Altcoin season in 90 days (37 of 49), not in a year (16)
Over the 90 days to 15 October 2026, 37 of 49 coins beat Bitcoin, right on the usual altcoin season line. Over the full year, only 16 did.
Every chart we publish on TradingView, newest first: one question, the measurement behind it and the answer. Open any idea to read it in full.
Over the 90 days to 15 October 2026, 37 of 49 coins beat Bitcoin, right on the usual altcoin season line. Over the full year, only 16 did.
Over a year of 50 Bybit crypto coins, the next hour rose 57% of the time after a 5% hourly drop, against 49% for any hour.
From January 2010 to September 2026, September was gold's worst month: a median of -3.1%, up in only 5 of 17 years. August the best, +3.6%.
Over the year to 15 October 2026, 7 of 48 coins clearly followed ETH more than Bitcoin, led by NEAR. XRP leaned to Bitcoin. 40 showed no clear lean.
All 159 Volume Oracle regime changes on Nvidia 4H warned the candle before. The yellow came on 113 times: 59 changes, 54 recoveries.
After the KOSPI's 30 earlier falls of 10% or more since 1985, it rose the next month 53.3% of the time, against 53.7% after the other months.
Over the year to 15 October 2026, the typical one of 49 coins closed the same way as Bitcoin on 75% of days. ETH did so on 87%, AKE on 55%.
After the chip index's 43 falls of 10% or more since 1994, it rose the next month 52.4% of the time, against 57.2% after the other months.
SHELL rose 115% on Bybit on 13 November 2026 while Bitcoin fell 12.7% that week. We checked the top minute by minute, and who paid funding.
Since 1986, in the month after WTI closed at $100 or more, the S and P 500 fell 46.7% of the time, against 35.5% after the other months. A lean.
33 Retail Sales hours since February 2024: USD/JPY moved 1.97 times an ordinary hour, the S and P 500 futures 1.24. Luck never reached 1.97.
Over the year to 15 October 2026, silver moved 1.87% for each 1% gold moved, and 2.16% on gold's down days. Every year since 2010 sat between 1 and 2.
Of 1,700 BMSB up crosses on 1D, 56.8% crossed back within 5 days. The 12.4% that lasted 20 days carried 72.5% of the move.
Across 50 Bybit coins over 54 weeks, a green week was followed by another green week 35.4% of the time, against 45.7% after a red week. A lean, not proof.
Over the year to 15 October 2026 a normal day moved natural gas 4.5%, silver 4.4% and oil 4.2%, all more than Bitcoin’s 3.4%.
Since 1978 the S and P 500 rose on 54.3% of its Fridays the 13th and on 54.1% of its other Fridays; gold since 1985, 55.9% against 53.9%.
Over the year to 15 October 2026, the typical coin of 50 closed green on 45.6% of days. Greener coins did a little better: -32% against -45%.
From January 2023 to October 2026, WTI oil's typical weekend gap was 0.26%, 5 times a weekday break, and 77% filled on the Monday.
32 PPI hours since February 2024 against an ordinary hour: US stock futures 2.16 times, the dollar index 2.11, Bitcoin 1.41. CPI moves them more.
In 78 weeks since 2018 Bitcoin dominance jumped a full point. The next week it rose again 47% of the time, less than after ordinary weeks.
Of 2,112 Augury Grid rows that turned Bull or Bear from a status, 48.0% were already on the table just before. Random order: 18.9%.
Since 1970, in the month after US prices rose 0.4% or more, the S and P 500 fell 49.5% of the time, against 34.8% after the other months.
Since 2006, in the 10-year yield's biggest weekly rises the Nasdaq 100 rose 0.45%; in its biggest falls it lost 1.50%. Since 2022 it flipped.
On 8 November 2026, 25 of 50 coins sat in the bottom fifth of their 52-week range, none in the top. Bitcoin 34%, TRUMP 0%, INJ 79%. In 30 days, 4 new lows.
Since 2001, in the 38 months WTI rose 10% or more, the US airline index fell in 25, 65.8%, against 45.9% in the other months.
From 2010 to 2025, copper and the S and P 500 moved the same way on 61% of days. On the S and P's drops of more than 2%, copper fell 87% of the time.
AppLovin rose 71.2% from 30 October to 6 November 2026, its second best week since 2021, while the Nasdaq 100 fell 4.9%. Where the move came from.
Since 1975, after gold fell 10% or more in two months, the next month rose 19 of 27 times, 70.4%, against 51.4% after any month.
After 281 volume spikes on 50 Bybit coins, 3 times normal, the next day moved 1.63 times an ordinary day and went up 50% of the time.
Over 160 weeks, WTI's EIA report hour moved 1.054% in a typical week, against 0.996% for the same hour on other days: only 6% more.
From January 2012 to September 2026, October was Bitcoin's best month and August its worst, but 55% of shuffled calendars made a gap as wide.
From 1988 to 2025, the Russell 2000 beat the S and P 500 in 50.3% of weeks, 48% to 54% in every decade. Since 2007 it moves more.
After Augury Grid showed ⏳ MACD on the 4 hour chart, the row turned within two days 12.6% of the time. After ⏳ Trend, 0.9%.
50 Bybit coins against the Nasdaq 100 over a year: the median coin's correlation 0.36, Bitcoin's 0.48. Take Bitcoin's move out: 0.05.
Over the 12 months to October 2026, the Nasdaq 100 rose 9.8% and Bitcoin fell 29.3%. Yet their daily moves lined up in 10 of the 12 months.
In 2026 the jobs data moved BTC, ETH and SOL 2.54 times an ordinary hour, first again. Housing rose to second, prices fell to sixth and GDP stayed last.
Since 1978, after silver's 5% jumps it was higher a month later 46.6% of the time, after its 5% drops 52.0%, after any day 49.5%.
Of each month's 10 best of 50 Bybit coins, 39 of 120 were back in the 10 best the next month and 36 in the 10 worst. Chance: about 24.
On EUR/USD, London's high and low were both final after two hours on 2 of 258 days. Janus's session lines move until the session ends.
After Intel's 33 earlier 10% drop days since 1972, it was higher a month later 54.5% of the time, against 55.5% after any day.
Over a year, the US session held 44% of the hours and 48% of the movement in 50 Bybit coins, and it was the busiest session for 46 of them.
EUR/USD, GBP/USD and USD/JPY all moved most from 10:00 to 11:00 New York last year, and least at 17:00. Every hour, in pips.
Every US midterm since 1930: the S and P 500 rose from the vote to New Year in 15 of 24, 62.5%, against 66.2% over the same weeks of other years.
The gold to silver ratio was 78.9 on 30 October 2026, lower than 70% of the last five years. Its range: 46.3 in January 2026 to 104.8 in April 2025.
On the 4 hour chart, Harmonic's Extreme Oversold came after a 3.9 ATR drop at the median, and the print candle's low held for two days only 17.8% of the time.
From 2011 to 2025, gold and the US dollar index moved in opposite directions on 66% of days. On the dollar's 1% drops, gold rose 85% of the time.
Of 102 drops of 20% on 50 coins, 59% were won back within a year. NEAR, DOT, OP and 9 more from March 2024 never were.
Over the year to October 2026, the week after Bybit's crowd was most long beat a normal week 49.1% of the time on 50 coins, and after most short 48.0%.
On WTI oil's 4 hour chart, a top came within 10 candles of a Plutus +2 sigma breach 46.4% of the time, against 31.6% after any candle. Timing, not direction.
Over the past year, Thursday was the busiest weekday for 23 of 50 Bybit coins. But a coin's own busy day did not last: only 9 kept it.
Over the year to October 2026, the week after each of 50 Bybit coins' 10 biggest up days ran 3.1% behind a normal week and beat it only 37% of the time.
Oil's normal week moved 9.2% high to low last year, natural gas 9.1% and copper 5.2%. Oil and gas both moved more than Bitcoin's 8.1%.
On 28 September we said payrolls had never had a quiet hour. It printed again on 2 October: 13 of 13. The week of 2 November in UTC, release by release.
Since 1988 the SPX's worst week of a year has been a median -4.6%, the Nasdaq 100's -6.2% and the Russell 2000's -5.8%. All three shared it in 11 years.
47 of 50 coins moved less on a weekend day than a weekday, median 0.71x. Bitcoin 0.46x. BR, AKE and STRK didn't slow down. The full table.
On the S and P 500's 98 days down 2% or more since 2015, Bitcoin rose on 32% of them, against 52% on other days. Gold rose on 52%.
IRYS rose 793% on Bybit from 20 to 23 October 2026, then gave back 75%. We checked the spike minute by minute, and who paid funding.
Over the year to 15 October 2026, AKE, STRK and ADA lost 11 days in a row, the longest of 50 coins. The median coin's longest run was 7 days.
Bitcoin's normal week moved 8.1% high to low, like Tesla's 8.0%. The median coin 17.5%, AKE 45.0%. All 50 coins, and the year before.
From 2016 to 2025, Microsoft's daily moves tracked the Nasdaq 100 most closely of big tech's seven, a correlation of 0.86. Tesla least, 0.57.
A SuperTrend flip on the 4 hour chart ran a typical 2.9% on Bitcoin and 12.6% on ZRO, but about 2 ATR on both. The full table for 44 coins.
From 1988 to September 2026 the S and P 500 made a new all-time closing high in 162 of 465 months. The month after one looked like any other.
On Bitcoin's 22 drop days of 3% or more, the median coin fell 5.0% against BTC's 4.2%. USELESS fell 7.9%, XMR 2.5%. The full 50-coin table.
From 2010 to 2025, a normal week in natural gas moved 10.15% high to low in winter and 7.79% in summer. Winter was wider in 12 of 16 years.
After Volume Oracle's Breakeven, the move kept going two days later 38.9% of the time, against 46.0% for other coins on the same candle. 525 cases, TRUMP shown.
Over 140 months since 2015, Bitcoin's first 3 days of a month rose on 59% of days, the rest 51.5%. The gap faded after 2020.
A normal week moved Tesla 8.0%, Meta 7.4% and Apple 4.5% over the last year, high to low. Five of seven had their biggest week on results.
Over 324 weekends of CME Bitcoin futures, 55.5% of gaps filled the same day and 95.6% by now. Since 29 May 2026 there are no new ones.
Over the last 52 weeks a normal week moved gold 5.1%, silver 11.0% and copper 5.3%, high to low. Silver's normal week more than doubled in a year.
Since 1990 the VIX first closed above 30 on 34 days. A month later the S and P 500 was typically up 3.1%, against 1.2% for any day.
AKE printed a 10% wick on 91 days in a year, Bitcoin on none. The full table of daily wicks for 50 coins on Bybit.
CoreWeave fell 25.2% from 9 to 16 October 2026. The five gaps at the open took it down 25.0%; the five trading sessions added up to just 0.2%.
46 of the 50 most traded coins move most at 14:00 or 15:00 UTC, 10:00 to 11:00 in New York. The full 50-coin table.
A long held all year on Bybit paid 27.1% of its size in funding on USELESS and 2.6% on BTC. LSK longs were paid 172%. The full table, 50 coins.
TradingView flags Housing Starts as high impact. Over 11 releases its hour moved crypto 1.09 times an ordinary hour, and Bitcoin 0.56 times on its own hours.
On 50 coins over two years, closes at least 0.25 ATR above last week's high were still above it a day later 57% of the time. Smaller breaks, 46%.
We measured 50 coins over a year of 4 hour bars. In ATR they dip alike before a move works. In percent, Bitcoin needed 2.3% and AKE 11.7%.
On 8 October 79 of the 150 most traded coins on Bybit hit their low in the same hour. Four days on, W and MAGIC had doubled and the median coin made 10.8%.
1,121 Pentarch Momentum Thrust Bearish prints on 1H crypto. The coin kept falling the next hour 36.2% of the time, against 41.5% for same-hour peers.
633 Exit Overbought prints on 4H across 76 Bybit perpetuals. Twelve hours later the coin was more often higher than its peers, unless longs were crowded.
The jobs report missed and Bitcoin topped within fifteen minutes. DXY and VIX barely moved. Oil and the 10-year yield did, then the alts fell together.
A level 1 ATR from price was reached within a day 49% of the time on BTC, ETH and SOL over 180 days. At 4.5 ATR, 3%. How to check your own levels.
Bitcoin topped Wednesday 23 September. Thirteen of the 29 biggest Bybit alts topped on the weekend, three or four days later, and fell less than the rest.
JOLTs has the biggest median hour of any US release, 3.69 times normal, but a third of its hours were quiet. Payrolls is smaller and has never had a quiet one.
Four coins ran 2x to 4x on funding near zero. LSK ran 5.7x in two days and its shorts paid up to 2 percent an hour. Two public numbers tell those runs apart.
303 US releases over 399 days, ranked by how much the hour containing each moved BTC, ETH and SOL. The jobs data moves crypto more than inflation does.
Six OmniDeck subsystems fired 3,687 times in eight weeks and landed on 3,420 different bars: 92.8 percent unique. Here is how to run the count yourself.
Ten alts, 998 four-hour bars. On the worst days correlation barely moved, 0.18 to 0.17, while the share of the book closing red went from 51 to 91 per cent.
Solana ran 11.91 per cent and reads 78.58. Bitcoin ran 10.79 and reads 86.27. The oscillator was measuring the path, not the distance.
Naked points of control turn price 50.2 per cent of the time. Ordinary value area edges turn it 47.9, and that second number is the finding.
The widest stop produced our best numbers and almost no edge: plus 0.077 R against plus 0.075 for random entries on the same bars.
STRONG appeared at 3% and CRITICAL at 46% in the same week, and the best possible threshold beat a constant guess by one bar in twenty-eight.
Four decisions inside our own multi-timeframe scanner, none of them visible in the table you read.
We rebuilt our own indicators to test them. The rebuilds were wrong, and every check we had said they were fine.
The stop was 5.17% away. The exchange’s exit was 4.16% away. Only one of those can happen first, and it is not yours.
A test of whether agreement between separate tools predicts what happens next, run across three markets and three timeframes. The headline said yes.
A test for whether stacking indicators helps, run across 22 markets and 25,900 days. The shuffled version won.
A method for checking whether any price level is doing anything, and what it returned on 54,000 days of data.
Five outputs turned bullish at the April 2026 low in ES1!, twelve days and 305 points apart. A method for pricing your own definition of confirmation.
A twenty minute test that tells you whether stacked indicators are confirming each other or just repeating one opinion.
A climax mark on ETH that price ran 14% past, a touchdown mark seven weeks earlier that worked, and the replay test that proves neither has moved.
A breakdown mark in May that everything agreed with, a touchdown mark in July that nothing did, and the replay test anyone can run to check neither has moved.
No idea matches. Try another filter or word.
Videos in English, on our YouTube channel.
Past measurement described after the fact. Nothing here is a recommendation or an indication of future results.