State of the Wheel 2026: What 9,380 Spins Show About How People Pick

Most writing about name wheels is guesswork about how people use them. We counted instead. Every spin on the NameWheel wheel sends back three facts and nothing else: how many names were on it, which mode it was in, and when it happened. Here is what 9,380 of those spins show, along with the device split of our visitors and every run of our public fairness test.

Where the Numbers Come From

When someone spins the main wheel on namewheel.org, the page sends one small message to our server with the number of names on the wheel and the mode it is in. The server adds the time and, if the person is signed in, a link to their account. That is the whole record. The names themselves never leave the browser, so this report could not tell you the most popular name on a wheel even if we wanted it to.

• Period: the first spin was logged on 24 July 2026 and this snapshot was taken on 29 October 2026, a little over three months.
• Scope: the main wheel at namewheel.org only. Spins on themed wheel pages, embedded wheels, the mobile apps and the AI assistant connector are not in these counts.
• Our own spins are included. Testing the site adds spins we cannot separate out. They are a small share of the total, but they are there.
• Devices come from a different source: the anonymous page speed reports browsers send when a visitor leaves a page, which record phone or computer and nothing about the person. Those started on 16 October 2026, so the device split covers 1,536 page views over two weeks.

Times are in UTC. With visitors spread across the United States, Europe and Asia, a single local time would mislead, so read the hour chart as a world clock rather than as anyone's school day.

People Spin for Groups, Not Coin Flips

The typical wheel holds 13 names. The average is far higher at 45.3, pulled up by a long tail of very large lists. Only 3.6% of spins were a choice between two options, the yes or no and this or that decisions a coin could settle. A wheel, it turns out, is a group tool.

Three groups stand out:

• Six to ten names (30.6% of spins) is the biggest band. That is a table of friends, a small team, a family deciding who does what.
• Eleven to thirty names (18.7% plus 13.0%) is the size of a class or a department. One spin in eight lands on a list of 21 to 30 names, which is exactly the range of a typical classroom: calling on students, picking presenters, choosing who answers next.
• More than 100 names (8.3%, or 782 spins) is the giveaway and raffle tail. Lists this long are almost always entries: comments, ticket numbers, email sign ups. 22% of all spins had more than 30 names on the wheel.

Small choices of two to five options make up just 15.7% of spins. If you are building a tool for picking from a group, this is the shape to design for: a list that fits on screen at around a dozen names, and that still works at several hundred.

A Note on Eight Names

The single most common list size was exactly eight names, with 905 spins or 9.6% of the total. We would love to report that eight is humanity's favourite group size, but the honest reason is simpler: the wheel opens with a sample list of eight names, and many people spin it once to see how it works before typing their own. Eight is in the six to ten band above, so that band is partly a measure of curiosity. Even if every one of those spins were a test, six to ten names would still be one of the two largest bands.

A Quarter of Spins Remove the Winner

NameWheel has a normal mode, where the winner stays on the wheel, and an Eliminate mode, where each winner is taken off before the next spin. 24.1% of all spins (2,259) happened in Eliminate mode. A further 1.4% used Tally mode, which keeps a running count of how often each name wins.

Eliminate mode is how people turn a one winner tool into an ordering tool: who presents first, second and third, the order of turns in a game, the sequence of a draft. One in four spins being part of a sequence says that a lot of people come to a wheel to put a whole group in order, one spin at a time.

That finding is why we built the marble race name picker, which ranks a whole list in a single race, and why the team generator exists alongside the wheel.

A Weekday Tool

Spins on an average weekday ran 1.8 times higher than on an average weekend day. Thursday was the busiest day with 1,768 spins; Saturday was the quietest with 791.

Party games and Friday night decisions exist, but they are not what drives a name wheel. Classrooms, team meetings and workplace rituals do. The shape is the same one schools and offices produce: a working week that builds towards midweek, a softer Friday, and a quiet weekend.

The Hours People Spin

By hour, the busiest windows were 18:00, 14:00 and 13:00 UTC. That is the morning and early afternoon in the eastern United States, when American schools are in session. A second, smaller rise in the early UTC hours lines up with the school and working day in Asia and Australia.

The quietest hours were 23:00 and 22:00 UTC, late evening in Europe and early evening on the American east coast, once the school day there has ended. Nights do not stop a wheel, but school bells clearly start one.

Seven in Ten Visits Come From a Computer

Across the 1,536 page views we have device data for, 70% came from a computer. On the homepage, where the wheel lives, it was 72%. Most of the web is read on phones, so this is the reverse of the usual pattern, and it fits everything above: a wheel is usually something one person runs on a laptop for a whole room, through a projector, a smartboard or a shared screen on a call.

That has a practical lesson for anyone running a group pick. The person spinning is rarely the person who needs to read the result. Large text, a clear winner moment and a sound people can hear across a room matter more than they would on a personal app. It is also why the OBS overlay and the full screen mode exist.

Signed in spins made up 22.6% of the total. Most people spin without an account, which is how it should be for a free tool; accounts are for saving wheels, Pro features and certified draws.

Draws That Need Proof

For prizes and anything else where the result might be questioned, NameWheel offers certified draws. The entry list is fingerprinted, the server commits to a secret seed before the draw, and the finished record is signed and published at a public link where anyone can check it. In this period there were 68 completed certified draws with a median of 14.5 entries, and people locked 21 entry lists ahead of a draw so they could prove later which list was used.

Certified draws are a small share of all wheel use, as they should be. Most picks are low stakes: whose turn it is, who answers next. When stakes are real, the gap between "I spun a wheel" and "here is a record you can check" matters, and the verified wheel spin page explains how it works.

168,000 Test Spins in the Fairness Lab

The Fairness Lab lets anyone run thousands of simulated spins in their own browser and see how evenly the wins spread. Every run logged so far used eight names. Each run records one number: the gap between the name that won most and the name that won least, as a share of a perfectly even count. 12 runs have been logged, 168,000 spins in total.

A fair wheel never comes out perfectly even. Short runs always show wide gaps and long runs narrow them. To judge the runs, we simulated 3,000 fair runs at 500 and at 5,000 spins, and 600 at 50,000 spins, and marked the range that 90 in 100 fair runs fall inside.

All three runs of 50,000 spins and all three runs of 5,000 spins sit inside the range chance produces, and the gaps shrink as runs get longer, exactly as fair randomness predicts. Five of the six 500 spin runs do too.

The sixth is the odd one, and we show it rather than leave it out. The very first run, on 31 July, came back with a gap of just 4.1%. At 500 spins, a result that even happens in fewer than one fair run in a thousand. Too even is as suspicious as too lopsided, and every run after it has behaved normally. One run cannot tell us what happened that day, which is exactly why the lab lets anyone run it again.

What to Take From This

1. Pick the tool for the job. For one winner, spin. To put a whole group in order, use Eliminate mode or a marble race. To split people into groups, use the team generator rather than spinning one name at a time.
2. Plan for the room. Seven in ten visits come from a computer, and a wheel run from a laptop is usually shown to a whole group. Turn on full screen and sound, and say the winner out loud.
3. Large lists deserve a record. One spin in twelve has more than 100 names, usually entries for something of value. If people entered, they may ask how the winner was chosen. A certified draw answers that before anyone asks.
4. Fair does not look even. In a class of 25, the same student coming up twice in a week is normal randomness. If you want everyone called once, remove names as they are picked.

Questions About This Report