The Side Facing Up Before a Coin Flip Wins Slightly More Often, and One Flipper Hit 60%

Ice Water, Math Quizzes, and Shocks: The 1948 Breastfeeding Experiment That Showed Stress Can Slow Milk FlowIn 1948, researchers stressed a nursing mother with ice water, toe pulling, and math problems with shocks. Each time, her baby got less milk. Thanksgiving 2026 Almanac Outlook Calls for Upper Midwest Snow and Eastern Rain as Travelers Plan AheadThe Old Farmer’s Almanac predicts snow in the Upper Midwest and rain in the East for Thanksgiving 2026. Here is how to plan travel safely and stay healthy. Cervical Cancer Screening Gaps Persist in the South and West as At-Home HPV Testing Expands for WomenA new study shows uneven cervical cancer screening across U.S. regions. Here are current Pap and HPV schedules and at-home testing options. Families with Children of Different Citizenship Status Can Check Each Child’s Eligibility for Medicaid or CHIP SeparatelyA JAMA Network Open study finds that over 1 million U.S. children live in families with mixed citizenship status. Learn how Medicaid and CHIP eligibility is decided for each child. Mental Health Women with Autoimmune Diseases Were Nearly 50% More Likely Than Men to Report Being Given a Psychiatric MisdiagnosisWomen with autoimmune diseases were nearly 50% more likely than men to report a psychiatric misdiagnosis in a 6,607-person survey. What to know. National Depression Screening Advice Was Linked to More Therapy and Less Antidepressant-Only Use Among Distressed AdultsAfter depression screening became routine advice, distressed adults used more therapy and fewer antidepressants alone. What patients can ask. After Bend Elementary School Shooting, Oregon Officials Urge Families to Watch Children for Signs of Trauma and GriefAfter the Bend school shooting, Oregon officials urge families to check in with kids. Trauma signs by age and where to find help. Teens and Young Adults with Depression Had Four Times the Odds of a Prescription Drug Use DisorderYouths with depression had 4 times the odds of a prescription drug use disorder in a national survey. What parents should know about the findings. Innovation FDA-Approved Brytly Becomes First Over-the-Counter Apraclonidine Eye Drop, Limited to Twice Daily with Contact Lens PrecautionsFDA approved Brytly, the first OTC apraclonidine drop for eye redness. Use it twice daily at most and see a doctor for eye pain or vision changes. A Vine Famous for Copying Plastic Leaves Showed No Mimicry in a Two-Year Greenhouse RetestA vine said to mimic plastic leaves showed no host-driven mimicry in a new preprint. Its leaves changed shape on their own, often by season. The Side Facing Up Before a Coin Flip Wins Slightly More Often, and One Flipper Hit 60%In 350,757 coin flips, coins landed on their starting side 50.8% of the time. The bias depended on the person, and one flipper hit 60%. Cockroach Milk Packs 3 Times the Energy of Dairy, and Its Ig Nobel Crystals Now Inspire Protein DesignCockroach milk crystals hold over 3 times the energy of dairy milk by weight. The 2016 discovery won a 2026 Ig Nobel, but it was never tested as food. Healthy Living Berkeley Springs Boil Water Advisory Now Covers Most Town Customers as Health Officials Check Affected RestaurantsMost Berkeley Springs Water Works customers remain under a boil water advisory. What it means for drinking, cooking, brushing teeth, and restaurants. Buried Cotton Underwear Exposed Lawns as the Sleepiest Soil, While Some Gardens Were Badly OverfedSwiss volunteers buried 2,000 pairs of cotton underwear. Lawns broke them down the least, and some gardens held far more phosphorus than plants need. Houston Pharmacy Behind Recalled Compounded Tirzepatide Vials Drew 14 FDA Inspection Observations, Including Exposed Skin and Research-Grade IngredientsFDA inspectors cited 14 sterility and manufacturing concerns at Houston RX, the maker of recalled compounded tirzepatide. What GLP-1 users should check. Medicare Advantage Members Who Rely on Wrong Doctor Listings Would Owe Only In-Network Costs Under CMS ProposalA CMS proposal would cap costs at in-network rates for Medicare Advantage members misled by wrong doctor listings. It would start in 2028, so verify now. Covid – 19 FDA Proposes Requiring Companies to Report Food Additives That Now Reach Grocery Shelves Without Federal ReviewFDA has released its proposed GRAS rule requiring companies to notify the agency about food additives. Here is what the proposal does and does not do. Federal Regulators Warn Some Hair Dryer Brushes Lack the Part That Stops ElectrocutionCPSC warns hair dryer brushes sold on Walmart.com and TikTok Shop lack immersion protection. The seller has not agreed to a recall or refund. A Recalled Teething Toy Broke the Pattern Parents Were Told to Look ForA newly recalled pull-string teething toy has a different model number and design from earlier recalls, and one choking incident has been reported. Trying to Lower Blood Sugar, Cholesterol or Blood Pressure? Different Workouts Suited Different MarkersA pooled analysis of 53 trials found that different exercise types helped different metabolic markers. Most of the evidence was low certainty. KIRILL KUDRYAVTSEV | AFP via Getty Images Share on Twitter Share on Facebook Share on Pocket A coin toss is supposed to be the fairest way to settle a dispute. A marathon experiment involving 350,757 flips found that it is very slightly tilted, not by the coin, but by the person flipping it.

Coins landed on the same side they started on 50.8% of the time, according to the study, which won the 2024 Ig Nobel Prize in Probability and was published after peer review in the Journal of the American Statistical Association in 2025. The bigger surprise was how much the effect varied from person to person, and how it seemed to fade with practice.

The project, led by František Bartoš and Eric-Jan Wagenmakers of the University of Amsterdam, set out to test a prediction made in 2007 by Stanford statistician Persi Diaconis and colleagues. Their physics model suggested that human flips add a slight wobble, so a coin spends a bit more time in the air with its starting face up. They estimated a same-side probability of about 51%.

Earlier coin marathons could not settle the question. Statistician Karl Pearson flipped a coin 24,000 times in the 19th century, and mathematician John Kerrich flipped 10,000 times in the 20th, but neither recorded which side each coin started on. A later set of 40,000 flips by Janet Larwood and Priscilla Ku at the University of California, Berkeley, gave mixed results: Larwood’s flips showed a small same-side bias, while Ku’s showed almost none.

For the new test, 48 people flipped coins from dozens of currencies and denominations, recording whether each landed on the side it started on. Many took part in all-day flipping marathons lasting up to 12 hours. Each person flipped in runs of 100, choosing the first starting side at random and then flipping each coin from whichever side had just landed up. Flips had to be caught in the hand, and participants videotaped their sessions so the data could be checked.

Of the 350,757 flips, 178,079 landed on the starting side. The researchers found strong evidence for this same-side bias, but no bias toward heads or tails overall: heads came up almost exactly half the time, and no coin showed a meaningful lean.

The 50.8% average hides wide differences. Individual same-side rates ranged from about 48.7% to 60.1%, according to the study’s person-by-person data. The highest rate came from a participant who contributed 2,800 flips.

“Our experiment suggests it’s not [up to] the coin, but the person flipping the coin,” Bartoš told ABC News in Australia in 2024.

The four people with rates above 53% each flipped fewer than 4,000 times. That fits another pattern in the data: the bias tended to shrink as people flipped more. The authors estimated that about 10,000 flips, or roughly 10 hours of practice, might be enough to virtually eliminate it, possibly because flippers improve with experience. They noted that fatigue or waning attention could also explain the decline.

An informal replication cited in the published paper pointed the same way. When 820 people each flipped a coin 10 times, about 50.9% of the flips landed on the starting side.

The bias is tiny for any single toss, and it does not mean heads or tails is favored in general. If the starting side is chosen at random, the outcome is still a fair 50/50.

But in the published paper, the authors calculated that someone who bet a dollar on each of 1,000 coin flips and knew the starting side would come out about $16 ahead on average. That edge is bigger than the casino’s advantage in six-deck blackjack against a player using perfect strategy, though smaller than the house edge in single-zero roulette.

They also pointed to a $5 million double-or-nothing coin flip on the Amazon Prime show “Beast Games,” where they estimated that knowing the starting face would have been worth about $95,000 in expected value. Their advice for high-stakes tosses is to keep the coin’s starting position hidden.

The study has limits. Participants knew what was being tested, so the authors could not completely rule out that some influenced their flips, consciously or not, though they judged that unlikely based on the videos and how closely the data matched the 2007 physics model. The experiment also counted only coins caught in the hand, not coins that were spun or allowed to bounce.

A fair flip matters because people lean on coins for weightier choices than who kicks off. In a separate field experiment published in 2021 in The Review of Economic Studies, economist Steven Levitt had undecided people flip a coin over decisions such as quitting a job. Those told by the coin to make a change were more likely to do so and reported being happier six months later.

Those findings came with caveats. Participants volunteered online, happiness was self-reported, and many people did not follow the coin. Levitt described the results as suggestive rather than definitive.

For anyone settling a choice with a flip, the coin data offer one easy fix: have someone else flip, or cover the coin before anyone calls it.

Not quite. In 350,757 flips, coins landed on their starting side 50.8% of the time.

No. Heads and tails came up almost equally, and no coin showed a meaningful heads or tails bias.

Human flips add a slight wobble that keeps the starting face up a little longer in the air, as a 2007 physics model predicted.

No. Individual rates ranged from about 48.7% to 60.1%, and the bias tended to shrink with practice.

Hide which side is facing up before the flip, or have a neutral person flip the coin.