XBOXGamePreservation.com is a community voting site launched in February 2026. It tracks which Original XBOX and XBOX 360 titles players most want added to the backward compatibility program on current hardware. Registered users browse a catalog of 2,677 games and vote for the ones they want brought forward. Each account can vote on a game once, with a limit of 10 votes per day, and votes can be cast from anywhere on the site: search results, the leaderboard, the trending list, or a game's own page.
This report draws on two sources. The first is the complete vote table: 770,266 votes cast by 77,368 accounts between February and June 2026. The second is aggregate site analytics covering page-level traffic and country-level visitors. A vote here is a registered, rate-limited expression of demand, not a purchase commitment, and the report treats it accordingly.
Observations vs. interpretations: this report covers the complete voting record, not a sample, so findings are stated as direct counts rather than estimates, and no confidence intervals apply. Where the report interprets behavior, those readings are presented as evidence-based explanations of the data, not directly observed fact.
At a glance
Backward compatibility demand is concentrated but highly addressable. Skylanders generated the largest overall demand, while Transformers is the strongest source of baseline demand once acquisition spikes are set aside. Because audiences overlap heavily, a 25-game wave selected for breadth would reach 85.9% of all voters. That demand spans 190 countries and is made up of several distinct communities rather than one homogeneous audience.
Demand ranking
The raw vote totals. Skylanders dominates, by a factor of 3x over the next franchise. But raw demand alone is misleading: a franchise can be popular for many reasons, and high vote counts don't tell you whether you're serving one community or many.
Vote surges over time
Demand didn't accumulate evenly, and the surges come in two distinct types. Platform-wide spikes: the week of May 4, twelve of the top twenty franchises peaked simultaneously, the signature of the site itself going viral. Franchise-specific surges: individual communities mobilizing independently, Transformers in March and Skylanders in May, whose late-May surge at 200x its median week dwarfs everything in the dataset.
Audience overlap
How many voters cast ballots for both franchises? Raw overlap is intuitive but biased toward large franchises: a franchise with 30,000 voters will naturally share more voters with anything else, regardless of genuine affinity.
Affinity lift
Raw overlap counts how many voters two franchises share. Affinity lift asks whether that overlap is larger than the franchises' sizes would predict by chance. A score of 1.0x means no relationship; above 2.0x indicates a genuine community cluster. Red cells fall below 1.0x (less overlap than chance); green cells show clusters.
Jaccard similarity
Jaccard similarity normalizes audience overlap by the union of both franchises' voters. It answers: of everyone who voted for either franchise, what share voted for both? This is the most size-neutral measure of genuine audience cohesion.
Audience archetypes
The analyses together reveal that franchises cluster into distinct archetypes, and knowing which archetype a franchise belongs to changes how you'd think about serving it.
A global audience
Visitors from 190 countries. Traffic data is a proxy, reflecting who finds the site rather than who votes, but the geographic spread reframes what this demand represents. Europe collectively (27%, including the UK) is the second-largest region after the US, and Latin America (12.5%) is a substantial third, ahead of any single European country. For a platform holder weighing backward compatibility, this is evidence that demand for these games is a global market signal, not a regional niche.
Traffic data only; does not reflect voter distribution. February–June 2026.
Traffic & engagement
302K unique visitors since launch in February 2026, roughly four months. Of those, 77K became voters, a visitor-to-voter rate of 25.6%. Users can vote from anywhere on the site without ever opening a game's own page, which makes the game-page traffic figures a measure of how people first arrived, not of how much they cared.
Where traffic lands
How visitors are distributed across the site's main sections, each shown as a share of the busiest page. These are independent destinations, not steps in a funnel: a visitor can land anywhere and vote from there.
Each bar is that section's unique visitors as a percentage of homepage visitors. The figures overlap (one person may visit several sections) and do not sum to 100%.
What page traffic does and doesn't measure
Because a vote can be cast from anywhere on the site, a game's page-visit count is a discovery signal, not a demand signal. It indicates how often that page was someone's way into the site, which usually means its link was shared somewhere, rather than how much the game was wanted. The vote total already measures demand. The two can diverge sharply.
The practical implication: the leaderboard and search, not individual game pages, are where most voting happens. Page-level traffic is a useful map of which titles draw people into the site, and little more.
Voter behavior: most people vote once and leave
How do 77,368 people actually use a demand-signal platform? The mechanics matter: each user may vote on a game only once, with a cap of 10 votes per day. Within those rules, the dominant pattern is a single visit: two-thirds of multi-vote users cast everything inside one 10-minute window, and the median of 7 votes sits just under the daily cap. But the 17% who come back are a genuine power-user tier. Because a game can only be voted on once, supporting many games means returning across many days: 475 users have backed 100+ different games, 74 have backed 300+, and 14 have backed 500+, each one a distinct title they wanted brought forward.
Distribution of votes per user. The 6–10 cluster sits right at the 10-per-day cap.
Power laws at both ends
Concentration shows up on both sides of the platform. A small core of voters supplies an outsized share of votes, and a small set of games absorbs an outsized share of demand. 463 of 2,677 games have fewer than ten votes each.
The single-franchise voter
35% of all voters, 27,390 people, voted for exactly one franchise and nothing else. More than half of them belong to a single fandom: Skylanders. This is the sharpest evidence for the isolated-community archetype.
The publisher stakes
Backward compatibility is partly a rights question, and the demand data shows where ownership is concentrated, though feasibility depends on factors beyond this dataset. Titles published by Activision account for 39.4% of all demand, and since Vivendi Games merged with Activision in 2008 (bringing Sierra with it), the Vivendi/Sierra catalog's 5.0% effectively sits under the same roof, pushing the Activision Blizzard total to roughly 44%. With Microsoft's acquisition complete, that share nominally sits in-house. The catch: much of it is licensed IP. Spider-Man, X-Men, and Marvel titles need Disney's sign-off, and Transformers needs Hasbro's.
One block stands out. Skylanders, at 163K votes and 21% of all demand, is wholly Activision-owned IP, which makes it the largest block of demand with the fewest external rights-holders. That said, ownership is only one input to feasibility. The Skylanders titles relied on the Portal of Power peripheral, whose support within the backward compatibility program is an open question, and even wholly-owned games can carry licensed music, engine middleware, or other third-party components that need clearing. Fewer external rights-holders than the licensed catalog, but rarely zero.
Publisher attribution reflects each game's original publisher; many of these catalogs have since changed hands.
The delisted cohort
In mid-May, delisted games were added to the voting catalog: titles that are already backward compatible on XBOX but are no longer available for digital sale, or were never made available digitally. With XBOX embracing a digital-first lineup through the XBOX Series S and the all-digital Series X, these games are increasingly unobtainable for fans. Voting has been open for under a month, so totals aren't comparable to the rest of the catalog, but the early signal is strong.
Non-franchise games: demand with no machine behind it
Several hundred games in the catalog belong to no franchise: single releases with no sequel, no toy line, no fandom infrastructure behind them. They collected 40,127 votes (5.2% of the total) from 13,607 distinct users. These votes can't ride a slate. Nobody voted for Alpha Protocol because they were already voting for its sequels; there are none. Each of these votes is a deliberate, individual choice.
The leaders: Alpha Protocol (1,221 votes), Gun (1,166), Wet (958), Eternal Sonata (937), Brute Force (880), and The Saboteur (864).
Accounts that voted for at least one non-franchise game voted for more games overall than the typical account (a median of 10 versus 7), suggesting these titles are found by users who browse deeper into the catalog.
A prioritization framework
Vote totals are one input, but they conflate three things a decision-maker needs separated: how big the demand is, how concentrated it was in time, and how its audience is structured. The analyses below pull those apart, and they change the answer to "what should ship first."
How dependent is each franchise on its campaign?
The campaign dependency ratio: what share of a franchise's voters touched nothing else on the platform? High values reflect an audience concentrated within a single fandom; low values indicate support drawn from the platform's broader cross-catalog base. Half of Skylanders' voter base exists only for Skylanders. X-Men, The Simpsons, and Halo draw overwhelmingly from the broader audience.
Dependency describes the shape of an audience, not the value of its demand. A vote from a single-franchise Skylanders fan counts exactly as much as a vote from a broad cross-catalog voter; both are real people who want a game back. What the ratio adds is context for how that demand was reached and who it represents, which matters for planning a release, not for judging whether the demand is legitimate.
Baseline demand vs. event-driven demand
After removing each franchise's two largest acquisition weeks, the top of the leaderboard reorders: Transformers becomes the platform's largest source of baseline demand, while Skylanders remains the largest source of total demand. Skylanders' support is 82% concentrated in its two biggest weeks; every other top franchise sits between 35% and 54%.
| Franchise | Raw | Rank | Excl. top 2 wks | Adj. | Surge |
|---|---|---|---|---|---|
| Transformers | 54,867 | 2 | 32,111 | 1 | 41% |
| Skylanders | 163,396 | 1 | 29,254 | 2 | 82% |
| Spider-Man | 37,047 | 3 | 20,078 | 3 | 46% |
| Sonic The Hedgehog | 34,143 | 4 | 19,365 | 4 | 43% |
| Need for Speed | 31,563 | 5 | 14,487 | 5 | 54% |
| The Simpsons | 18,620 | 6 | 8,756 | 6 | 53% |
| Crash Bandicoot | 12,487 | 8 | 7,453 | 7 | 40% |
| X-Men | 14,612 | 7 | 7,160 | 8 | 51% |
The optimized wave: wide beats deep
If XBOX shipped a single 25-game backward compatibility wave, licensing and rights set aside, which slate would serve the most distinct fans? Because voters vote the slate, one game per franchise satisfies most of its fandom. A simple selection method makes this concrete: repeatedly pick the game that adds the most voters not yet served by an earlier pick (the standard "greedy" approach), limited to one title per franchise. The optimized wave reaches 85.9% of all 77,368 voters. The naive top-25-by-votes list, which spends six slots on Skylanders alone, reaches 79%. Notably, allowing repeat franchises would lift coverage only to 86.0%: once a franchise's flagship is in the slate, a second entry from it adds almost no one new, which is the clearest possible evidence that breadth beats depth.
| # | Game | Votes | New voters | Cumulative |
|---|---|---|---|---|
| 1 | Skylanders: Spyro's Adventure | 29,002 | 29,002 | 37.5% |
| 2 | Transformers: War for Cybertron | 12,864 | 10,089 | 50.5% |
| 3 | The Simpsons: Hit & Run | 11,517 | 6,358 | 58.7% |
| 4 | Need for Speed: Most Wanted (2005) | 8,580 | 3,914 | 63.8% |
| 5 | Sonic Heroes | 8,264 | 3,112 | 67.8% |
| 6 | Jet Set Radio Future | 6,782 | 2,188 | 70.7% |
| 7 | Spider-Man: Shattered Dimensions | 8,756 | 1,615 | 72.7% |
| 8 | Def Jam: Fight for NY | 4,527 | 1,460 | 74.6% |
| 9 | Army of Two: The 40th Day | 2,718 | 919 | 75.8% |
| 10 | Forza Motorsport 4 | 2,747 | 875 | 76.9% |
| 11 | Deadpool | 8,050 | 723 | 77.9% |
| 12 | Crash Twinsanity | 3,118 | 653 | 78.7% |
| 13 | Burnout 3: Takedown | 3,663 | 599 | 79.5% |
| 14 | Minecraft: XBOX 360 Edition | 3,224 | 554 | 80.2% |
| 15 | Dead or Alive Xtreme 2 | 1,613 | 532 | 80.9% |
| 16 | MechAssault | 1,340 | 494 | 81.5% |
| 17 | Blur | 2,883 | 482 | 82.2% |
| 18 | Buffy the Vampire Slayer | 916 | 411 | 82.7% |
| 19 | Guitar Hero III: Legends of Rock | 5,972 | 388 | 83.2% |
| 20 | Marvel: Ultimate Alliance | 3,212 | 376 | 83.7% |
| 21 | Silent Hill 4: The Room | 2,195 | 369 | 84.2% |
| 22 | 007: Quantum of Solace | 1,552 | 355 | 84.6% |
| 23 | WWE SmackDown vs. Raw 2007 | 1,480 | 344 | 85.1% |
| 24 | Halo 2 | 2,949 | 330 | 85.5% |
| 25 | Armored Core 4 | 1,247 | 294 | 85.9% |
Selection method: at each step, choose the game that adds the most voters not already served, limited to one title per franchise. Removing the limit changes coverage by about 0.1%. Rights complexity is not modeled; several entries carry licensing constraints that would shape a real slate.
Coverage economics
Demand is concentrated enough that modest waves go far: 50 games cover half of all expressed demand, and 250 cover three-quarters. The split by console generation is 68% XBOX 360 to 32% Original XBOX, but per-game demand is nearly identical, so generation share mostly reflects catalog size.
Cumulative share of all votes captured by the top N games. Demand is spread widely: it takes 50 games to reach half of all votes, and 250 to reach three-quarters.
Putting it together, prioritization comes down to four questions: How many people want it? (volume). Was the demand spread across the period or concentrated in a few large events? (event-driven vs. distributed). Does it come from one mobilized fandom or the broad audience? (structure). And how hard is it to ship? (rights). At the platform level, the optimized wave shows that breadth beats depth: once a franchise's flagship ships, additional entries reach almost no one new.
How the metrics were calculated
Each measure is computed directly from the complete voting record. None involves sampling or estimation.
Affinity lift. For two franchises, divide the number of voters who supported both by the number expected if voting were independent (voters of A × voters of B ÷ total voters). 1.0 means the overlap is exactly what sizes predict; above 1.0 means more than chance, below means less.
Jaccard similarity. For two franchises, divide the voters who supported both by the voters who supported either. Runs from 0 (no shared voters) to 1 (identical audiences). Unlike affinity lift, it is not adjusted for chance.
Campaign dependency ratio. The share of a franchise's voters who supported no other franchise. High means an audience that arrived for that franchise alone; low means one drawn from the broader population.
Baseline vs. event-driven demand. Remove each franchise's two highest-vote weeks and re-rank the remainder. Franchises holding rank drew support across the period; those falling sharply drew it from short events. "Surge share" is the percentage of votes from those two weeks.
Optimized coverage selection. To find a 25-game slate reaching the most distinct voters, proceed one game at a time, each step adding the game bringing in the most voters not already covered. This is the standard greedy approach to coverage problems. Finding the provably optimal slate is computationally infeasible at this scale, but greedy is the established approximation and, on this data, the result is stable.
Demand is many stories in one number
A vote count looks like a single number. This report has shown it compresses at least four different stories. Volume is not the same as steady accumulation: remove each franchise's two largest acquisition weeks and the leaderboard reorders, with Transformers becoming the largest source of baseline demand. Volume is not structure: half of the largest franchise's voters exist only for that franchise, while X-Men and The Simpsons draw almost entirely from the broad audience. Demand is not navigation: voters vote the slate from the leaderboard. And demand is not feasibility: the largest blocks of demand run through layered rights, expired licenses, and peripheral hardware.
Most importantly, the data suggests backward compatibility demand is not one audience but a collection of distinct communities, each with its own structure, acquisition pattern, and degree of overlap with the broader ecosystem. The practical synthesis is the optimized wave. Because fans vote in slates, one well-chosen game per franchise serves most of its supporters, and 25 titles selected for breadth, one per franchise, would reach 86% of every person who has ever voted on the platform. The list of what players want is long. The list of what it would take to satisfy most of them is surprisingly short.
Data: XBOXGamePreservation.com · 770,266 votes · 77,368 unique voters · 361K site visitors across 190 countries · February–June 2026