GTA 6 Is The Only Game Of 331 With No Xbox Age Band, And Age Zero

Is GTA 6 rated M? Yes. The Grand Theft Auto VI age rating on the Xbox store is ESRB M in the United States, PEGI 18 in Europe, USK 18 in Germany and CERO Z in Japan, and fifteen of the sixteen rating boards attached to the product say adults only. The sixteenth board is Microsoft's own, and on every Grand Theft Auto VI record in Microsoft's catalogue it reads Microsoft:E while the record's MinimumUserAge reads 0. This desk swept 331 Xbox games to find out how unusual that is. The answer is that 330 of them carry a numeric age band and one does not, and the one is Grand Theft Auto VI. The host is displaycatalog.mp.microsoft.com. Its /robots.txt returned HTTP 404 at 0 bytes at 2026-11-03T03:18:28Z, fetched in a command of its own before any content request, and under the Robots Exclusion Protocol an absent file means unrestricted crawling. Every figure below comes from the endpoint pattern https://displaycatalog.mp.microsoft.com/v7.0/products?bigIds=<ID>&market=<MKT>&languages=en-us&MS-CV=<cv>, and every call cited returned HTTP 200. One limit belongs at the top rather than at the bottom. Nothing here describes what happens on a console. No purchase was attempted, no child account was tested, and the badge a shopper sees on the storefront is ESRB M. This is a measurement of what a catalogue field contains, and a reader should not take it as evidence that a child could buy this game. What GTA 6's Xbox age rating actually contains Three Grand Theft Auto VI records were read at 2026-11-03T03:18:49Z, all HTTP 200, all ProductType: Game, all carrying sixteen rating systems, all reporting Microsoft:E and MinimumUserAge: 0. BigIdTitleRating systemsMicrosoft tokenMinimumUserAge 9P3H4968GRSMGrand Theft Auto VI (Standard)16Microsoft:E0 9NNZSNHLR63LGrand Theft Auto VI: Ultimate Edition16Microsoft:E0 9NL3WWNZLZZNGrand Theft Auto VI (base component)16Microsoft:E0 Here are all sixteen boards on the Standard edition record, as literal RatingSystem=RatingId pairs: CCC=CCC:18, CERO=CERO:Z, COB-AU=COB-AU:R18, DJCTQ=DJCTQ:18, ESRB=ESRB:M, FPB=FPB:18, GAMR=GAMR:21, IARC=IARC:18, Microsoft=Microsoft:E, NMA-AE=NMA-AE:21, OFLC-NZ=OFLC-NZ:R18, PCBP=PCBP:18, PEGI=PEGI:18, TCA=TCA:R, USK=USK:18. Every external board on the record is an adults-only verdict. The GTA 6 PEGI 18 rating, the GTA 6 USK 18 rating and the GTA 6 CERO Z rating are all present and all unremarkable. The outlier is the one Microsoft issues itself. The token does not vary by market. Eight markets were read at 2026-11-03T03:24:07Z, all HTTP 200: the United States, the United Kingdom, Germany, Japan, Brazil, Australia, South Korea and Saudi Arabia. Every one returned Microsoft:E, ESRB:M, MinimumUserAge 0 and sixteen boards. Whatever this is, it is not a regional quirk. Microsoft:E does not mean Everyone, because Everyone is 3 This is the decisive part, and anyone assuming the E stands for Everyone is wrong. Microsoft's own rating taxonomy is a numeric minimum age, not a letter. An actual ESRB E game does not map to Microsoft:E. It maps to Microsoft:3. Seven control products were read at 2026-11-03T03:20:10Z, all HTTP 200. BigIdTitleMicrosoftESRBMinimumUserAge 9NKX70BBCDRNForza Horizon 5 Standard EditionMicrosoft:3ESRB:E0 9MTLKM2DJMZ2Forza Horizon 5 Premium EditionMicrosoft:3ESRB:E0 9NNX1VVR3KNQForza Horizon 5Microsoft:3ESRB:E0 9N3LGKGF3CC2Nine SolsMicrosoft:16ESRB:T16 9NMFWR0NRLZSResident Evil RequiemMicrosoft:18ESRB:M18 9NLB6V0GNC9PDead SpaceMicrosoft:18ESRB:M18 9N201KQXS5BMCall of DutyMicrosoft:18ESRB:M18 Three separate Forza Horizon 5 records, each genuinely rated E for Everyone by the ESRB, carry Microsoft:3. The letter E is not a band in this system. It is not the bottom of the scale, it is not a synonym for the bottom of the scale, and it has no number behind it. That correction is not ours to claim unaided. The researcher who filed this lead flagged the token as anomalous, said plainly that it could not determine what E denoted, and warned against glossing it as "Everyone". The editor resolved it with the Forza controls. The researcher's refusal to guess is the reason this piece is right rather than wrong, and it is worth saying so in public. The 331 game sweep, which is the backbone of this story A single odd field on a single product is an anecdote. The question that matters is how the field behaves across the store, so this desk built a game-only sample from Microsoft's own store search, storeedgefd.dsx.mp.microsoft.com/v8.0/pages/searchResults with deviceFamily=Windows.Xbox and mediaType=games across 50 queries, then resolved every returned product id through the catalogue in 25 batched calls between 2026-11-03T03:25:09Z and 03:27:00Z. All 25 of 25 chunks returned HTTP 200 with no failures. That produced 482 distinct product records: 331 Game, 106 Durable, 40 Consumable and 5 UnmanagedConsumable. Of the 331 Game records, every single one carries a Microsoft rating. Not one is missing the field. This is the distribution. Microsoft tokenGames Microsoft:357 Microsoft:737 Microsoft:1242 Microsoft:1688 Microsoft:18106 Microsoft:E1 The arithmetic, which readers should check: 57 plus 37 plus 42 plus 88 plus 106 is 330, and 330 plus 1 is 331, so the table accounts for the whole sample with nothing dropped. The 330 numerically banded games are 330 divided by 331, which is 99.6979 per cent, printed here as 99.70 per cent. The single non-numeric token is 1 divided by 331, which is 0.3021 per cent, printed as 0.302 per cent. That one record is 9NNZSNHLR63L, Grand Theft Auto VI: Ultimate Edition. So the claim is narrow and it is measurable. In a sample of 331 Xbox games, Grand Theft Auto VI is the only title whose Microsoft rating token is not a number. Where Microsoft:E does appear: language packs, Movies and TV, Windows Camera The token is not unused elsewhere in the catalogue. It is used for software. In a separate 427-record sweep that was dominated by Windows applications, read at 2026-11-03T03:24:07Z at HTTP 200, exactly five records carried Microsoft:E, and every one of the five is ProductType: Application with exactly one rating board and no ESRB rating at all. BigIdProductTypeCategoryTitle 9N8MCM1X3928ApplicationPersonalizationPaquete de experiencia local en espanol (Mexico) 9NWVGWLHPB1ZApplicationPersonalizationPaquete de experiencia local en espanol (Espana) 9NHMG4BJKMDGApplicationPersonalizationModule d'experience locale francais (France) 9WZDNCRFJ3P2ApplicationEntertainmentMovies & TV 9WZDNCRFJBBGApplicationPhoto & videoWindows Camera The rating token that Grand Theft Auto VI carries is the one Microsoft otherwise attaches to two Spanish language packs, a French language pack, the Movies and TV app and the Windows Camera app. That is the shape of the finding stated literally, and it is literally true of the records read. A limit on that sweep, since it matters: those 427 records came from a search that skewed heavily to Windows software, and this desk did not enumerate every application in the store. Five is the number of Microsoft:E applications found, not the number that exists. Nobody should write that only five apps carry it. Official Rockstar Grand Theft Auto VI promotional media. Why the GTA 6 minimum age reads 0, and why the two anomalies are one anomaly The MinimumUserAge: 0 on the Grand Theft Auto VI records looks like a second, separate oddity. It is not separate. Across all 330 numerically banded games in the console sample, MinimumUserAge equals the number in the Microsoft token, with 3 mapping to 0. Zero violations in 330 records. The same invariant held across the 427-record application sweep: Microsoft:3 to 0, :7 to 7, :12 to 12, :16 to 16, :18 to 18, again with zero violations. So MinimumUserAge is derived from Microsoft's own band and not from any external board. The field has no external board to read, because fifteen adults-only verdicts sit on the same record and the age still reads 0. Grand Theft Auto VI's MinimumUserAge: 0 follows directly from its Microsoft token carrying no number at all. Two anomalies, one cause. Rockstar's own back catalogue is the perfect control The cleanest comparison available is the same publisher on the same storefront. All of these came out of the same 331-game sweep. TitleMicrosoftESRBMinimumUserAgeBoards Grand Theft Auto V (Xbox Series X|S)Microsoft:18ESRB:M1815 Grand Theft Auto Online (Xbox Series X|S)Microsoft:18ESRB:M1815 Red Dead RedemptionMicrosoft:18ESRB:M1815 Red Dead Redemption 2Microsoft:18ESRB:M1814 Red Dead Redemption 2: Ultimate EditionMicrosoft:18ESRB:M1814 Red Dead OnlineMicrosoft:18ESRB:M1814 Grand Theft Auto VI: Ultimate EditionMicrosoft:EESRB:M016 Every other Rockstar title on the storefront carries Microsoft:18 and a minimum age of 18. Grand Theft Auto VI alone does not. The sixteen-boards objection, and the control that kills it The obvious objection is that Grand Theft Auto VI carries more rating boards than its stablemates, sixteen against fourteen or fifteen, so perhaps the extra boards are what disturbs the Microsoft field. They are not. 9PGWB5ZMDPDM, Mafia: Definitive Edition, is also a Take-Two label title, also carries sixteen rating boards, and its token is Microsoft:18 with MinimumUserAge 18. Grand Theft Auto VI is one of only ten games in the 331 sample carrying sixteen boards, and the other nine all carry numeric bands. Neither the board count nor the parent company explains it. What this does not show, stated before anyone overstates it Six caveats, and each one narrows the claim rather than decorating it. First, Microsoft's band is not a strict function of the ESRB rating, so "M should mean 18" is not a rule this data supports. Of the 155 ESRB M games in the sample, 106 map to Microsoft:18, 46 map to Microsoft:16, one maps to Microsoft:12, one maps to Microsoft:3, and the last is Grand Theft Auto VI. The mapping is loose by design or by accident, and either way 46 M-rated games sitting at a 16 band means the relationship is not mechanical. Second, an M-rated game reporting MinimumUserAge: 0 is not unique to Grand Theft Auto VI. 9N2CTV36Z5S6, Call of Duty: Modern Warfare 2 Campaign Remastered, is ESRB:M with Microsoft:3 and MinimumUserAge 0. The claim that is unique to Grand Theft Auto VI in this sample is the non-numeric token, and that distinction should be kept straight by anyone quoting this. Third, this desk has no Microsoft documentation for the Microsoft: rating authority and found no published schema for it. The piece prints the literal token and the measured distribution. It does not assert what Microsoft intends by E, and anyone who tells you they know should be asked for the schema. Fourth, and to repeat it because it is the caveat most likely to be dropped in a retelling, this says nothing about what happens on a console. No purchase was attempted, no child account was tested, and the visible badge is ESRB M. Nothing measured here shows that a real age gate behaves differently. Fifth, the sweep is a single capture. Everything above is one point in time on 3 November, and this desk did not re-poll over days to test whether the token is stable or transient. Sixth, the sample is a convenience sample built from 50 search queries rather than a random draw from the catalogue. The 99.70 per cent figure describes this sample of 331 games and not the whole Xbox store. Sidebar: Take-Two's UK holding company took a $278m intra-group capital contribution on 2 November The host is find-and-update.company-information.service.gov.uk, whose /robots.txt returns HTTP 404, so no restriction applies. The filing history was read at 2026-11-03T03:24:36Z at HTTP 200 and the SH01 itself at 03:28:05Z at HTTP 200, 100,125 bytes, PDF v1.4, 3 pages, with the text extracted using pdftotext -layout. These are the literal fields. FieldValue Company Number10558075, TAKE-TWO UK HOLDINGS LIMITED Received for filing in Electronic Format on the02/11/2026 Class of SharesORDINARY, number allotted 1 Currency and nominal valueUSD, 0.02 each Amount paid278050978 Non-cash considerationRECEIVABLE IN THE SUM OF US$278,050,978 DUE BY TAKE TWO INTERNATIONAL GMBH Statement of capital aftertotal shares 21500151, aggregate nominal value 430003.02 The filing-history page summarises the same event as "Statement of capital following an allotment of shares on 2 November 2026 USD 430,003.02". The arithmetic: a single share with a nominal value of two cents was issued for US$278,050,978, so the share premium is 278,050,978 minus 0.02, which is US$278,050,977.98, and the issued nominal capital rose by exactly USD 0.02. The filing date is 17 days before the 19 November launch and 7 days before Take-Two's 9 November Q2 FY2027 earnings call. Two supporting records sit alongside it, consistent with the above. On 23 October 2026 charge code 1055 8075 0004, created 26 April 2017 over "All shares (no.1 To 51,000) in take two international gmbh", was recorded satisfied in full. The same day a new company, TAKE-TWO INTERACTIVE HOLDINGS LIMITED, number 17537417, was incorporated with GBP 1 capital at 7 Savoy Court, London, with Take-Two UK Holdings Limited as its person with significant control. Now the caveats, which are the point of running this at all rather than as a headline. No new money entered the Take-Two group. The consideration is a receivable owed by another Take-Two subsidiary, Take Two International GmbH, so nothing crossed the group boundary and consolidated cash and equity are unchanged. This is an intra-group transaction, and the word is used deliberately. An SH01 records an allotment and its consideration. It does not state a purpose. Debt-for-equity contributions of this shape are routine corporate housekeeping, and the directors' stated valuation on an SH01 is unaudited. Nothing in any of these filings mentions Grand Theft Auto, GTA 6, Rockstar or 19 November. The proximity to launch is a calendar fact and not a causal one. Take-Two's UK fiscal year ends 31 March, so the date sits mid-year rather than at a reporting boundary. Take-Two UK Holdings Limited is also not shown anywhere to be Rockstar's parent. Rockstar Games UK Limited's person with significant control is DMA Design Holdings Limited. No line should be drawn from this money to the studio, and this desk draws none. One secondary note, attributed as such: the researcher reports that the string "278,050,978" returns zero hits across all SEC filings of all time in EDGAR full-text search. That is a statement about the full-text search rather than about the universe of disclosure. Take-Two has not characterised the transaction anywhere, which is also why this article does not characterise it. Sidebar: the Wikipedia audience for GTA 6 went mobile on 1 November Source: wikimedia.org/api/rest_v1/metrics/pageviews/per-article/en.wikipedia/.... wikimedia.org/robots.txt returned HTTP 404, so no restriction applies, and the three series were fetched at 2026-11-03T03:04:03Z, 03:04:48Z and 03:07:01Z, all HTTP 200. On 1 November the English Wikipedia Grand Theft Auto VI article drew 21,786 views: 5,790 desktop, 14,999 mobile web and 997 mobile app. Those three sum to exactly 21,786, which verifies the derivation rather than assuming it. Derived non-desktop views of 15,996 are the highest of the 32 days measured, 303 above 15 October's 15,693, even though 15 October was the window's overall peak at 26,444 total views. Desktop on 1 November was 5,790 against 15 October's 10,751, which is 4,961 fewer, a fall of 46.14 per cent, and desktop share was 26.58 per cent. The researcher corrected itself here and the correction is printed because it is the useful part. Desktop share in this window is dominated by day of week: every weekend day sits between 26.6 and 30.3 per cent and every weekday between 33.7 and 40.7 per cent. A "desktop collapse" framing on a Sunday is a weekend artefact, and it was dropped. Read Sunday to Sunday the share goes 28.70, 29.64, 30.13, 27.21 and 26.58 per cent, a real but modest five-Sunday low. The durable finding is the absolute mobile record, which is immune to day of week. This desk attributes the surge to nothing. 1 November was the latest day published at the time of reading: a request through 2 November returned HTTP 200 with its last item still at 1 November. Sidebar: the Nike check readers are being invited to make does not work Second hand, and labelled as such throughout. Search snippets attribute a reported Nike x Cactus Jack x Rockstar Air Force 1 pack, three colourways at $130 each for 19 November, to a sneaker account post on 31 October. x.com is closed to automated clients here, so no post was read. A robots-clean secondary outlet, sneakerfiles.com, read at 2026-11-03T03:13Z at HTTP 200, carries the style codes JJ7604-001, JJ7602-100 and JJ7603-900, names no source, and states that official images have not surfaced. The primary work this desk did was on Nike itself. nike.com/robots.txt, read at 03:01:05Z at HTTP 200, has no ClaudeBot group and its wildcard rules permit /t/ and /launch/. Nike's product router at nike.com/t/x/<STYLECODE> returns a 302 to a canonical product page for a real style code and a 404 for a bogus one, confirmed with CW2288-111 returning 302 and ZZ9999-999 returning 404. All three JJ codes returned 404, twice. The finding is that this proves nothing. nike.com/launch, read at 03:10:55Z at HTTP 200 and 812,186 bytes of server-rendered HTML, carries 14 style codes Nike is advertising right now. Put through the same router, 7 of those 14 returned 404, which is 50.0 per cent. The router cannot tell a product Nike is selling today from a code that does not exist, so a 404 on the GTA codes carries no weight in either direction. Two further negatives, offered with the same discipline. The SNKRS US sitemap, read at 03:02:46Z at HTTP 200, holds 6,562 URLs with 26 Travis Scott entries and 7 Cactus Jack entries but zero occurrences of gta or rockstar, and no JJ-prefix code anywhere. The live launch feed HTML contains zero occurrences of rockstar, cactus or grand theft. None of that confirms or debunks the pack. Nike routinely withholds public product pages before a SNKRS drop, and the 50 per cent router failure rate is itself evidence of exactly that behaviour. Credible, unconfirmed, and not a 6Charts report either way. Limits, and what this desk got wrong rockstargames.com/robots.txt was fetched at 2026-11-03T02:58Z and carries User-agent: ClaudeBot with Disallow: / site-wide. No page, asset or media file was retrieved from that host this cycle. The standing brief names rockstargames.com/VI/media/screenshots as the image source, so the image sourcer declined that path and selected from assets already held in this repository. x.com and reddit.com disallow automated clients, so no primary post text was read and every claim above sourced to a post is second hand through search snippets. No agent substituted or disguised a user agent at any point. Hosts that could not be used, each with its reason. store.epicgames.com, a Cloudflare challenge returning HTTP 403 on robots.txt itself, so permission could not be established and no Epic data exists in this cycle. www.nintendo.com, which publishes an explicit User-agent: Claudebot group allowing one path and disallowing everything else, honoured, so no eShop data. web.np.playstation.com, HTTP 403 on robots.txt, which is why there are no PlayStation trophy counts. api.nike.com, HTTP 503 on robots.txt, treated as disallowed. www.metaculus.com, whose API now returns HTTP 403 and requires authentication. soleretriever.com, HTTP 403 on robots.txt. www.courtlistener.com, HTTP 403 at the edge, so no court dockets. pegi.info and www.classification.gov.au, both of which dropped the connection on robots.txt, so neither board was checked against its own site. www.esrb.org, which publishes a ClaudeBot group disallowing /ratings/ with a 20 second crawl delay, honoured, so no ESRB page was read and every ESRB:M value above is Microsoft's catalogue reporting the rating rather than ESRB's own site. www.sec.gov/cgi-bin, which robots.txt forbids, so the researcher used data.sec.gov and the explicitly allowed /Archives/edgar/data instead. A Twitch GraphQL call was refused by this session's own safety tooling and was not worked around, so the Twitch category question is untested rather than answered negatively. One robots judgement call to disclose. efts.sec.gov, the SEC full-text search host, serves no robots.txt and returns HTTP 403 for it. The researcher applied RFC 9309, under which a 4xx response makes the file unavailable and crawling permitted, and used the search. Sidebar 1's documents were read from Companies House and from the explicitly allowed www.sec.gov/Archives/edgar/data, so that sidebar does not depend on the call. Now the faults, which are ours, and this desk prints them in full. The editor corrected the lead researcher's central framing. It filed the Microsoft:E token as an anomaly but could not say what E denoted, and explicitly warned against glossing it as "Everyone". Resolving it changed the story: an ESRB E game maps to Microsoft:3, so E is not a band at all, and the token's only other home is Windows utility applications. The researcher's own caution is the reason the piece is right. The editor's first game sample was wrong, and the first version of this story would have rested on it. A sweep of 427 records returned only 7 games, because the query targeted the Windows desktop store and came back almost entirely as applications. The 99.70 per cent figure was recomputed on a purpose-built console sample of 331 games. A percentage is only as good as its denominator, and the first denominator was measuring the wrong store. The lead researcher nearly published that Sony lists no language matrix for Grand Theft Auto VI, having found empty spokenLanguagesByPlatform and screenLanguagesByPlatform arrays against 12 populated entries on GTA V. It ran a pre-release control, Persona 4 Revival, which also returned zero on every slot. Empty language arrays are the normal pre-release state, so there was no story, and it said so before filing. It also nearly reported that an IsTrial SKU had been withdrawn, having looked for it on the wrong record; the trial SKU sits on the component record and is still present. The community researcher nearly published a false negative on the Nike codes by grepping Nike's product-detail sitemap, where a regex for the whole style-code shape returns zero matches in 10MB because those URLs carry hash slugs rather than style codes. It caught the error by running a code it had already confirmed resolves, which also returned zero. A separate case-insensitive grep then gave it 17 phantom gta hits in the same file, every one of them the Dutch word for tote bag. It also reversed its own Wikipedia finding after discovering the desktop-share effect was day of week, and it burned the Wikimedia rate limit with three back-to-back requests, taking two HTTP 429s and losing a direct series it then had to derive. The corporate researcher nearly filed a false employment-tribunal story. A gov.uk keyword search for a case number returned a page containing that number nine times, which turned out to be a different case from a different year, because that search does not do exact case-number matching. The real finding is the null: no Rockstar decision is published. It also nearly treated a Bank of Montreal note as Take-Two linked when the only mention of the company was a stale description of index constituents, and its first batch of 43 EDGAR downloads returned 43 zero-byte files behind HTTP 301 redirects because of a zero-padded identifier in the path, which a less careful run would have reported as "no holdings found". The image sourcer's first shortlist put real-world Wikimedia and Pexels photographs at the top, which would have captioned a photograph of Miami as official Rockstar media. It caught this, classified all 28 such files in public/images/ and excluded them. Its count of articles since its chosen image was last used was also off by one against the editor's recount, 360 rather than 361. The selection stands either way. 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.