Sandworm Playground player count and stats
At our latest snapshot on 16 November 2026 the game had 15,400 players online, 7,408,730 visits and 21,772 favourites. Two reads 1.98 hours apart showed 136,442 new visits - about 68,910 an hour - and the all-time peak reported by public trackers is 52,574 players on 15 November.
Snapshot
Two reads, both from the game API
Every number below is read straight from Roblox's public game endpoint, not from a tracker panel or an estimate.
| Metric | Mon, 16 Nov 2026 03:16:00 GMT | Mon, 16 Nov 2026 05:15:00 GMT | Change |
|---|---|---|---|
| Visits | 7,272,288 | 7,408,730 | +136,442 |
| Playing | 17,708 | 15,400 | -2,308 |
| Favourites | 21,462 | 21,772 | +310 |
| Up votes | 4,919 | 4,970 | +51 |
| Down votes | 635 | 649 | +14 |
Like rating at the second read: 88.45%. The window between the two reads was 1.98 hours, so the per-hour figures are a short sample, not a daily average.
Growth
What a two-hour measurement actually tells you
Visits only ever go up, so a difference between two reads is a real rate. Concurrent players are not - they follow time zones.
Visits per hour
68,910
measured over 1.98 h
Visits per day (rate)
1,653,842
the same rate extrapolated, not observed
Favourites per hour
156.6
310 in the window
Between the two reads the concurrent count moved from 17,708 to 15,400 - down, not up, while visits climbed by 136,442. That is not a contradiction. Visits accumulate forever; concurrent players are whoever is in a server at that second. Our first read was late morning in Asia, the second was mid-morning in Europe and the middle of the night in North America, so the same growing game can post a smaller live number an hour later without losing anyone.
This is why single “player count” screenshots are worthless for judging a Roblox game, and why a two-point measurement of visits is worth more than either number alone. It is also why we only publish the extrapolated daily rate as a rate, clearly labelled, rather than as a claim about tomorrow.
Servers
A 100-server census, and what it says about scale
Roblox exposes public server lists, so the server structure is measurable rather than guessed.
Server cap
10
Servers sampled
100
Average occupancy
8.29 / 10
Estimated live servers
~1,858
Every server is capped at 10 players, and the servers that exist are mostly full: our sample of 100 public servers held 829 players between them, with 100 of them at seven or more. Dividing the live player count by that average gives roughly 1,858 servers running at once - an estimate, because we can only page through the public list, not count every server.
Ten-player servers matter for how the game feels. There is no mega-server where the whole player base collides; you get a small crowd, a shared destructible map, and a rotation as people leave. It also means “the game is empty” screenshots are usually just a badly timed server, not an empty game.
Trackers
What public trackers add, and where they differ
Two public trackers follow this game. Neither is official, and their peaks do not always agree.
| Source | Reported peak | What it is |
|---|---|---|
| Tracker A | 52,574 on 15 Nov | Tracks since 8 Nov; its live value matched our API read exactly at the same minute |
| Tracker B | 52.6K, global rank #43 | Also reports a 88.5% rating and 14.1-minute average session |
| An earlier crawl | 44,514 | A stale read captured the same afternoon, since superseded by both trackers |
The number to quote
52,574 is the peak as of 16 November, from the tracker that also matched our own API read on the same afternoon. Anything lower is an older snapshot; anything higher needs a timestamp before you believe it.
Method
How these numbers are collected
All primary figures come from Roblox's public JSON endpoints: the game record for visits, playing, favourites and votes; the badges endpoint for the level funnel; the game-passes endpoint for prices and unlock levels; and the public server list for the census. None of it requires an account, a key or a scraper that violates anything.
The limitation is frequency. We read on demand rather than on a timer, so the site shows a timestamp next to every number, and a page that says 05:15Z is describing that moment and not “now”. If you need a live figure, the game page on Roblox is one click away and always current.
We do not publish retention, daily active users or country splits, because none of those are available from the public API. A site that gives you a country breakdown for this game is either buying panel data or making it up; we would be guessing.
Reading counts
How to tell whether this game is actually growing
Three numbers, three different jobs. Mixing them up is how people conclude a growing game is dying.
Visits only ever increase and never reset, so a difference between two reads is genuine growth. That is the number to use for “is this getting bigger”. Concurrent players is a live gauge that swings with time zones, update windows and the Roblox front page - useful for “is it busy right now”, useless for trend. Favourites is the closest public proxy for intent: someone pressed a button, which is a stronger signal than a visit that may have lasted nine seconds.
Our two reads illustrate the trap neatly. Visits rose by 136,442 while concurrent players fell by 2,308. Both numbers are true. If you screenshot the second one and post it, you can claim the game is collapsing during the fastest growth week it will ever have.
For a longer view, the daily history is the only public series: the average sat at about 1,700 players on 7 November and about 5,400 on 13 November, with single-day peaks of 3,155 and 8,365 on those two dates. The jump happened on 11 November, when the daily average went from about 1,700 to about 5,600. That is one week of data, not a year, and it is aggregated by sites we do not control - so treat it as direction, not measurement.
Limits
What this page cannot tell you
Retention
The API exposes totals, not cohorts. We cannot say what share of yesterday's players came back, and any site that claims to is using panel estimates or inventing numbers.
Country split
Regions are not published per game. Popularity in one market can be inferred from time-of-day peaks, and that is all it is - an inference.
Revenue
Pass sales are not public. We know the prices and the roster, not how many people bought anything.
Historical peaks
Roblox does not publish an all-time peak. Every peak figure you see is a tracker's best observation, which is why two trackers disagree.
We would rather show four honest limits than four confident estimates. If you need a live figure, the game page on Roblox is always current; it is linked at the top and bottom of every page here.
Cadence
How often these numbers change
Visits move every second, favourites every few minutes, badge counts every few hours, and prices almost never - in our 1.98-hour window across two reads, every pass kept the same price and the same availability. That difference in cadence is why the store pages carry a “last verified” note while this page carries a timestamp: the two kinds of data go stale at completely different speeds.
The data on this page is a snapshot, not a live feed. If the numbers here are older than you need, the official game page always shows the current count, and we re-read it every time we touch the site.
Related pages
Last updated 2026-11-16. Both snapshots read 2026-11-16 (03:16Z and 05:15Z). Badge deltas use the same two reads.