Is the wheel random? Our fairness test

We ran the wheel's real picking code 16,004,000 times, in a browser and in Node.js, and checked every result. Here is exactly how a winner is picked, what we tested, every number we got, and what a test like this cannot tell you.

The short answer

16,004,000spins of the real code
0 of 24tests flagged a bias
0times the wheel stopped on a different slice than the one picked

None of the 24 tests found a bias. With 2, 6, 10 and 37 entries, every entry won its fair share to within a small fraction of a percentage point, which is the amount chance alone gives. The spin time, the order of wins and elimination mode made no difference either. This is a strong check, but it is a sample, not a proof: see what this test cannot tell you.

How the wheel picks a winner

Everything happens in your browser, in this order:

  1. A random number. The wheel asks your browser for random 32-bit numbers with crypto.getRandomValues, the cryptographically secure generator built into every modern browser, which draws on randomness collected by your device's operating system. It asks for 256 at a time and uses them one by one. Each is a whole number from 0 to 4,294,967,295, divided by 4,294,967,296 to give a number u from 0 up to (not including) 1. The wheel never uses Math.random, the older and weaker generator.
  2. The winner. With n entries, the winner is entry number floor(u × n): the bottom 1/n of the range picks the first entry, the next 1/n the second, and so on. This happens the moment you press Spin, before the wheel moves.
  3. Where it stops. A second random number picks a stopping point between 12% and 88% of the way across the winning slice, so the pointer never lands on or near a line between two slices and there is never a disputed result.
  4. The animation. The wheel then turns from wherever it is resting to that point, adding whole turns for show: 5 at the 3 second setting, 8 at 5 seconds, 13 at 8 and 19 at 12, plus up to one more to reach the slice. It slows down with a smooth ease-out curve. The curve, the number of turns and your device's speed only change how the spin looks; the winner was fixed in step 2.
  5. The result. The winner shown is the entry picked in step 2. Our test also reads the wheel's final angle every time and checks that the pointer really is on that slice.

These are the lines that do it, copied by the test script straight out of the wheel's JavaScript file:

var RB = new Uint32Array(256), RI = 256;
function rnd() { if (RI >= 256) { crypto.getRandomValues(RB); RI = 0; } return RB[RI++] / 4294967296; }

var n = it.length, a = 2 * Math.PI / n, win = Math.floor(rnd() * n);
var target = -(win * a + (0.12 + rnd() * 0.76) * a);

An entry you type twice gets two slices, and twice the chance. That is the only way to weight an entry, so the weighted test below uses repeated entries.

Equal shares, down to the last value

Because u comes from one of 4,294,967,296 (232) equally likely numbers, we can count exactly how many of them pick each entry, without any sampling. With 2, 4, 8 or any power of two entries, every entry gets exactly the same count. Otherwise 4,294,967,296 does not divide evenly, so some entries get one more value than others. The difference is real but far too small to matter or to measure:

EntriesValues per entryEntries with one extraLargest difference in chance
22,147,483,648nonenone
6715,827,882 or 715,827,8834 of 61 in 715,827,882 of the chance
10429,496,729 or 429,496,7306 of 101 in 429,496,729 of the chance
37116,080,197 or 116,080,1987 of 371 in 116,080,197 of the chance
10042,949,672 or 42,949,67396 of 1001 in 42,949,672 of the chance
1,0004,294,967 or 4,294,968296 of 1,0001 in 4,294,967 of the chance

On a 10-entry wheel, for example, 6 entries win with probability 429,496,730 / 4,294,967,296 and 4 with 429,496,729 / 4,294,967,296: a gap of about one in 430 million of a 10% chance. You would need many quintillions of spins to notice it. Even at the 1,000-entry maximum, the gap is 1 in 4,294,967 of each entry's 0.1% chance. The test script also checks the boundaries between every pair of neighbouring entries, so we know the computer's arithmetic matches this count exactly.

How we tested it

The real code, not a copy. The test script opens the wheel's JavaScript file, cuts out the random number buffer, the spin function and the function that reads which slice is under the pointer, and runs them unchanged. Only the page around them is replaced: no drawing, no sound, and the animation jumps straight to its final frame. To check the animation as well, 2,000 spins per run play every frame at 60 frames a second.

Two random number generators. The same test ran in headless Chromium 141.0.7390.37, using the browser's own crypto.getRandomValues, and in Node.js v22.22.0, whose crypto.getRandomValues comes from a different library (OpenSSL). Nothing is seeded, because crypto.getRandomValues cannot be seeded; every run gives new numbers.

What ran. 1,000,000 spins each on wheels of 2, 6, 10 and 37 entries; 1,000,000 spins of a wheel with one entry on 3 slices, one on 2 and one on 1; 250,000 spins of a 10-entry wheel at each spin time; 500,000 back-to-back pairs on 6 entries; and 250,000 full draws of a 4-entry wheel in elimination mode, removing each winner as the wheel does. The wheel was never reset between spins, as on the site.

The test. For each wheel we counted the wins and compared them with what a fair wheel gives, using Pearson's chi-square test. The p-value is the chance that a perfectly fair wheel would give counts at least this uneven. Fair wheels give p-values spread evenly between 0 and 1, so a single small one is expected now and then.

Decided in advance. We fixed the rule before the run: a test counts as a bias if its p-value is below 0.001 and a fresh run of that test comes back below 0.001 again. If one had, we would have fixed the wheel instead of publishing this page.

Run it yourself. The script needs Node.js and Playwright. It takes the path of the wheel's JavaScript file, finds the three functions by their text (and stops if they are not there), runs every test above in Node.js and then in headless Chromium, and writes the counts, chi-square values and p-values to a JSON file. Our run took about 2 minutes on one processor core. The chi-square p-values come from the regularized incomplete gamma function, which the script checks against textbook table values before it starts.

Run on November 13, 2026. Chromium 141.0.7390.37 (headless, Playwright); Node.js v22.22.0 (OpenSSL 3.5.4). Wheel code: assets/stw.js, SHA-256 9aa82eda9563c49d… (the file this site serves now).

Results

Every test, both runs. A p-value of 0.001 or more means no sign of bias.

Chromium 141.0.7390.37

TestSpinsChi-squaredfp-valueResult
2 entries, equal slices1,000,0000.0210.88No sign of bias
6 entries, equal slices1,000,0001.5750.90No sign of bias
10 entries, equal slices1,000,0007.9190.54No sign of bias
37 entries, equal slices1,000,00030.32360.74No sign of bias
Repeated entries as weights (3 : 2 : 1)1,000,0002.7120.26No sign of bias
10 entries, 3 second spin250,0007.4290.59No sign of bias
10 entries, 5 second spin250,00017.3090.04No sign of bias
10 entries, 8 second spin250,0003.7390.93No sign of bias
10 entries, 12 second spin250,0007.7190.56No sign of bias
Winner independent of spin time1,000,00030.15270.31No sign of bias
Back-to-back pairs, 6 entries1,000,00037.05350.37No sign of bias
Elimination mode: all 24 orders of 41,000,00021.40230.56No sign of bias

Node.js v22.22.0

TestSpinsChi-squaredfp-valueResult
2 entries, equal slices1,000,0000.4810.49No sign of bias
6 entries, equal slices1,000,0004.4750.48No sign of bias
10 entries, equal slices1,000,00017.7590.04No sign of bias
37 entries, equal slices1,000,00030.75360.72No sign of bias
Repeated entries as weights (3 : 2 : 1)1,000,0000.5520.76No sign of bias
10 entries, 3 second spin250,0008.1890.52No sign of bias
10 entries, 5 second spin250,00010.6890.30No sign of bias
10 entries, 8 second spin250,00015.2790.08No sign of bias
10 entries, 12 second spin250,0007.4190.59No sign of bias
Winner independent of spin time1,000,00036.66270.10No sign of bias
Back-to-back pairs, 6 entries1,000,00027.97350.79No sign of bias
Elimination mode: all 24 orders of 41,000,00021.40230.56No sign of bias

Across the 24 tests, the smallest p-value was 0.04 and 2 fell below 0.05, against about 1.2 expected by chance alone. "df" is degrees of freedom: one less than the number of outcomes counted.

Equal slices: how far each entry was from its fair share

The biggest gap between an entry's share of wins and its fair share, in the browser run:

EntriesFair shareBiggest gapIn standard errorsChi-square (df)p-value
250%0.008 pts0.150.02 (1)0.88
616.667%0.041 pts1.091.57 (5)0.90
1010%0.046 pts1.537.91 (9)0.54
372.703%0.035 pts2.1330.32 (36)0.74

A gap is measured in standard errors because that is the size chance alone gives: with 1,000,000 spins, a fair 10-entry wheel's entries usually land within about ±0.06 points of 10%. On a 37-entry wheel, at least one of the 37 entries lands beyond 3 standard errors in about 1 run in 11 by chance. The single biggest gap in either run, entry 4 on the 10-entry wheel in the Node.js run (2.49 standard errors above), got a fresh 5,000,000-spin browser run of its own: there it was 1.02 standard errors above, and that whole wheel scored chi-square 2.81 (df 9, p = 0.97). A biased slice would have stayed off.

2 entries, 1,000,000 spins in the browser: gap from a fair share, in standard errors−4−20+2+4Entry 1: 499,923 wins (expected 500,000), −0.15 standard errorsEntry 2: 500,077 wins (expected 500,000), +0.15 standard errors12
2 entries, 1,000,000 spins in the browser: gap from a fair share, in standard errors. Bars above the line won a little more than 1 in 2, below a little less. Dashed lines: ±2 standard errors.
Every count for 2 entries
EntryWinsShareGap from 1 in 2
1499,92349.992%−0.008 pts
2500,07750.008%+0.008 pts
6 entries, 1,000,000 spins in the browser: gap from a fair share, in standard errors−4−20+2+4Entry 1: 166,774 wins (expected 166,667), +0.29 standard errorsEntry 2: 166,529 wins (expected 166,667), −0.37 standard errorsEntry 3: 166,260 wins (expected 166,667), −1.09 standard errorsEntry 4: 166,846 wins (expected 166,667), +0.48 standard errorsEntry 5: 166,783 wins (expected 166,667), +0.31 standard errorsEntry 6: 166,808 wins (expected 166,667), +0.38 standard errors123456
6 entries, 1,000,000 spins in the browser: gap from a fair share, in standard errors. Bars above the line won a little more than 1 in 6, below a little less. Dashed lines: ±2 standard errors.
Every count for 6 entries
EntryWinsShareGap from 1 in 6
1166,77416.677%+0.011 pts
2166,52916.653%−0.014 pts
3166,26016.626%−0.041 pts
4166,84616.685%+0.018 pts
5166,78316.678%+0.012 pts
6166,80816.681%+0.014 pts
10 entries, 1,000,000 spins in the browser: gap from a fair share, in standard errors−4−20+2+4Entry 1: 99,995 wins (expected 100,000), −0.02 standard errorsEntry 2: 99,851 wins (expected 100,000), −0.50 standard errorsEntry 3: 99,743 wins (expected 100,000), −0.86 standard errorsEntry 4: 100,175 wins (expected 100,000), +0.58 standard errorsEntry 5: 100,100 wins (expected 100,000), +0.33 standard errorsEntry 6: 100,331 wins (expected 100,000), +1.10 standard errorsEntry 7: 99,540 wins (expected 100,000), −1.53 standard errorsEntry 8: 100,204 wins (expected 100,000), +0.68 standard errorsEntry 9: 99,645 wins (expected 100,000), −1.18 standard errorsEntry 10: 100,416 wins (expected 100,000), +1.39 standard errors12345678910
10 entries, 1,000,000 spins in the browser: gap from a fair share, in standard errors. Bars above the line won a little more than 1 in 10, below a little less. Dashed lines: ±2 standard errors.
Every count for 10 entries
EntryWinsShareGap from 1 in 10
199,99510.000%−0.001 pts
299,8519.985%−0.015 pts
399,7439.974%−0.026 pts
4100,17510.018%+0.018 pts
5100,10010.010%+0.010 pts
6100,33110.033%+0.033 pts
799,5409.954%−0.046 pts
8100,20410.020%+0.020 pts
999,6459.965%−0.036 pts
10100,41610.042%+0.042 pts
37 entries, 1,000,000 spins in the browser: gap from a fair share, in standard errors−4−20+2+4Entry 1: 26,932 wins (expected 27,027), −0.59 standard errorsEntry 2: 27,175 wins (expected 27,027), +0.91 standard errorsEntry 3: 26,887 wins (expected 27,027), −0.86 standard errorsEntry 4: 26,857 wins (expected 27,027), −1.05 standard errorsEntry 5: 26,721 wins (expected 27,027), −1.89 standard errorsEntry 6: 26,992 wins (expected 27,027), −0.22 standard errorsEntry 7: 27,099 wins (expected 27,027), +0.44 standard errorsEntry 8: 27,085 wins (expected 27,027), +0.36 standard errorsEntry 9: 27,135 wins (expected 27,027), +0.67 standard errorsEntry 10: 27,007 wins (expected 27,027), −0.12 standard errorsEntry 11: 27,070 wins (expected 27,027), +0.27 standard errorsEntry 12: 26,893 wins (expected 27,027), −0.83 standard errorsEntry 13: 27,045 wins (expected 27,027), +0.11 standard errorsEntry 14: 27,143 wins (expected 27,027), +0.72 standard errorsEntry 15: 26,899 wins (expected 27,027), −0.79 standard errorsEntry 16: 26,998 wins (expected 27,027), −0.18 standard errorsEntry 17: 26,808 wins (expected 27,027), −1.35 standard errorsEntry 18: 26,892 wins (expected 27,027), −0.83 standard errorsEntry 19: 27,039 wins (expected 27,027), +0.07 standard errorsEntry 20: 27,251 wins (expected 27,027), +1.38 standard errorsEntry 21: 27,157 wins (expected 27,027), +0.80 standard errorsEntry 22: 26,932 wins (expected 27,027), −0.59 standard errorsEntry 23: 27,166 wins (expected 27,027), +0.86 standard errorsEntry 24: 26,980 wins (expected 27,027), −0.29 standard errorsEntry 25: 27,212 wins (expected 27,027), +1.14 standard errorsEntry 26: 27,087 wins (expected 27,027), +0.37 standard errorsEntry 27: 27,109 wins (expected 27,027), +0.51 standard errorsEntry 28: 27,189 wins (expected 27,027), +1.00 standard errorsEntry 29: 27,147 wins (expected 27,027), +0.74 standard errorsEntry 30: 26,789 wins (expected 27,027), −1.47 standard errorsEntry 31: 26,681 wins (expected 27,027), −2.13 standard errorsEntry 32: 26,902 wins (expected 27,027), −0.77 standard errorsEntry 33: 27,257 wins (expected 27,027), +1.42 standard errorsEntry 34: 27,122 wins (expected 27,027), +0.59 standard errorsEntry 35: 27,286 wins (expected 27,027), +1.60 standard errorsEntry 36: 27,071 wins (expected 27,027), +0.27 standard errorsEntry 37: 26,985 wins (expected 27,027), −0.26 standard errors171319253137
37 entries, 1,000,000 spins in the browser: gap from a fair share, in standard errors. Bars above the line won a little more than 1 in 37, below a little less. Dashed lines: ±2 standard errors.
Every count for 37 entries
EntryWinsShareGap from 1 in 37
126,9322.693%−0.010 pts
227,1752.718%+0.015 pts
326,8872.689%−0.014 pts
426,8572.686%−0.017 pts
526,7212.672%−0.031 pts
626,9922.699%−0.004 pts
727,0992.710%+0.007 pts
827,0852.709%+0.006 pts
927,1352.714%+0.011 pts
1027,0072.701%−0.002 pts
1127,0702.707%+0.004 pts
1226,8932.689%−0.013 pts
1327,0452.705%+0.002 pts
1427,1432.714%+0.012 pts
1526,8992.690%−0.013 pts
1626,9982.700%−0.003 pts
1726,8082.681%−0.022 pts
1826,8922.689%−0.014 pts
1927,0392.704%+0.001 pts
2027,2512.725%+0.022 pts
2127,1572.716%+0.013 pts
2226,9322.693%−0.010 pts
2327,1662.717%+0.014 pts
2426,9802.698%−0.005 pts
2527,2122.721%+0.018 pts
2627,0872.709%+0.006 pts
2727,1092.711%+0.008 pts
2827,1892.719%+0.016 pts
2927,1472.715%+0.012 pts
3026,7892.679%−0.024 pts
3126,6812.668%−0.035 pts
3226,9022.690%−0.013 pts
3327,2572.726%+0.023 pts
3427,1222.712%+0.009 pts
3527,2862.729%+0.026 pts
3627,0712.707%+0.004 pts
3726,9852.699%−0.004 pts

Repeated entries as weights

A wheel of 6 slices with A on 3, B on 2 and C on 1 should give A half the wins, B a third and C a sixth.

EntrySlicesWinsShareExpected
A3 of 6500,21150.021%50.000%
B2 of 6332,66233.266%33.333%
C1 of 6167,12716.713%16.667%

Chi-square 2.71 with 2 degrees of freedom, p = 0.26.

Spin time

The same 10-entry wheel at each Spin time setting. Each row is tested on its own, and the last line tests whether the four settings differ from each other (a chi-square test of independence on the 4 × 10 table).

Spin timeSpinsLowest entryHighest entryp-value
3 seconds250,0009.92%10.09%0.59
5 seconds250,0009.87%10.13%0.04
8 seconds250,0009.94%10.05%0.93
12 seconds250,0009.91%10.06%0.56

Do the settings differ? Chi-square 30.15 with 27 degrees of freedom, p = 0.31. As expected: the spin time is read after the winner is picked.

Back-to-back spins

If one spin could influence the next, some of the 36 possible pairs on a 6-entry wheel would come up more than others. Over 500,000 pairs, chi-square 37.05 (df 35), p = 0.37. The same entry won twice in a row 16.69% of the time, against 1 in 6 (16.67%) for a fair wheel.

Elimination mode

Drawing all 4 entries of a wheel one by one, removing each winner, can give 24 different orders. Each should come up 1 time in 24 (4.17%). Over 250,000 full draws all 24 orders appeared, from 4.10% to 4.23% of draws; chi-square 21.40 (df 23), p = 0.56.

Where the wheel stopped

In all 16,000,000 quick spins and 4,000 fully animated ones, the pointer ended on the picked slice every time, between 12.00% and 88.00% of the way across it, never on a line.

Try it in your browser

This runs 100,000 spins of the same two steps (crypto.getRandomValues, then floor(u × n)) in your browser, and tests the counts the same way. It takes well under a second. Most runs give a p-value above 0.05; about 1 run in 20 gives less by chance, and that is what a fair wheel does.


Press the button to run the test.

What this test cannot tell you

  • It is a sample, not a proof. A test of a few million spins can show that there is no bias big enough to see. With 1,000,000 spins on 10 entries, a slice that won 10.15% instead of 10% would stand out by about 5 standard errors, but a much smaller bias could hide. The exact count of values per entry above covers the formula; the test covers the code and the random numbers coming into it.
  • We rely on your browser's random number generator. crypto.getRandomValues is built for security work and is far beyond what a test of millions of spins can check, so its quality comes from its design and the operating system behind it, not from this page. We tested Chromium (the engine of Chrome and Edge) and Node.js. We did not run Safari or Firefox for this page; they provide the same function, built separately.
  • It tests the code, not your device. A browser extension, a changed copy of the page or a broken device could in principle change what runs. The wheel's JavaScript is public, so anyone can compare it with the lines above.
  • Random is not the same as even. A fair wheel can pick the same entry three times running, and over a few dozen spins some entries will win much more than others. Elimination mode is the way to make sure everyone is picked once. The odds of repeats are worked out on how the wheel works.
  • Other tools on the site. This page tests the wheel. The team picker and the number tools use the same random numbers in their own code, which this page does not test.

Sources

Questions about this test? Please contact us.

Frequently asked questions

Is the wheel really random?
Yes, as far as any test can show. The winner is picked by your browser's cryptographically secure random number generator (crypto.getRandomValues), and in our test of 8,002,000 spins of the real picking code in Chromium, and the same again in Node.js, no wheel size showed a bias. A test like this can find a flaw but cannot prove there is none.
Does the spin time or the starting position change who wins?
No. The winner is picked before the wheel moves. Spin time only sets how many turns the animation makes, and the starting position only changes how far it turns to reach the picked slice. In our test, the winners at the 3, 5, 8 and 12 second settings did not differ (p = 0.31).
Can SpinTheWheelPro choose or see the winner?
No. The pick happens in your browser and nothing is sent to our server, not your entries and not the result. The code that does it is in the wheel's JavaScript file, which any browser can open and read.
Is an entry that is on the wheel twice more likely to win?
Yes, on purpose. Each line is its own slice, so an entry typed twice has two slices and twice the chance. That is how you give an entry more weight. In our test, an entry on 3 of 6 slices won 50.02% of 1,000,000 spins (expected 50%).
Why is one of the p-values small?
Because we ran 24 tests, and with that many, a few small p-values happen by chance alone: about 1 test in 20 is expected below 0.05 and 1 in 100 below 0.01. We decided before the run that only a p-value below 0.001 that came back on a fresh run would count as a bias. The smallest this time was 0.04 (10 entries, 5 second spin).
Can I check it myself?
Yes. The test further up this page runs 100,000 spins of the same formula in your own browser and shows the chi-square result. Your browser's developer tools also show the wheel's JavaScript, including the two lines that pick the winner.