Hot hand fallacy: a streak is not a hotter next bet

The belief is easy to say. Someone just won, scored, or spun the color you like. You treat the next try as more likely to go the same way.

The streak is the evidence. The next result is supposed to inherit it.

Gilovich, Vallone, and Tversky opened their 1985 paper with that basketball version. Players and fans, they wrote, tend to believe a player's chance of hitting a shot is greater after a hit than after a miss. They called the pattern the hot hand and streak shooting.

The paper is about skilled shots, not a casino wheel. A shooter can be tired, guarded, or close to the basket. A wheel pocket cannot.

The casino mistake is to carry the belief across that gap. Casino knowledge is the shelf this page sits on. Adults 18+ only.

A hot hand fallacy 2026 reading is the same belief with a newer calendar. The paper's date did not expire. A streak in October 2026 still does not edit the next independent draw.

Say the hot hand fallacy out loud before the stake moves.

Study 1 asked 100 basketball fans from Cornell and Stanford. They played at least occasionally, and they watched at least five games a year. Ninety-one percent said a player has a better chance of making a shot after making the last two or three than after missing the last two or three.

They also gave a number. For a hypothetical player who shoots 50% from the field, the average estimate was 61% after a make and 42% after a miss. Every respondent put the after-make figure at least as high as the after-miss figure.

Study 2 used field goals from 48 home games of the Philadelphia 76ers and their opponents in the 1980–81 season, recorded by the team statistician. Table 1 covers nine 76ers. The weighted chance of a hit was 51% after a hit and 54% after a miss.

That runs against the hot-hand guess.

Serial correlations were negative for eight of the nine players. Only one player, Dawkins, differed from zero enough for the paper to mark it significant, and that correlation was also negative. The authors wrote that the records provided no evidence for a positive correlation between successive shots.

That fan gap is the hot hand fallacy in numbers.

A controlled Cornell shooting test, summarized in the abstract, found that previous shots changed players' predictions and did not change their performance. This page does not retype every later table. The claim it uses is the one the authors printed: belief on one side, and no positive dependence in the samples they showed.

Later arguments about that statistic are not reviewed here, because those papers were not the file we opened. Do not upgrade a 1985 sample into a law about every athlete who has ever lived.

The same paper names the opposite error. A locally representative view of chance, they wrote, induces the belief that heads is more likely after a long run of tails than after a long run of heads. They called that the gambler's fallacy.

One belief says the streak continues. The other says the streak must break. They point at different bets.

They share a cause: treating a short run as if it had used up, or stored, probability.

The belief that a color is due is the break-the-streak version. This page is the continue-the-streak version. On a fresh wheel, both are wrong for the same reason.

The pocket count does not read the history board.

Basketball was a fair place to test the continue-the-streak idea, because a person is not a wheel. The authors looked and, in those records, did not find the rise fans expected. A casino wheel does not need that debate.

Nothing in the mechanism stores a Purple.

How swingy a reel is is a design fact about a game. A hot streak on that reel is still a sample. Volatility tells you the swings are large.

It does not tell you the next spin inherited the last one.

Take the 33-slot wheel on this site. Sixteen slots are Purple, sixteen are Silver, and one is Green. Purple's chance is 16/33 on the first spin and 16/33 on the spin after four Purples.

The history strip is a diary. It is not an input.

Purple and Silver pay 2x, then a 5% fee comes off the winning payout. The edge after that fee is about 7.88%. Green pays 14x and, after the same fee, carries an edge near 59.70%.

A "hot Green" story would have you raise a stake on the worse edge because the rare slot finally appeared. The appearance did not change the fraction.

The price built into a payout is how that 7.88% is named. The name stays put after a streak. If you double the stake because the color feels live, you double the average price.

You do not buy a better chance.

A pencil version of the same fractions is the worksheet. Use it when a streak makes the percent feel negotiable. The percent is the arithmetic.

The feeling is the fallacy.

A lead at a poker table can pick up the same mood. Raising because a short session feels hot spends chips on a mood. The cards did not get easier.

Example A, a bigger stake on a live color. Your usual Purple bet is $5. After four Purples you bet $20.

The chance is still 16/33. The extra $15 is priced at about 7.88% after the win fee, an average loss near $1.18 on that extra stake. One spin will not lose $1.18.

It will pay the 2x ticket, minus the fee, or pay nothing. The $1.18 is the average you bought by sizing up.

Example B, the paper's own number, which is not a casino price. Fans guessed a 50% shooter would hit 61% after a make. That 11-point gap was a belief inside a questionnaire.

It is not a payout a book posted, and this page will not turn 61% into a wager. Quoting the fans is not permission to bet them.

Example C, the two errors on one spin. The board shows six Purples. One person bets Silver because Purple is "due." Another bets more on Purple because Purple is "hot." Both are treating 16/33 as if it moved.

It did not. The two bets can both lose to Green, which is 1/33 whether the board is loud or blank.

None of the examples is a system for catching a streak or fading one. A system that needs the past to edit the next independent draw is the fallacy with a notebook.

The rows compare objects. A belief does not have a rake. A game does.

The second table is the belief, not a strategy card.

How PvP compares

Roulette will show recent colors. The list is allowed to be streaky. Fairness recomputes the slot from a seed committed before the spin.

A match says the seed was not swapped after you saw the streak. It does not say the streak caused the slot.

Sixteen Purples in a row would still leave the next spin at 16/33 for Purple. That sentence is the whole protection this page offers. It is not a pick.

If the strip is why the stake grew, the stake is the leak. Close it before the next round, not after the one that "owes" you.

Read it once when the board looks persuasive.

• [ ] I can say the belief in one line: a streak raises the next chance.
• [ ] I can say the other belief in one line: a streak makes the opposite due.
• [ ] I know the 1985 fans guessed 61% after a make for a 50% shooter.
• [ ] I know the 76ers weighted figures they printed were 51% after a hit and 54% after a miss.
• [ ] I am not treating a basketball sample as a promise about a wheel.
• [ ] Purple is 16/33 after any history on this site's wheel.
• [ ] Raising the stake multiplies the edge. It does not edit the chance.
• [ ] I did not turn this page into a system for riding or fading streaks.
• [ ] The money is spare. A hot board is not a reason to use rent.
• [ ] If I cannot name the pocket count, I do not bet the feeling.

The list is a pause. It is not a selection method. Passing it means you noticed the belief.

It does not mean the next spin is a good purchase.

Do not play if you are under 18, or under the age where you live. Do not play with money assigned to rent, food, or a bill. A clear description of a fallacy does not make the game cheaper.

Do not use the 76ers table as a tip sheet. Those rows are a study of shots. They are not a forecast for tonight's wheel, and they are not an edge you can buy.

If a streak is how you decide to chase, name the hot hand fallacy and stop. Use responsible gambling and 1-800-GAMBLER in the US and Canada. The board will keep printing runs.

You do not have to fund them.