About RoVitals

What this is

RoVitals tracks live player counts and statistics for 44,560 Roblox experiences — concurrent players (CCU), 24-hour and observed peaks, visits, favorites, ratings, historical charts and estimated developer earnings.

Data sources and coverage

We collect each experience's public player count, visits and other counters from Roblox's game-details API. Game names, descriptions, icons and creator details are supplied by Roblox and game creators. Historical charts, observed peaks, averages and growth calculations are produced by RoVitals from the snapshots we collect. We do not have access to creators' private analytics or per-game revenue.

New games are discovered automatically from the Roblox charts across many countries and devices, keyword search sweeps, and creators' full catalogs. Opening a game page we don't know yet (/game/<place id> with any id from a roblox.com URL) can add it to tracking after Roblox resolves the experience, subject to request limits. Our catalogue is a discovered subset of Roblox, not every experience on the platform. Ranks and platform totals refer to this tracked catalogue; its coverage changes as games enter or leave it. The player count history adds up each tracked experience's sampled average players for every hour. It is an approximate tracked total with changing coverage, not a simultaneous platform-wide census; new games can also increase it as the catalogue grows.

Collection times and missing history

The normal collection schedule depends on the latest observed player count: at least 1,000 players, every 5 minutes; 50–999, every 10 minutes; 1–49, every 30 minutes; zero, every 6 hours. These are targets, not guarantees: API failures, unavailable games and tracker delays can leave gaps. Collection timestamps identify the source snapshot, not the time you opened the page. A snapshot is marked stale after two normal collection intervals without a new reading.

History begins at the tracking-start date shown on each game page. An observed peak is the largest sample we have collected since that date, not a verified lifetime record. A higher peak before tracking or between collections can be missed. Where a primary source published a higher peak from before tracking began, the game page cites it separately with its date; it is never counted as an observed peak. Different games enter tracking on different dates, so observed-peak rankings have unequal historical coverage. Raw snapshots remain available for 30 days; hourly and daily summaries are retained for older history. The homepage's raw-sample count is current retained storage, not a lifetime total.

Games are sampled at different moments. Adding their latest counts gives an approximate tracked-catalogue total, not a simultaneous platform census. A concurrent player count measures players online at a moment; visits count entries and repeat visits. Neither is a count of unique daily or monthly users. Public data cannot reliably distinguish genuine users from automated activity.

Averages, growth and charts

Each hourly average is the mean of its collected player counts, rounded to a whole player. The 24-hour and 7-day averages weight these hourly values by their sample counts. Daily values average the available hourly averages equally; the 30-day metric then weights those daily values by their sample counts. Consequently the 30-day value is not a pure mean of raw samples. None of these methods reconstructs unobserved time; varying collection frequency and gaps can affect comparisons. Analytics are normally recomputed every 10 minutes.

Player growth compares the average for the selected 24-hour, 7-day or 30-day window with the preceding equal-length window. Percentage change is (current average − previous average) / previous average × 100. No percentage is shown when the preceding average is zero or unavailable. Short histories can cover only part of a window. If a positive rolling mean is unavailable, the stored 24-hour metric falls back to the latest count; longer metrics fall back to the available shorter-window metric.

Visit and favorite gains subtract the earliest available counter within the window from the latest counter, with negative changes clamped to zero. Hourly boundaries are used for 24-hour and 7-day analytics; daily boundaries for 30-day analytics. These are approximate rolling windows, especially when data is missing. Daily tables use UTC; today is incomplete. A visit-day record compares successive daily ending counters, so a gap between days can span more than one day.

The 24-hour and 7-day charts use raw snapshots. The 30-day chart uses hourly averages for players and ending counters for visits; the full-history chart uses daily summaries. Chart averages are not the peak figures. Initial charts and tables are included in the page; range controls request additional data. Player-count tiles refresh separately from the page's tables and analysis.

Current and historical audience analysis

Our game audience analyses update from the recorded observations. Rolling performance summaries describe the latest available periods. Selected games also have calculated four-week comparisons that advance after each complete Monday–Sunday UTC week. Their figures and factual summaries are generated from the data, not a human editorial review. The current partial week is excluded. These selected games are not a representative sample of all Roblox experiences.

A game qualifies only when all 672 distinct hourly buckets in the four-week window contain at least one observation. Each Monday–Sunday UTC week gives its 168 hourly means equal weight. This differs from the rolling, sample-weighted metrics above and allows comparisons between complete weeks with the same weekdays. Missing hours are never filled with zero. If the latest four complete weeks have gaps, the comparison is omitted rather than replaced by an older period. A populated bucket confirms that observations exist, not continuous coverage throughout that hour.

The typical-hour figures summarize the 168 hourly means in the final complete week of the displayed period. The median is their midpoint; the middle 50% runs from the 25th to the 75th percentile. Quartiles use type-7 linear interpolation between sorted values, rather than rounding to a single observation. These describe sampled concurrency, not how many different people played.

To describe the daily rhythm, we take the median of the 28 hourly means for each UTC clock hour across the full four-week period. The busiest and quietest displayed hours are the highest and lowest of those medians; exact ties use the earlier UTC hour. Game pages show those hours in your time zone, using the period's final day for the time-zone conversion. Weekend comparisons give equal weight to 192 Saturday/Sunday hourly means and separately to 480 Monday–Friday hourly means across the same four weeks. Those differences can reflect both weekly patterns and changes over the period; they do not isolate a weekend effect.

Peaks are the largest recorded samples in the named window, not continuously measured or verified lifetime records. A weekly change does not establish its cause, player retention, unique visitors, revenue or future performance. Event explanations require independent evidence. Send corrections through our contact page.

Separately, collapsed historical studies preserve their original observations and reviewed interpretations. The archived study covers Sep 14 – Oct 11, 2026, from a read-only export taken on October 15, 2026 and reviewed on 2026-10-15. Its findings remain tied to that period; current counts do not rewrite them. Published audience reports and historical articles retain the same distinction.

Trending and ratings

The trending score compares the 24-hour average with the 7-day average: (avg24h − avg7d) / (avg7d + 25) × log10(avg24h + 10) × 100. The smoothing term reduces extreme percentages from tiny audiences; the logarithm weights audience size. This is our discovery heuristic, not a Roblox metric or a forecast. Positive rating is simply upvotes divided by all votes, expressed as a percentage; it is not a five-star review score.

Avatar Marketplace prices and history

The avatar Marketplace uses Roblox's public catalog, item, thumbnail and collectible resale APIs. Prices are in Robux. For limited items, an original sale price is different from the current lowest resale offer; an item originally released free can cost Robux on resale. Offers can change or disappear before you reach Roblox. Purchases and trades take place on Roblox; RoVitals does not hold items or execute transactions.

Recent Average Price (RAP) and dated sales charts are supplied by Roblox. RAP describes recent sales and is not an assigned trading value, an appraisal or a promise of what an item will sell for. RoVitals price observations start when we first collect an item; they are separate from Roblox's resale history. Missing prices and collection gaps are not zero-price sales. Marketplace searches and detail refreshes are cached and paced independently of game tracking.

Price-drop comparisons identify their reference. A fresh Roblox RAP can be used for a limited item; otherwise an observed historical reference needs several separate days of prices. These comparisons describe a price difference, not a guaranteed saving or profit. Read the actual reference date and availability before comparing offers. Item wishlists remain on this browser and do not send alerts.

Revenue model and assumptions

Displayed developer earnings are a model, not reported sales or a measurement of a game's monetization. Daily USD is the stored 24-hour average concurrency × 24 × an assumed earned-Robux rate per player-hour × $0.0038. Genre rates are RoVitals assumptions; Roblox does not publish these per-game rates. No estimate is shown when that input is zero or unavailable. Examples from the current model:

The displayed daily range divides and multiplies the model estimate by three. It is not a statistical confidence interval. The 30-day projection multiplies the central daily model by 30 and assumes unchanged activity and monetization. Actual earnings can fall outside the range. Purchases, revenue shares, creator eligibility, audience mix and expenses are not observed.

Roblox's official DevEx guidance describes different rates and eligibility rules. Our fixed $0.0038 assumption does not model older earned balances or the higher rate for eligible US adult spending. Treat earnings estimates as a rough scenario, not financial data.

Affiliation

RoVitals is a fan-made statistics site. It is not affiliated with, endorsed by, or sponsored by Roblox Corporation. All game names, trademarks and images belong to their respective owners.

Get in touch

Corrections, creator enquiries and data questions are welcome — see our contact page. Our privacy policy and terms of use explain how data is handled.

Browse all tracked games