Every Powerball draw publishes how many tickets hit each prize tier. Ratios between tiers reveal how many people were holding the drawn structure — the one thing that changes a ticket's expected value without touching its odds.
Live · updates daily
Draws with data
1,412
2015-10-07 → 2026-10-07
Median tickets / draw
13.4M
estimated from tier ratios
Avg over-play of drawn 3-sets
0.996
± 0.004 · chance = 1.000
Draws where players over-held the structure
33%
over-play index above 1
Draws made of birthday numbers were held by 1.47× as many players as draws made of high numbers
Most tickets are quick picks, which are uniform, so the overall average sits near 1. The action is in the split: the minority who choose their own numbers choose small ones.
Draws with four or five whites ≤ 31: over-play 1.31 ± 0.02 (164 draws). Draws with one or none: 0.89 ± 0.00 (345 draws). Overall mean 0.996 ± 0.004 from the 3-white tiers; the noisier 4-white tiers agree at 0.994 ± 0.007.
Birthday numbers: the more drawn numbers ≤ 31, the more people were holding them
Over-play index of the drawn 3-sets, grouped by how many of the five white balls were 31 or below. Whiskers are ± one standard error.
Table view
Bucket
Draws
Over-play
± SE
0
58
0.866
0.005
1
287
0.897
0.003
2
542
0.934
0.002
3
361
1.048
0.005
4
142
1.274
0.014
5
22
1.551
0.044
Numbers ≤ 12 (day-and-month picks)
Grouped by how many drawn whites were 12 or below.
Table view
Bucket
Draws
Over-play
± SE
0
542
0.923
0.003
1
585
0.984
0.004
2
244
1.119
0.012
3+
41
1.422
0.040
Consecutive pairs
Do people over-pick runs like 23-24?
Table view
Bucket
Draws
Over-play
± SE
No consecutive pair
1041
0.993
0.004
Has a consecutive pair
371
1.006
0.010
Sum of the five whites
Low-sum draws are the ones made of small numbers — the birthday effect again, seen from the other side.
Table view
Bucket
Draws
Over-play
± SE
Low sum (<159)
467
1.148
0.008
Mid sum (159–194)
473
0.953
0.002
High sum (≥195)
472
0.890
0.002
The red ball: which Powerballs get over-picked
Share of tickets holding each Powerball, relative to 1/26, pooled across every draw of that ball. Bars above the line are over-played.
Table view
PB 07: 1.131 (48)
PB 11: 1.092 (48)
PB 13: 1.050 (53)
PB 09: 1.044 (58)
PB 05: 1.039 (60)
PB 10: 1.029 (50)
PB 04: 1.026 (66)
PB 12: 1.025 (48)
PB 03: 1.024 (58)
PB 08: 1.022 (45)
PB 18: 1.006 (61)
PB 17: 1.001 (45)
PB 06: 0.995 (52)
PB 21: 0.983 (62)
PB 19: 0.982 (51)
PB 14: 0.977 (68)
PB 22: 0.974 (48)
PB 26: 0.974 (54)
PB 23: 0.973 (52)
PB 25: 0.970 (60)
PB 02: 0.964 (56)
PB 15: 0.959 (46)
PB 20: 0.957 (61)
PB 24: 0.955 (63)
PB 16: 0.938 (40)
PB 01: 0.930 (59)
White balls: estimated over-play by number
Ridge regression of each draw's over-play index on which numbers were drawn (69 indicators). Exploratory — read the shape, not any single bar.
Table view (most to least over-played)
#07: 1.142
#11: 1.115
#09: 1.102
#03: 1.099
#12: 1.098
#08: 1.093
#10: 1.076
#05: 1.076
#19: 1.071
#17: 1.070
#06: 1.065
#04: 1.064
#23: 1.059
#21: 1.054
#27: 1.050
#02: 1.049
#01: 1.045
#18: 1.044
#13: 1.041
#24: 1.035
#16: 1.033
#28: 1.033
#22: 1.033
#15: 1.029
#25: 1.027
#29: 1.025
#20: 1.024
#14: 1.021
#26: 1.015
#31: 1.014
#30: 1.013
#33: 1.003
#32: 0.992
#39: 0.979
#36: 0.975
#38: 0.974
#44: 0.974
#37: 0.974
#42: 0.973
#45: 0.971
#35: 0.971
#54: 0.969
#47: 0.968
#34: 0.967
#40: 0.966
#52: 0.963
#69: 0.961
#43: 0.960
#41: 0.959
#46: 0.959
#53: 0.957
#49: 0.952
#51: 0.951
#48: 0.950
#55: 0.949
#59: 0.947
#50: 0.946
#57: 0.946
#56: 0.944
#58: 0.936
#60: 0.936
#66: 0.933
#62: 0.929
#68: 0.929
#61: 0.929
#63: 0.927
#67: 0.925
#64: 0.925
#65: 0.913
Most and least over-held winning draws
The over-play index for individual draws. High values: many players were holding pieces of that structure.
Most over-held
2024-07-10
07 11 12 27 46 26
1.97
2022-10-10
03 06 11 17 22 11
1.93
2024-09-16
08 09 11 27 31 17
1.87
2022-08-31
07 08 19 24 28 01
1.85
2019-02-09
01 02 03 07 39 25
1.85
2018-04-18
09 10 12 17 23 09
1.83
2016-05-21
05 07 09 23 32 26
1.82
2017-09-27
08 10 21 23 25 22
1.81
Least over-held
2017-06-21
14 46 61 65 68 13
0.74
2022-07-06
32 36 49 62 69 13
0.79
2016-08-10
23 56 61 64 67 12
0.79
2015-10-17
48 49 57 62 69 19
0.80
2020-02-08
35 49 50 59 66 06
0.80
2024-07-24
16 42 59 63 68 13
0.80
2016-09-21
01 28 63 67 69 17
0.80
2018-05-23
20 54 56 61 64 07
0.80
So what?
This is the only lever a number-picking method actually has.
Every combination has the same chance of being drawn. What differs is how many people you would split with. A structure that players over-hold by 20% pays, on average, a jackpot share about 20% smaller. Balliqa's filters happen to steer away from the most over-played shapes (High/Low needs numbers above 34; Tens Diversity spreads the decades), but they were never designed for it. This page is the first step toward deciding whether they should be.
How it works
The data
For every draw since October 2015, powerball.com lists how many tickets nationwide hit each of the nine prize tiers. We collect those counts once a day. Nothing here uses sales figures; they aren't published per draw.
Backing out sales
A tier's winner count is tickets sold × share matching k whites × share holding that Powerball. The 3-white tiers with and without the Powerball share everything except the red ball, so their ratio gives the Powerball's share directly. The three lowest Powerball tiers together cover 99.8% of all tickets holding that red ball, which pins down sales.
The over-play index
With sales known, the 3-white tiers tell us how many tickets carried three of the drawn numbers, compared with the C(69,5) expectation. 1.00 means exactly as many as chance; 1.10 means ten percent more people were holding that structure. It is a property of the drawn numbers, so pooling draws by their features reveals what players over-pick.
What it does not say
Nothing here changes the odds of any combination. A less-popular structure has the same 1 in 292 million chance; it simply shares a jackpot with fewer people if it hits. Per-number indices come from a regression on 1,000+ draws and are exploratory.