Quick Pick vs Self-Pick: What Lottery Data Actually Shows

The stat everyone cites

You have probably seen the claim: 70% to 80% of lottery jackpot winners used quick pick (computer-generated random numbers). This is true. It is also meaningless in the way most people interpret it.

The reason 70-80% of winners used quick pick is that 70-80% of all tickets sold are quick picks. If 75% of tickets are quick picks and 75% of winners used quick picks, quick pick has no advantage. It just has more volume.

This is survivorship bias applied to lottery data. It would be like observing that most car accidents involve cars with four wheels and concluding that four-wheeled cars are more dangerous.

What the odds actually say

Every combination has the same probability of being drawn. 1-2-3-4-5 with Powerball 6 has the same 1 in 292,201,338 chance as any quick pick combination. The drawing machine does not know or care how the numbers on your ticket were selected.

This is not a simplification. It is the mathematical reality. The balls in the machine have no memory of previous draws, no preference for patterns, and no awareness of how the numbers on tickets were chosen. Each draw is an independent random event.

Quick pick does not increase your odds of winning. Self-pick does not increase your odds of winning. Nothing increases your odds of winning except buying more tickets (which increases them linearly but at a terrible cost-to-probability ratio).

Where self-pick has a real disadvantage

While the odds of any combination being drawn are identical, the odds of sharing a prize are not. Self-pick players tend to choose numbers with human patterns: birthdays (1-31), lucky 7s, sequences, geometric patterns on the play slip.

These patterns mean that certain combinations are held by more players. If 7-14-21-28-35 is played by 5,000 people and a random quick pick combination is played by 3 people, the quick pick holders would each receive a much larger share of a split jackpot.

Quick pick numbers are distributed more uniformly across the entire number range. They are less likely to cluster in the 1-31 birthday range. They are less likely to form recognizable patterns. This means quick pick tickets are less likely to duplicate other players' selections.

The practical impact: if you win with quick pick numbers, you are slightly more likely to be the sole winner. If you win with birthday-based self-pick numbers, you are slightly more likely to share the prize.

The data from actual drawings

Looking at Powerball and Mega Millions winners:

The $2.04 billion Powerball winner in 2022 used a quick pick. The $1.602 billion Mega Millions winner in 2024 used a quick pick. The $1.537 billion Mega Millions winner in 2018 used a quick pick.

The $758.7 million Powerball winner in 2017 (Mavis Wanczyk) used a quick pick. The $590.5 million Powerball winner in 2013 (Gloria MacKenzie) used a quick pick.

Among smaller jackpot winners, the split is roughly proportional to the overall quick pick vs self-pick ratio. There is no statistically significant difference in win rates between the two methods when adjusted for volume.

Some self-pick winners have won large jackpots too. Richard Lustig, who wrote a book about lottery strategies, won seven times using self-selected numbers (though his wins were mostly modest prizes, and his total spending likely exceeded his total winnings).

The randomness quality argument

Some players argue that quick picks from lottery terminals are "more random" than human-selected numbers. This is partially true. Humans are terrible random number generators. When asked to pick random numbers, people avoid repeats, prefer odd numbers, cluster around the middle of the range, and gravitate toward numbers with personal significance.

Lottery terminal RNGs produce genuinely random selections that cover the full number range uniformly. Over thousands of tickets, quick pick selections produce a more even distribution across all numbers than human selections do.

But "more random" does not mean "more likely to win." A perfectly random selection and a terribly non-random selection both have the same probability of matching the draw. The randomness only affects the likelihood of sharing a prize, not the likelihood of winning one.

The "system" players

A subset of self-pick players use systems: wheeling, frequency analysis, pattern avoidance, or other structured approaches. These systems select numbers based on criteria (avoiding popular numbers, covering ranges evenly, matching historical patterns).

None of these systems change the probability of any combination being drawn. What some of them do is reduce the probability of sharing a prize (by avoiding popular numbers) and increase the number of combinations covered (wheeling).

A wheeling system that generates 20 tickets covering 12 numbers does give you 20 chances instead of 1. But it costs 20 times as much. The expected value per dollar spent is identical to buying 20 random quick picks. The wheel just structures which combinations you buy.

The honest answer

Quick pick and self-pick have the same odds of winning. Quick pick has a marginal advantage in expected prize size because the numbers are less likely to duplicate other players' choices. Self-pick has no mathematical advantage of any kind.

If you enjoy picking your own numbers, pick them. The experience of choosing has value to you even if it does not change the math. If you do not care and just want a ticket, quick pick is fine and has the slight anti-splitting benefit.

The only genuinely bad strategy is picking extremely popular combinations (all low numbers, obvious sequences, previous winning numbers) and hoping to win big. You might win, but you will share with thousands.

Use a random number generator if you want the anti-splitting benefit of quick pick but want to choose from your computer rather than the store terminal. The mathematical outcome is the same.