Research · re-tested every night on our own data
55 things crypto traders say, each tested on our own price history and re-tested every night: 2 hold up, 5 are false, the opposite held and 48 come down to luck. Each verdict is read from a study's own numbers, so it changes when the history does. Past patterns, not financial advice.
past the study's bar, the older and the newer half of the history agreeing
“Bitcoin's best days come right after its worst.”
Of Bitcoin's 20 best days since 2014, 7 came within 7 days of one of its 20 worst, against 6% with the worst days' dates slid anywhere else in the record.
The study →updated
“New Binance listings bleed out.”
Of 523 coins listed on Binance since 2018, 73% sat under their listing close six months later, and they did −29.6% against the market on average.
The study →updated
the other way, past the same bar
“CME gaps always fill.”
85% of Bitcoin's gaps filled within 30 days, but 87% of the levels just as far the other way were reached too. The gap's level came first 42% of 319 times, where a coin flip gives 50%.
The study →updated
“When shorts pile in, a squeeze follows.”
On 2,788 coin-days since 2019 when shorts paid 50% a year or more on one of the 100 most traded altcoins, the coin did −3.0% against the other coins over the next 7 days on average (−4.2% in the middle case), ahead 35% of the time.
The study →updated
“Altcoins that crash 90% always come back.”
Of the 230 times an altcoin we hold from its first day fell 90% or more from its all-time high at least three years ago, it closed above that high again within three years 27 times (12%); coins that stopped trading count.
The study →updated
“A coin that crashes 20% bounces back.”
After 2,635 crashes of 20% or more with no news behind them, the coin did −5.4% against the market over the next week.
The study →updated
“Chase the pump: a coin up 20% keeps running.”
After 6,412 pumps of 20% or more with no news behind them, the coin did −1.1% against the market the next day and −1.5% over the week.
The study →updated
the numbers lean the saying's way, not past the bar: luck could make them
“Uptober: Bitcoin goes up in October.”
Bitcoin rose in 9 of 12 Octobers, +16.3% on average, against 54% of all 153 months; with twelve months to choose from, one looks this good 79% of the time by luck alone.
The study →updated
“Bitcoin's summer half, May to October, lags November to April.”
Over 12 years Bitcoin did +46% from May to October on average and +44% from November to April; November to April came out ahead in 7 of 12 years.
The study →updated
“Be greedy when others are fearful.”
Bitcoin bought on the 243 days of extreme fear since 2018 did +4.3% over the next 30 days on average, against +3.9% after the other days.
The study →updated
“When Bitcoin makes a new high, the alts follow.”
After the 72 days Bitcoin closed at a new all-time high, the 50 most traded altcoins did +35.4% against Bitcoin over the next 90 days, against −2.6% after the other days.
The study →updated
“Altcoins fall harder than they rise with Bitcoin.”
In 5 of 6 full years since 2020, the typical one of the 50 most traded altcoins fell further on Bitcoin's down days than it rose on its up days. Over the year to Oct 14: 1.45 times Bitcoin's fall on its down days against 1.41 times its rise on its up days. With its drift against Bitcoin taken out, its beta was higher on the down days in 6 of the 6 years.
The study →updated
“The altcoins that swing hardest with Bitcoin pay the most.”
Over 28 quarters since 2019, the half of the 50 most traded altcoins that had moved most with Bitcoin over the year before did +0.6% a quarter against the other half on average, ahead in 14 of them.
The study →updated
“When longs pay through the nose, the coin is about to dump.”
On 187 coin-days since 2019 when longs paid 100% a year or more on one of the 100 most traded altcoins, the coin did +3.1% against the other coins over the next 7 days on average (0.0% in the middle case), ahead 31% of the time.
The study →updated
“An altcoin down 90% is a bargain.”
Bought at the close of a 90% crash and held a year, 341 altcoins did −11% against Bitcoin on average (−44% in the middle case); held together, day by day, +5% a year against the other coins.
The study →updated
“New all-time highs lead to more highs.”
After 1,745 new all-time highs, the coin did +2.0% against the market over the next week on average, beating it 47% of the time.
The study →updated
“Bitcoin goes wild on CPI day.”
On 57% of the 69 US CPI release days since 2021, Bitcoin moved more than on its usual weekday of the 20 before; on the other weekdays, 50% did.
The study →updated
“A Puell Multiple of 4 marks Bitcoin's top.”
Bitcoin's Puell Multiple read 4 or more in 2 stretches since 2014 with a year after it: a year after Dec 4, 2017, Bitcoin was −66% from that day's close; a year after Jan 5, 2018, Bitcoin was −78% from that day's close. Too few days for a luck test.
The study →updated
“Under its power-law line, Bitcoin is a bargain.”
On the 1,102 days since 2017 Bitcoin closed at 0.50 to 0.99 times its power-law line - the line fitted on the closes up to each day - it did +170.2% over the next year on average, against +19.3% after the other days.
The study →updated
“Bitcoin peaks 12 to 18 months after a halving.”
Of the 3 halving cycles on Bitcoin's closes since 2016, the peak came 12 to 18 months after the halving in all 3: 17.2 months after the 2016 halving (Dec 16, 2017); 17.9 months after the 2020 halving (Nov 8, 2021); 17.5 months after the 2024 halving (Oct 6, 2025). A handful of cycles is no proof.
The study →updated
“The Pi Cycle Top indicator calls Bitcoin's cycle top within days.”
Of the 3 halving cycles on Bitcoin's closes since 2014, the Pi Cycle Top fired within 30 days of the cycle's peak in 1; the 2020 cycle peaked on Nov 8, 2021, 210 days from its nearest firing; the 2024 cycle peaked on Oct 6, 2025 without it firing.
The study →updated
“When funding runs hot, the top is in.”
On the 230 days Bitcoin's funding ran at 30% a year or more since 2019, Bitcoin did +1.0% over the next 30 days on average, against +4.6% after the other days.
The study →updated
“When everyone is talking about Bitcoin, the top is in.”
On the 151 days since 2015 that Bitcoin's Wikipedia page drew 3× its usual readers or more, Bitcoin did +14.1% over the next 90 days on average, against +22.6% after the other days.
The study →updated
“Money printing pumps Bitcoin.”
On the 730 days since 2015 the US money supply (M2) was growing 10% a year or faster, Bitcoin did +28.5% over the next 90 days on average, against +20.8% after the other days.
The study →updated
“When Wall Street panics, Bitcoin is a bargain.”
On the 226 days since 2014 the VIX stood at 30 or more, Bitcoin did +9.6% over the next 30 days on average, against +5.0% after the other days.
The study →updated
“When stocks are in a bear market, Bitcoin goes down too.”
On the 342 days since 2014 the Nasdaq stood 20% or more under its high, Bitcoin did +8.3% over the next 90 days on average, against +22.0% after the other days.
The study →updated
“When Americans pay up on Coinbase, Bitcoin rises.”
On the 164 days since 2021 Bitcoin's Coinbase premium was the highest of its year, Bitcoin did +6.0% over the next 30 days on average, against +1.3% after the other days.
The study →updated
“Bitcoin follows the lunar cycle: up as it waxes, down as it wanes.”
Over 157 lunar months since 2014, Bitcoin moved +2.9% on average in the waxing half of the month (up in 91 of 157) and +1.9% in the waning half (up in 79 of 157; a luck test of +0.8).
The study →updated
“When the kimchi premium runs hot, the top is in.”
On the 647 days since 2017 Bitcoin cost 3% to 10% more in Korea than elsewhere, it did +5.6% over the next 90 days on average, against +17.1% after the other days.
The study →updated
“When the dollar falls, Bitcoin rises.”
On the 593 days since 2014 the dollar index fell 1.5% or more in a month, Bitcoin did +43.2% over the next 90 days on average, against +15.4% after the other days.
The study →updated
“When stablecoins are printed, Bitcoin goes up.”
On the 1,209 days the stablecoins' supply grew 6% or more in a month since 2017, Bitcoin did +17.7% over the next 90 days on average, against +10.7% after the other days.
The study →updated
“A death cross means more pain ahead.”
After 574 death crosses among the 100 most traded coins, the coin did −2.6% against the market over the next 30 days on average, beating it 38% of the time.
The study →updated
“When ETH/BTC is cheap, Ether catches up.”
On the 406 days ETH/BTC stood under 0.80 of its 200-day average, Ether did +1.5% against Bitcoin over the next 90 days, against +0.8% after the other days.
The study →updated
“Negative funding sets up a short squeeze.”
On the 294 days Bitcoin's funding was under zero since 2019, the shorts paying the longs, Bitcoin did +2.8% over the next week on average, against +0.7% after the other days.
The study →updated
“Exchange listing news pumps the coin.”
On the day of a Binance spot listing article, coins that already traded did +2.8% against the market on average (75 articles).
The study →updated
the numbers do not even lean the saying's way
“Once altcoin season starts, the alts keep winning.”
After the 228 days our altcoin season index read 75 or over, the 50 most traded altcoins did −15.9% against Bitcoin over the next 90 days, against −0.1% after the other days.
The study →updated
“A little every week beats putting it all in at once.”
On 1,519 of 4,305 first days since 2014, $100 a week for a year ended with more Bitcoin than $5,200 put in at once; in the middle case all at once held 20% more. Each Jan 1 as one case: the weekly plan ended with 9.0% fewer coins on average over 12 years.
The study →updated
“Token unlocks dump the price.”
Across 182 unlocks of 0.5% of the supply or more, the coin did −0.2% against the market over the 30 days before and +5.6% over the 30 days after.
The study →updated
“Bitcoin far under its 200-day average is a bargain.”
On the 781 days Bitcoin closed under 0.80 times its 200-day average, it did +5.6% over the next 90 days on average, against +23.0% after the other days.
The study →updated
“Under its 2-year moving average, Bitcoin is the bargain of the cycle.”
On the 804 days since 2015 Bitcoin closed under its 2-year moving average, it did +136.1% over the next year on average, against +152.9% after the other days.
The study →updated
“CPI day dumps Bitcoin.”
Bitcoin closed lower on 48% of the 69 US CPI release days since 2021, against 51% of the other weekdays.
The study →updated
“A Puell Multiple under 0.5 marks Bitcoin's bottom.”
On the 96 days since 2014 Bitcoin's Puell Multiple read under 0.5, Bitcoin did +85.6% over the next year on average, against +146.9% after the other days.
The study →updated
“Bitcoin follows the stock-to-flow model.”
On 35% of the 2,764 days since PlanB published the stock-to-flow model (Mar 22, 2019), Bitcoin closed within a factor of 2 of its price, and on none since the 2024 halving; the middle close sat at 32% of it, and on Oct 14, 2026 Bitcoin closed at 10% of its $810,575.
The study →updated
“Extreme greed comes before a fall.”
After the 95 days of extreme greed since 2018, Bitcoin did +31.6% over the next 30 days on average, against +3.1% after the other days.
The study →updated
“When miners capitulate, the bottom is in.”
After 35 recoveries of Bitcoin's hash ribbons since 2014 (the hash rate's 30-day average back over its 60-day after miners capitulated), Bitcoin did +9.7% over the next 90 days on average, against +19.4% after the other days.
The study →updated
“Fed rate cuts pump Bitcoin.”
After 11 Fed rate cuts since 2019, Bitcoin did −1.6% over the next 30 days on average, against +5.2% after the other days.
The study →updated
“Low volatility comes before a big move.”
After the 686 quietest days since 2014 (the calmest 10% of their year), Bitcoin's largest swing over the next 30 days averaged 20.7%, against 22.9% after the other days.
The study →updated
“Bitcoin dumps when the US market opens.”
Over 3,212 days since 2017, Bitcoin's US hours (13:00-21:00 UTC) moved +0.048% a day on average (a luck test of +1.3); the late hours after them, 21:00-24:00 UTC, +0.053%.
The study →updated
“When hedge funds load up on Bitcoin shorts, the price drops.”
On the 630 days since 2018 hedge funds' CME position was the most short of its year, Bitcoin did +5.2% over the next 30 days on average, against +4.3% after the other days.
The study →updated
“The dip is a bargain: Bitcoin does better after a big fall.”
On the 1,384 days since 2014 Bitcoin closed 15% to 30% under its highest close of the last 90 days, it did +1.9% over the next 30 days on average, against +7.3% after the other days; 30% or more under it, +4.4%.
The study →updated
“A golden cross starts a rally.”
After 669 golden crosses among the 100 most traded coins, the coin did −1.6% against the market over the next 30 days on average, beating it 37% of the time.
The study →updated
“RSI under 30 means a bounce is coming.”
After 3,126 RSI falls under 30 among the 100 most traded coins, the coin did −0.7% against the market over the next 7 days on average, beating it 41% of the time.
The study →updated
“RSI over 70 means a drop is coming.”
After 4,131 RSI rises over 70 among the 100 most traded coins, the coin did +1.3% against the market over the next 7 days on average, beating it 42% of the time.
The study →updated
“Weekends, Mondays and Fed weeks move crypto.”
Not one test of the days of the week, the weekend, the turn of the month, the quarter's last week, the options expiry week or the Fed's week held on both parts of the history (116 tests in all).
The study →updated
“Big caps move first, then the small caps catch up.”
A day after the big caps moved, the smaller coins followed with a correlation of −0.09 to +0.04, and 0 of 67 catch-up trades held on both parts of the history.
The study →updated
Each saying is a question one of our standing studies asks of the whole price history every night: the coins it is about, measured against the market over the same days, or Bitcoin on those days against its other days.
Before trading costs and funding. Every study lists its dated revisions, so a verdict that changes leaves a record.