How many Mega Millions numbers in a draw are odd?
Mega Millions has been drawn as 5 numbers from 70 plus a Mega Ball from 24 since April 2025. Everything on this page counts only the 161 draws since that change.
Across 161 Mega Millions draws the five numbers drawn have included 2.37 odd numbers on average, where chance predicts 2.50.
The two sit 0.13 of a number per draw apart, or 1.56 standard errors, with the record below the prediction. Every number in a draw came up odd in 3 draws, against a predicted 4.3.
What was measured one count per draw
Every number in the pool is odd or even, so a single count settles a whole draw. Count the odd numbers in it and the even ones are whatever is left. Draw 161 was 1328333548, and three of them are odd, 133335, so it counts three. That one count, draw by draw, is the whole measure, and every figure below is read out of it.
Chance predicts the count out of each draw's own pool. Of the 70 numbers in today's pool, 35 are odd and 35 are even, which is a dead heat. The two sides are the same size, so the predicted count in a draw of five is exactly 2.50, and chance favours neither side by any amount at all. This game has drawn from one pool of 70 numbers throughout, so that same prediction stands behind every draw on record. The per draw sum is kept anyway, because it stays right if the pool ever changes.
The record 161 draws, 6 rows
Red marks a figure outside the range ordinary variation produces
| Odd numbers in the draw | Chance | Draws |
|---|---|---|
| 0 | 4.32.7% | 116.8% |
| 1 | 24.415.1% | 22-0.52 SE |
| 2 | 51.832.2% | 53+0.20 SE |
| 3 | 51.832.2% | 50-0.30 SE |
| 4 | 24.415.1% | 22-0.52 SE |
| 5 | 4.32.7% | 31.9% |
Each row is a count of odd numbers and the number of draws that produced it, beside the number chance predicts and the share of the record that comes to. The prediction behind a row is summed from each draw's own pool, so a row means the same thing across the whole record, and no row gathers a tail of several counts together, so the comparison between them is level. The commonest count is 2, and chance puts the same count first. Two rows are predicted fewer than five draws each, too few for a standard error to mean much, so those rows carry their own share of the record instead and are left out of the red marking.
The table reads 6 counts in all, of which four carry a range, and the furthest of those from its prediction is 0.52 standard errors out. Chance alone predicts 2.5 of them that far out, so a furthest reading of that size is the ordinary result of reading several of them rather than anything about odd and even. Nothing in the table is printed in red: every count sits inside the range ordinary variation produces. The rows between them cover every draw read, so they lean on one another; a prediction for one row does not depend on what the others do, which is what keeps the figures above exact.
Set the two sides against each other and the record gives the even side 0.26 of a number per draw more than the other, where chance predicts a dead heat. That gap is twice the figure above less the draw size, so it says the same thing a second way rather than anything new. A draw of five numbers cannot split level, so one side holds the majority every time. The odd side has held it in 75 draws against a predicted 80.5, and the even side in 86 draws against a predicted 80.5. Chance splits those two predictions evenly, because every pool on record holds as many odd numbers as even. A clean sweep for one side is the far end of the same scale: all odd has happened three times against a predicted 4.3, and all even has happened eleven times against a predicted 4.3.
Recent trend last 100 draws, then blocks of 20
Over the last 100 draws the count has averaged 2.53 odd numbers against a predicted 2.50, which is 0.28 standard errors above it. Over that stretch the odd side ran 0.06 of a number per draw ahead of the other, where chance predicts a dead heat. A stretch that short carries a wide range of ordinary variation, which is what the band in the charts below is for.
Odd numbers per draw, one point per block of 20 draws
result for the blockrange ordinary variation allowsoutside that range
Share of draws with an odd majority, one point per block of 20 draws
result for the blockrange ordinary variation allowsoutside that range
Oldest block on the left, newest on the right, with the same eight blocks of 20 draws in both charts. The band is what chance predicts for those same draws, worked out from their own pools, and a point outside it is marked. One point in the first chart lies outside its band, and one in the second. The second chart moves in steps, because a block of 20 draws can only hold a whole number of majorities.
Applied to this draw draw 161 · 2026-10-20
Draw 161 was 1328333548 from a pool of 70, which holds 35 odd numbers and 35 even ones.
This draw leaned odd by one number, where chance predicts a dead heat. Which side was ahead in one draw says nothing about which side will be ahead in the next one, and the record above is the record of which side has been ahead.
Related statistics
Share this page
If this report does not answer your question, tell us what you need and we will look at building it. Contact