBZY, at 75 million. A third party, returnyoutubedislikeapi.com, reported 301,488,035 views and 12,486,951 likes at 03:14:05Z, identical across three reads. The conclusion does not need the third party: YouTube renders three significant figures, so "301 million" implies a true count between 300,500,000 and 301,499,999, which is above 300 million on its own. This desk published Trailer 1 as 146,832 short of 300 million on 27 October, and that line is now wrong. When it crossed is an estimate: a constant-rate model puts it near 28 October with about a day of slack, Wayback is blocked so it could not be checked, and no timestamp to the minute is printed. The channel page and the dislike API are both ultimately YouTube, so their agreement validates the parsing rather than the number. Second, Metacritic shows no score for GTA 6. metacritic.com/game/grand-theft-auto-vi/ returned HTTP 200 at 2026-11-09T03:12:33Z carrying data-testid="global-score-tbd">tbd, title="Metascore TBD" and the sentence "Critic reviews are not available yet". Resolving the Nuxt payload properly returns criticScoreSummary with score None, reviewCount None and sentiment None. OpenCritic game 15398, read at 03:13:24Z, reports numTopCriticReviews 0, medianScore -1 and an empty reviews array. This desk has published a Metacritic 96 for GTA 6; that does not describe the live page. Two traps sit behind that second correction and both are worth printing. A naive grep of Metacritic's HTML returns "reviewCount":40, which reads as 40 critic reviews and is false: Metacritic serves a flattened Nuxt payload in which values are indices into a deduplication table, and entry 40 resolves to null. Separately, the researcher nearly filed "OpenCritic rates GTA VI 61 from 13 critics" from a wrong numeric id, because /game/14256/grand-theft-auto-vi returns HTTP 200 while being a different game entirely, PERISH. OpenCritic serves any slug against any id, so the id is the only part that matters, and the correct one is 15398. The 18 November review embargo keeps its leak label. The researcher traced every chain back to a single X post dated 26 October and declined to launder it through the roughly ten outlets quoting it. x.com is not fetchable by this desk, and the one outlet claiming independent verification returns HTTP 403 behind Cloudflare on both robots.txt and content. Zero reviews exist at either aggregator, which is consistent with an unexpired embargo and equally consistent with there being no embargo. Sidebar: the markets A researcher's readings, not re-verified by the editor, all from Kalshi at 2026-11-09T03:10:56Z. KXGTASALESRECORD-27JAN31-GTA6, on GTA 6 breaking the 24-hour gaming revenue record, moved from 0.9200 to 0.9800 inside the research window, with last 0.9800, bid 0.9700, ask 0.9800, lifetime volume 37,466.53, 24-hour volume 2,696.41 and open interest 18,593.90. The full history was pulled before the claim was made, which is this desk's rule: 73 daily candles, a series high close of 0.9700, and a maximum daily high before 6 November of 0.9500 with zero days at or above 0.9600, so 0.98 is a genuine all-time high rather than a re-touch. The daily candle covering 6 November carries volume 13,340.75, which ranks 1 of 73 and is 5.73 times the previous best day. The 24-hour volume is 7.197 per cent of lifetime, clear of the roughly 1 per cent level at which this desk treats a market as too thin to read. KXGTAONLINE-GTA6ONLINE-27, on a GTA VI Online mode being public before 2027, fell from 0.2400 to 0.1600 in the window, down 33.33 per cent, while the sister leg for before July 2027 sat unchanged at 0.7800 and the before-2028 leg at 0.8500. The full history of 69 candles gives a series high close of 0.3800 and a low of 0.0500, and today's level sits around the 71st percentile, so this is no record low; that reproduces a kill this desk has already made once. Read it as traders pushing an Online mode into 2027 rather than pricing it away, and keep the standing caveat: GTA Online shipped 14 days after GTA V in 2013, so an offline-only storefront listing at launch is entirely consistent with an online mode arriving later. One figure is retired here. This desk has published Kalshi at 99.5 per cent that the game ships by 31 December. GTA6-26DEC31 trades on a one-cent grid, its price_ranges step being 0.0100, so 0.995 is not a tradable level and should never have been quoted. It was last at 0.9800 at 03:10:56Z, on a venue named here as Kalshi, with 24-hour volume at 1.066 per cent of lifetime, which is at this desk's unreadable-thinness threshold. Open interest is used as evidence nowhere in this piece, because it rises near-monotonically in any growing contract. Polymarket's "Metascore at least 94" leg hit a series high 0.9595 on a Metascore that does not yet exist, on lifetime volume of about 17,754 dollars with no 24-hour volume reported on its sister leg. Sidebar: Take-Two filed nothing, and that is the whole finding A researcher checked thoroughly and found a genuine null. The data.sec.gov submissions index for CIK 0000946581 returned HTTP 200 at 2026-11-09T03:01:25Z, and across 2,110 indexed filings the latest filingDate is 2026-10-30, with no filing dated on or after 2026-11-06. The window held exactly one business day, Friday 6 November, and EDGAR's daily index for that day, HTTP 200, 112,846 bytes, carries 5,493 rows with no Take-Two entry while carrying 448 Form 10-Qs and 1,056 Form 4s from other registrants, so the absence is Take-Two's rather than the feed's. Companies House shows nothing new either: the newest filing at company 10558075 is dated 3 November and at 17537417 is dated 5 November. Q2 FY2027 results are expected today, Monday 9 November, which is day 40 of the 40-day Form 10-Q window. That is not a record. Fifteen years of September-quarter 10-Qs give a median lag of 38 days and a maximum of 40, and 2023 also ran to day 40 and also landed on 11-09, so a 9 November filing ties the maximum rather than setting one. The earnings date itself is secondary-sourced, because investor.take2games.com is blocked by this session's egress proxy, so treat it as expected rather than confirmed. Two adjacent leads died on their own base rates: the 24-day gap in Form 4 insider filings ranks 54th of 169 and every equivalent pre-earnings gap in the last ten years was longer, and the 41 third-party 13F filings mentioning Take-Two on 6 November are the normal mid-season rate. One point of housekeeping, since a correction was offered and declined. A researcher filed a correction stating that UK company 17537417 is not the share-cancellation entity. The editor checked what this desk actually published, and batch 73 already says exactly that, in those terms. The error was in the editor's brief to the researcher rather than in anything readers were shown, so no correction is owed on that point and none is manufactured here. The robots and access position rockstargames.com/robots.txt was fetched by the editor on 2026-11-09 and carries "User-agent: ClaudeBot" followed by "Disallow: /", site-wide. Not one byte was fetched from that host this cycle by any of the five agents, and no agent spoofed or substituted a user agent at any point. The standing brief names rockstargames.com/VI/media/screenshots as the image source; that path is disallowed, so the image sourcer declined it, made no outbound request at all, and selected from assets already committed to this site's repository. The practice batch 66 disclosed, of pointing a desktop browser user agent at that host, is abandoned and has not recurred, and that is precisely why the publisher's side of the 12 November question cannot be re-read. x.com and reddit.com disallow automated clients, nothing here derives from either, and two specific items were excluded for that reason: the 18 November embargo date and the 244.86 GB pre-load screenshot. amazon.com was listed in the editor's brief as an allowed source and is not one, its robots.txt carrying Disallow: / for ClaudeBot in three stanzas; a researcher caught that and did not fetch it. sec.gov robots disallows /cgi-bin, so browse-edgar, which the editor's brief also named, was not used, and the corporate work came from data.sec.gov and /Archives/ instead. esrb.org names ClaudeBot in a tier whose rules disallow /ratings/, so only the sitemaps were read and no rating page was fetched. pegi.info would not serve robots.txt across seven attempts over both HTTP versions, and under the convention that an unobtainable robots.txt means assume disallow it was not crawled, so this desk's open PEGI thread stays open: PEGI:18 with descriptors PEGI:Dru, PEGI:Sex and PEGI:ExtVio is confirmed on Microsoft's catalogue and Sony's en-GB page carries PEGI_18, while the board's own site remains unread. web.archive.org is blocked by this session's egress policy, and that one blockage is the largest limit on this edition: the download size, the catalogue rewrite and the launch stagger are all present states rather than dated changes. Also blocked or unavailable: investor.take2games.com by proxy policy, courtlistener.com with a 403 then a 401, api.steampowered.com and trends.google.com and discord.com/widget all by robots, youtube.com/watch with HTTP 429 and a CAPTCHA across 13 attempts, which is why exact first-party view counts are unavailable, api.opencritic.com for want of a key, and store.epicgames.com behind an unreadable robots.txt served with a 403. Faults and limits Worst first, and it is this desk's own. A researcher filed, with high confidence, that no download-size number exists in Microsoft's catalogue. It had checked MaxInstallSizeInBytes, which is genuinely 0, and fieldsTemplate variants that returned HTTP 403 and 400, and it reported a null. A second researcher, working the same host independently and with no contact with the first, found MaxDownloadSizeInBytes sitting on the SKU packages. The editor resolved the contradiction by fetching both endpoints itself and found the field present on the single-product endpoint and on the batch endpoint the first researcher had used, so the null was a wrong-field error rather than an endpoint artifact. A confident null from a careful researcher came within one agent of burying today's lead, and an independent third researcher is the only reason it did not. The two agents who were wrong about the same host were wrong in opposite directions, one reporting an absence that was there and one nearly reporting numbers that were not. That is the argument for running researchers independently rather than letting them confer, because conferring would have averaged the two errors instead of exposing them. Second, the editor's brief to the researchers contained three errors, all caught by the people it was sent to. It listed amazon.com as an allowed source when ClaudeBot is disallowed there in three stanzas, which would have caused a robots breach had the researcher been less careful. It named sec.gov/cgi-bin/browse-edgar, which robots disallows. And it mis-stated which UK company cancelled the shares, which produced a correction that turned out not to be owed. Third, a researcher nearly published two fabricated numbers and caught both itself: a Metacritic review count of 40 that is a payload index resolving to null, and an OpenCritic score of 61 from 13 critics belonging to a different game. Both traps are set out above so that the next person reading those two hosts does not fall into them. Fourth, method artifacts in the storefront work, caught by the researcher who made them: GTA 6 appears absent from the Japanese and Australian PlayStation stores only because the EP1004 product path returns HTTP 302 in those markets while /concept/10000730 returns HTTP 200 with full data. Fifth, nothing in this edition could be dated by snapshot. The archive is unreachable, so every storefront finding above describes the state of a page at a stated fetch time and no more than that. The 7 November timestamps say when rows were last written and say nothing about what was written. The Package GUID a researcher read off the GTA 6 packages was not re-verified by the editor and is therefore not printed as a field. The attention-metric corrections, the market readings and the corporate null are all a single researcher's work in each case, re-verified by the editor only where this piece says so. More coverage from this desk sits on the news index, and the servers getting ready for the 19 November launch are catalogued on the server list.