PATTERN+ PREDICTIONS
Try Pro freeLog in
Updated

Patterns · every kind of event we measure · 9,938 past events

Every pattern we measure

Each kind of event, from unlocks and listings to Fed and CPI days, funding extremes, liquidations, ETF flows and Fear & Greed extremes, against the same coins' ordinary days 1, 7 and 30 days later. Colour only where the result is clear; a kind with too few cases says so. Past patterns, not financial advice.

as of 42 kinds of event40 with cases9 with a clear edge876 patterns109 this week34 coming up

Clearest edges: Delistings −30 pts the next dayExtreme fear readings −29 pts the next dayKorean discounts −26 pts the next dayHacks −17 pts the next day

The atlas · 42 kinds against their coins' ordinary days

Which events move price, and when

Each cell is the share of past cases higher that many days later minus the share of the same coins' ordinary days, in points; its bar is the 95% range. Green or red only when the whole range is clear of zero and both halves of history agree.

Listings and delistings

2 clear
  • DELISTDelistings134 cases
    −30−37…−19
    −24−31…−12
    −1−15…+18
  • PERPSPerp listings387 cases
    +50…+11
    −1−7…+4
    −8−13…−2
  • LISTINGSpot listings233 cases
    +5−4…+14
    −9−17…+1
    −9−18…0

Token supply

  • UNLOCKUnlocks249 cases · likely on 1
    0−7…+8
    +1−6…+9
    +6−1…+14
  • BURNToken burns20 cases
    −2−22…+18
    −12−30…+9
    +2−19…+21
  • HALVINGHalvings5 cases
    few5/20 days
    few5/20 days
    few5/20 days

US economic releases

1 clear
  • RETAILUS retail sales releases101 cases · on BTC · likely on 25
    −11−20…−1
    −1−11…+8
    −1−11…+8
  • PPIUS PPI releases100 cases · on BTC · likely on 71
    −3−12…+7
    −1−11…+9
    +2−8…+11
  • JOBSUS jobs reports96 cases · on BTC · proven on 1
    −1−11…+9
    +2−8…+12
    0−10…+10
  • PCEUS PCE inflation releases93 cases · on BTC · likely on 47
    0−10…+10
    +1−9…+11
    +2−8…+12
  • CPIUS CPI releases70 cases · on BTC · likely on 145
    +10−1…+21
    +4−8…+15
    0−12…+11
  • FEDFed rate decisions64 cases · on BTC · proven on 1
    +3−9…+15
    +1−11…+13
    0−12…+12
  • GDPUS GDP first estimates32 cases · on BTC · likely on 25
    +7−10…+22
    +6−11…+21
    +7−10…+22

Stocks, volatility and options

1 clear

Mood and attention

1 clear
  • FEARExtreme fear readings21 cases · on BTC · likely on 48
    −29−42…−8
    −19−36…+3
    −13−31…+9
  • ATTENTIONWikipedia attention spikes152 cases · likely on 2
    0−8…+9
    0−8…+9
    +3−6…+11
  • GREEDExtreme greed readings6 cases · on BTC
    few6/20 days
    few6/20 days
    few6/20 days

Money flows and premiums

1 clear

Leverage

  • FUNDING LOWFunding extreme lows1,854 cases · likely on 5
    −4−9…+1
    −3−8…+2
    +4−1…+9
  • FUNDING HIGHFunding extreme highs579 cases · likely on 1
    −5−11…+1
    −5−12…+1
    −4−10…+3
  • SHORTS SQUEEZEDShort liquidation spikes3 cases
    few2/20 days
    few2/20 days
    few2/20 days

No measured case yet: Long liquidation spikes

Protocols and networks

2 clear
  • USAGE UPUsage surges598 cases · likely on 4
    +40…+9
    +2−3…+7
    +7+3…+12
  • TVL JUMPTVL surges416 cases · likely on 2
    −4−9…+1
    −3−8…+3
    −6−11…−1
  • VOTEGovernance votes1,310 cases · likely on 1
    0−3…+3
    0−3…+3
    +1−2…+5
  • USAGE DOWNUsage slumps506 cases · likely on 5
    −2−7…+3
    −3−8…+2
    −3−8…+2
  • TVL DROPTVL crashes419 cases
    −2−8…+3
    −4−10…+1
    −6−11…0
  • DIFFICULTYBitcoin difficulty adjustments208 cases · on BTC · likely on 32
    −6−12…+1
    −1−8…+6
    0−7…+7
  • UPGRADENetwork upgrades16 cases
    few16/20 days
    few16/20 days
    few16/20 days

Hacks and funding rounds

1 clear
  • HACKHacks117 cases
    −17−25…−8
    −10−18…−1
    −3−11…+7

No measured case yet: Funding rounds

Every number in the atlas, as a table
Each kind of event against the same coins' ordinary days, counted by day; edges and ranges in points, as of Nov 19, 2026
EventAfterCases · daysHigher · ordinaryEdge95% rangeOlder · newer halfResult
Spot listings1 day233 · 11953% · 48%+5−4…+14+7 · +4no clear edge
7 days230 · 11739% · 48%−9−17…+1−7 · −10no clear edge
30 days212 · 10536% · 45%−9−18…0−6 · −12no clear edge
Perp listings1 day387 · 30055% · 49%+50…+11+8 · +3no clear edge
7 days382 · 29848% · 49%−1−7…+4−4 · +1no clear edge
30 days370 · 28840% · 48%−8−13…−2−9 · −6higher less often than on ordinary days
Delistings1 day134 · 6015% · 45%−30−37…−19−32 · −29higher less often than on ordinary days
7 days130 · 5917% · 41%−24−31…−12−26 · −22higher less often than on ordinary days
30 days48 · 2830% · 31%−1−15…+18−4 · +2no clear edge
Unlocks1 day249 · 16844% · 43%0−7…+80 · +1no clear edge
7 days246 · 16541% · 39%+1−6…+9−2 · +5no clear edge
30 days231 · 15537% · 31%+6−1…+14−1 · +12no clear edge
Hacks1 day117 · 11129% · 46%−17−25…−8−15 · −19higher less often than on ordinary days
7 days116 · 11033% · 43%−10−18…−1−2 · −18no clear edge
30 days109 · 10436% · 38%−3−11…+7−1 · −5no clear edge
Fed rate decisions1 day64 · 6456% · 53%+3−9…+15+5 · +1no clear edge
7 days64 · 6455% · 54%+1−11…+13+8 · −6no clear edge
30 days63 · 6356% · 55%0−12…+12+5 · −5no clear edge
US CPI releases1 day70 · 7063% · 52%+10−1…+21+4 · +17no clear edge
7 days70 · 7057% · 53%+4−8…+15−11 · +18no clear edge
30 days69 · 6955% · 55%0−12…+11−6 · +5no clear edge
US jobs reports1 day96 · 9652% · 53%−1−11…+9+4 · −6no clear edge
7 days96 · 9655% · 53%+2−8…+12−2 · +6no clear edge
30 days95 · 9555% · 55%0−10…+10−6 · +6no clear edge
US PPI releases1 day100 · 10050% · 53%−3−12…+7−4 · −1no clear edge
7 days99 · 9953% · 53%−1−11…+9−6 · +4no clear edge
30 days99 · 9957% · 55%+2−8…+110 · +4no clear edge
US PCE inflation releases1 day93 · 9353% · 53%0−10…+10+5 · −5no clear edge
7 days93 · 9355% · 54%+1−9…+11+15 · −12no clear edge
30 days92 · 9258% · 55%+2−8…+12−3 · +8no clear edge
US GDP first estimates1 day32 · 3259% · 53%+7−10…+22+21 · −8no clear edge
7 days32 · 3259% · 53%+6−11…+21+21 · −8no clear edge
30 days31 · 3161% · 54%+7−10…+22+6 · +9no clear edge
US retail sales releases1 day101 · 10142% · 53%−11−20…−1−16 · −6higher less often than on ordinary days
7 days100 · 10052% · 53%−1−11…+8−12 · +10no clear edge
30 days100 · 10054% · 55%−1−11…+80 · −2no clear edge
Bitcoin options expiries1 day94 · 9452% · 53%−1−11…+9+8 · −9no clear edge
7 days94 · 9456% · 53%+3−7…+13+10 · −4no clear edge
30 days93 · 9356% · 55%+1−9…+10−1 · +2no clear edge
Big stablecoin prints1 day134 · 13456% · 55%+1−7…+10+1 · +1no clear edge
7 days134 · 13455% · 56%−1−10…+7−4 · +1no clear edge
30 days134 · 13453% · 61%−8−17…0−10 · −6no clear edge
Bitcoin difficulty adjustments1 day208 · 20847% · 53%−6−12…+1−5 · −7no clear edge
7 days207 · 20753% · 53%−1−8…+6+2 · −4no clear edge
30 days206 · 20655% · 55%0−7…+7−2 · +3no clear edge
Extreme fear readings1 day21 · 2124% · 53%−29−42…−8−34 · −24higher less often than on ordinary days
7 days20 · 2035% · 54%−19−36…+3−24 · −13no clear edge
30 days20 · 2040% · 53%−13−31…+9+9 · −34no clear edge
Extreme greed readings1 day6 · 6—too few days
7 days6 · 6—too few days
30 days6 · 6—too few days
Big S&P 500 down days1 day56 · 5657% · 52%+5−8…+17+10 · +1no clear edge
7 days56 · 5648% · 52%−4−16…+9−1 · −6no clear edge
30 days55 · 5564% · 50%+140…+25+12 · +15higher more often than on ordinary days
Big S&P 500 up days1 day47 · 4753% · 51%+2−12…+16−5 · +9no clear edge
7 days47 · 4745% · 51%−6−19…+8−13 · +1no clear edge
30 days46 · 4650% · 47%+3−11…+17+3 · +2no clear edge
VIX spikes1 day22 · 2259% · 55%+5−16…+22+9 · 0no clear edge
7 days22 · 2241% · 56%−15−33…+5−9 · −21no clear edge
30 days21 · 2157% · 59%−2−23…+16−4 · 0no clear edge
Bitcoin implied volatility spikes1 day9 · 9—too few days
7 days9 · 9—too few days
30 days9 · 9—too few days
Bitcoin implied volatility lows1 day22 · 2264% · 50%+13−8…+30+4 · +22no clear edge
7 days22 · 2259% · 50%+9−11…+27+14 · +4no clear edge
30 days22 · 2241% · 48%−8−25…+13−16 · +1no clear edge
Coinbase premium spikes1 day43 · 4363% · 52%+11−4…+24+6 · +16no clear edge
7 days43 · 4360% · 52%+9−6…+22+2 · +15no clear edge
30 days43 · 4365% · 53%+12−3…+24+8 · +15no clear edge
Coinbase discounts1 day29 · 2955% · 51%+4−13…+21+5 · +3no clear edge
7 days29 · 2952% · 52%0−17…+17+5 · −5no clear edge
30 days28 · 2854% · 52%+1−16…+18−4 · +7no clear edge
Governance votes1 day1310 · 89549% · 49%0−3…+3+1 · 0no clear edge
7 days1307 · 89446% · 46%0−3…+3+1 · 0no clear edge
30 days1287 · 88443% · 41%+1−2…+5−3 · +6no clear edge
Funding extreme lows1 day1854 · 36642% · 46%−4−9…+1−3 · −6no clear edge
7 days1804 · 36040% · 42%−3−8…+20 · −5no clear edge
30 days1693 · 33740% · 36%+4−1…+9+2 · +6no clear edge
Funding extreme highs1 day579 · 22242% · 48%−5−11…+1−5 · −6no clear edge
7 days571 · 21842% · 47%−5−12…+1−1 · −10no clear edge
30 days532 · 19941% · 45%−4−10…+3+3 · −11no clear edge
Network upgrades1 day16 · 16—too few days
7 days16 · 16—too few days
30 days16 · 16—too few days
Halvings1 day5 · 5—too few days
7 days5 · 5—too few days
30 days5 · 5—too few days
Token burns1 day20 · 2050% · 52%−2−22…+18+10 · −14no clear edge
7 days20 · 2040% · 52%−12−30…+9−11 · −13no clear edge
30 days20 · 2055% · 53%+2−19…+21+13 · −8no clear edge
Big spot ETF inflow days1 day17 · 14—too few days
7 days17 · 14—too few days
30 days16 · 13—too few days
Big spot ETF outflow days1 day18 · 17—too few days
7 days18 · 17—too few days
30 days16 · 15—too few days
Short liquidation spikes1 day3 · 2—too few days
7 days3 · 2—too few days
30 days3 · 2—too few days
TVL crashes1 day419 · 31545% · 47%−2−8…+3−2 · −2no clear edge
7 days412 · 30941% · 45%−4−10…+1−8 · −1no clear edge
30 days393 · 29534% · 39%−6−11…0−6 · −5no clear edge
TVL surges1 day416 · 35444% · 48%−4−9…+1−4 · −4no clear edge
7 days414 · 35344% · 47%−3−8…+3−4 · −1no clear edge
30 days400 · 34137% · 43%−6−11…−1−11 · −2higher less often than on ordinary days
Korean premium spikes1 day133 · 10246% · 44%+2−7…+12−5 · +9no clear edge
7 days124 · 9830% · 37%−6−15…+3−13 · 0no clear edge
30 days97 · 8332% · 22%+90…+20−6 · +24no clear edge
Korean discounts1 day64 · 5718% · 44%−26−34…−14−36 · −16higher less often than on ordinary days
7 days62 · 5533% · 37%−4−15…+9−7 · −1no clear edge
30 days53 · 4633% · 23%+9−2…+24−6 · +25no clear edge
Usage surges1 day598 · 45852% · 47%+40…+9+4 · +5no clear edge
7 days587 · 45247% · 45%+2−3…+7+2 · +2no clear edge
30 days558 · 43248% · 41%+7+3…+12+7 · +8higher more often than on ordinary days
Usage slumps1 day506 · 39145% · 47%−2−7…+3−2 · −2no clear edge
7 days502 · 38742% · 45%−3−8…+2−7 · +1no clear edge
30 days475 · 37138% · 41%−3−8…+2−7 · +1no clear edge
Wikipedia attention spikes1 day152 · 13649% · 49%0−8…+9+3 · −2no clear edge
7 days152 · 13649% · 49%0−8…+9−1 · +2no clear edge
30 days142 · 12751% · 48%+3−6…+11+8 · −2no clear edge

Source: every measured event on record against its coin's own daily closes · fetched

Happening now and coming up

Events of the last 7 days and the next 30 that fit a pattern, each with that pattern's share higher in its past cases. A horizon with fewer than 8 cases shows no number.

Coming up · next 30 days · 34 events · the nearest 12 · the calendar →

  1. VOTE tomorrow

    UNI Uniswap

    1d50%7d45%30d38%higher, in past cases like it Governance votes of mid-cap tokens outside the big categories 273 cases · ⚠ flipped
  2. VOTE tomorrow

    AAVE Aave

    1d51%7d47%30d45%higher, in past cases like it Governance votes of mid-cap L1 and L2 tokens 550 cases
  3. VOTE in 2 days

    UNI Uniswap

    1d50%7d45%30d38%higher, in past cases like it Governance votes of mid-cap tokens outside the big categories 273 cases · ⚠ flipped
  4. VOTE in 2 days

    ENS Ethereum Name Service

    1d50%7d48%30d43%higher, in past cases like it Governance votes of small-cap tokens outside the big categories 256 cases · ⚠ flipped
  5. USAGE DOWN in 2 days

    ENA Ethena

    1d45%7d43%30d31%higher, in past cases like it Usage slumps of mid-cap tokens 98 cases
  6. USAGE DOWN in 2 days

    LINK Chainlink

    1d79%7d53%30d74%higher, in past cases like it Usage slumps of large-cap L1 and L2 tokens 19 cases
  7. VOTE in 2 days

    COMP Compound

    1d50%7d48%30d43%higher, in past cases like it Governance votes of small-cap tokens outside the big categories 256 cases · ⚠ flipped
  8. UNLOCK in 3 days

    LMTS Limitless

    1d49%7d41%30d30%higher, in past cases like it Unlocks of micro-cap tokens outside the big categories over 2% of supply 102 cases · ⚠ flipped
  9. UNLOCK in 3 days

    SKR Seeker

    1d38%7d48%30d37%higher, in past cases like it Unlocks of small-cap tokens outside the big categories of 0.5-2% of supply 29 cases
  10. VOTE in 4 days

    LDO Lido DAO

    1d50%7d48%30d43%higher, in past cases like it Governance votes of small-cap tokens outside the big categories 256 cases · ⚠ flipped
  11. VOTE in 5 days

    GNO Gnosis

    1d50%7d48%30d43%higher, in past cases like it Governance votes of small-cap tokens outside the big categories 256 cases · ⚠ flipped
  12. UNLOCK in 6 days

    TMX TermMax

    1d49%7d41%30d30%higher, in past cases like it Unlocks of micro-cap tokens outside the big categories over 2% of supply 102 cases · ⚠ flipped

What the price did after each kind

Each fan is what the price did after one kind of event, from the close before it to 1, 7 and 30 days later: the median, the middle half of its cases and 8 in 10 of them, every fan on the same scale. 37 of the 42 kinds have enough measured cases to draw one as of Nov 19, 2026.

The detailsevery kind in words, its patterns, and how it is counted

Kind by kind876 patterns

Every pattern, kind by kind

as of · the words behind each row of the atlas

Open a kind for its answer in words, the moves after its cases and its biggest patterns. Every kind has its own page with every past case; the search above finds any of the 876 patterns.

Listings and delistings

DELISTDelistings43 patterns−30 pts 1 day

After delistings, the price was higher the next day in 15% of 134 cases on 60 separate days, against 45% of the same coins' ordinary days: −30 points (95% range −37 to −19; older half −32, newer half −29). The edge is clear 7 days later too.

Median move −12.7% after 1 day, −16.3% after 7 days, −13.7% after 30 days; after 30 days, half the cases −30.4% to +1.3% and 8 in 10 −46.7% to +21.0%.

Explained and tested: What happens to a coin the day after a delisting notice?

25 more: find them with the search above, or from the page of delistings

PERPSPerp listings150 patterns−8 pts 30 days

After perp listings, the price was higher 30 days later in 40% of 370 cases on 288 separate days, against 48% of the same coins' ordinary days: −8 points (95% range −13 to −2; older half −9, newer half −6).

Median move −0.6% after 1 day, −7.1% after 7 days, −16.9% after 30 days; after 30 days, half the cases −41.6% to +18.5% and 8 in 10 −60.6% to +65.5%.

Explained and tested: What happens to a coin in the 30 days after a perpetual futures listing?

133 more: find them with the search above, or from the page of perp listings

LISTINGSpot listings127 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (233 cases on 119 separate days).

Median move −2.4% after 1 day, −8.4% after 7 days, −15.3% after 30 days; after 30 days, half the cases −40.0% to +15.4% and 8 in 10 −57.2% to +58.4%.

Explained and tested: What happens to a coin in the 30 days after an exchange listing?

110 more: find them with the search above, or from the page of spot listings

Token supply

UNLOCKUnlocks84 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (249 cases on 168 separate days). Coin by coin, it is a likely driver on 1 coin.

Median move −0.4% after 1 day, −1.9% after 7 days, −6.2% after 30 days; after 30 days, half the cases −22.1% to +8.0% and 8 in 10 −36.2% to +32.3%.

Explained and tested: What happens to a coin in the 30 days after a token unlock?

66 more: find them with the search above, or from the page of unlocks

BURNToken burns12 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (20 cases).

Median move −0.1% after 1 day, −0.9% after 7 days, +3.1% after 30 days; after 30 days, half the cases −6.5% to +13.9% and 8 in 10 −22.1% to +21.2%.

HALVINGHalvings5 cases

Too few to say: 5 separate days of cases, and a call needs 20.

US economic releases

RETAILUS retail sales releases3 patterns−11 pts 1 day

After US retail sales releases, the price was higher the next day in 42% of 101 cases, against 53% of the same coins' ordinary days: −11 points (95% range −20 to −1; older half −16, newer half −6). Measured on Bitcoin's own price. Coin by coin, it is a likely driver on 25 coins.

Median move −0.6% after 1 day, +0.3% after 7 days, +1.8% after 30 days; after 30 days, half the cases −6.2% to +18.0% and 8 in 10 −20.6% to +29.1%.

Explained and tested: What does Bitcoin do the day after a US retail sales release?

PPIUS PPI releases3 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (100 cases). Measured on Bitcoin's own price. Coin by coin, it is a likely driver on 71 coins.

Median move 0.0% after 1 day, +0.4% after 7 days, +2.2% after 30 days; after 30 days, half the cases −8.7% to +16.6% and 8 in 10 −19.8% to +31.0%.

JOBSUS jobs reports3 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (96 cases). Measured on Bitcoin's own price. Coin by coin, it is a proven driver on 1 coin and a likely one on 48 more coins.

Median move +0.1% after 1 day, +0.8% after 7 days, +2.2% after 30 days; after 30 days, half the cases −6.9% to +14.3% and 8 in 10 −18.0% to +30.5%.

Explained and tested: What does Bitcoin do in the week after a US jobs report?

PCEUS PCE inflation releases3 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (93 cases). Measured on Bitcoin's own price. Coin by coin, it is a likely driver on 47 coins.

Median move +0.2% after 1 day, +0.5% after 7 days, +2.5% after 30 days; after 30 days, half the cases −7.6% to +13.7% and 8 in 10 −17.9% to +29.4%.

CPIUS CPI releases3 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (70 cases). Measured on Bitcoin's own price. Coin by coin, it is a likely driver on 145 coins.

Median move +1.0% after 1 day, +1.1% after 7 days, +1.9% after 30 days; after 30 days, half the cases −11.3% to +14.5% and 8 in 10 −18.5% to +26.3%.

Explained and tested: What does Bitcoin do the day after a US CPI release?

FEDFed rate decisions3 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (64 cases). Measured on Bitcoin's own price. Coin by coin, it is a proven driver on 1 coin and a likely one on 87 more coins.

Median move +0.4% after 1 day, +0.7% after 7 days, +3.0% after 30 days; after 30 days, half the cases −8.5% to +13.1% and 8 in 10 −19.1% to +32.7%.

Explained and tested: What does Bitcoin do the day after a Fed rate decision?

GDPUS GDP first estimates3 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (32 cases). Measured on Bitcoin's own price. Coin by coin, it is a likely driver on 25 coins.

Median move +0.4% after 1 day, +1.4% after 7 days, +3.3% after 30 days; after 30 days, half the cases −8.2% to +13.9% and 8 in 10 −17.7% to +28.6%.

Stocks, volatility and options

S&P DROPBig S&P 500 down days3 patterns+14 pts 30 days

After big S&P 500 down days, the price was higher 30 days later in 64% of 55 cases, against 50% of the same coins' ordinary days: +14 points (95% range 0 to +25; older half +12, newer half +15). Measured on Bitcoin's own price. Coin by coin, it is a proven driver on 4 coins and a likely one on 129 more coins.

Median move +0.4% after 1 day, −0.1% after 7 days, +4.0% after 30 days; after 30 days, half the cases −6.3% to +15.0% and 8 in 10 −25.2% to +32.5%.

Explained and tested: What does Bitcoin do in the 30 days after a big S&P 500 down day?

OPEXBitcoin options expiries3 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (94 cases). Measured on Bitcoin's own price. Coin by coin, it is a likely driver on 77 coins.

Median move +0.2% after 1 day, +0.9% after 7 days, +2.4% after 30 days; after 30 days, half the cases −5.4% to +15.3% and 8 in 10 −16.4% to +31.2%.

S&P JUMPBig S&P 500 up days3 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (47 cases). Measured on Bitcoin's own price. Coin by coin, it is a likely driver on 55 coins.

Median move +0.4% after 1 day, −1.1% after 7 days, −0.5% after 30 days; after 30 days, half the cases −13.0% to +14.3% and 8 in 10 −23.5% to +31.6%.

VOL LOWBitcoin implied volatility lows3 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (22 cases). Measured on Bitcoin's own price. Coin by coin, it is a proven driver on 1 coin and a likely one on 29 more coins.

Median move +0.4% after 1 day, +0.4% after 7 days, −1.1% after 30 days; after 30 days, half the cases −7.2% to +9.6% and 8 in 10 −13.3% to +24.6%.

VIX SPIKEVIX spikes2 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (22 cases). Measured on Bitcoin's own price. Coin by coin, it is a likely driver on 9 coins.

Median move +0.3% after 1 day, −0.6% after 7 days, +0.7% after 30 days; after 30 days, half the cases −5.6% to +5.5% and 8 in 10 −10.7% to +18.5%.

VOL SPIKEBitcoin implied volatility spikes1 pattern

Too few to say: 9 separate days of cases, and a call needs 20. Measured on Bitcoin's own price. Coin by coin, it is a likely driver on 176 coins.

Median move +3.6% after 1 day, +4.8% after 7 days, +4.9% after 30 days; after 30 days, half the cases −3.0% to +7.6% and 8 in 10 −8.4% to +12.2%.

Mood and attention

FEARExtreme fear readings2 patterns−29 pts 1 day

After extreme fear readings, the price was higher the next day in 24% of 21 cases, against 53% of the same coins' ordinary days: −29 points (95% range −42 to −8; older half −34, newer half −24). Measured on Bitcoin's own price. Coin by coin, it is a likely driver on 48 coins.

Median move −3.5% after 1 day, −2.6% after 7 days, −2.6% after 30 days; after 30 days, half the cases −7.5% to +10.8% and 8 in 10 −13.2% to +33.0%.

ATTENTIONWikipedia attention spikes41 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (152 cases on 136 separate days). Coin by coin, it is a likely driver on 2 coins.

Median move 0.0% after 1 day, 0.0% after 7 days, 0.0% after 30 days; after 30 days, half the cases −14.9% to +16.8% and 8 in 10 −32.0% to +56.4%.

Explained and tested: What happens to a coin in the 30 days after a Wikipedia attention spike?

33 more: find them with the search above, or from the page of Wikipedia attention spikes

GREEDExtreme greed readings6 cases

Too few to say: 6 separate days of cases, and a call needs 20. Measured on Bitcoin's own price.

Money flows and premiums

KOREA DISCOUNTKorean discounts16 patterns−26 pts 1 day

After Korean discounts, the price was higher the next day in 18% of 64 cases on 57 separate days, against 44% of the same coins' ordinary days: −26 points (95% range −34 to −14; older half −36, newer half −16).

Median move −3.3% after 1 day, −6.4% after 7 days, −8.9% after 30 days; after 30 days, half the cases −22.7% to +11.0% and 8 in 10 −33.3% to +20.3%.

Explained and tested: What happens to a coin the day after a Korean discount?

7 more: find them with the search above, or from the page of Korean discounts

PRINTBig stablecoin prints3 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (134 cases). Measured on Bitcoin's own price. Coin by coin, it is a likely driver on 81 coins.

Median move +0.6% after 1 day, +0.8% after 7 days, +1.0% after 30 days; after 30 days, half the cases −7.7% to +12.2% and 8 in 10 −19.0% to +30.7%.

Explained and tested: What does Bitcoin do in the 30 days after a big stablecoin print?

KOREA PREMIUMKorean premium spikes18 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (133 cases on 102 separate days). Coin by coin, it is a likely driver on 1 coin.

Median move −0.6% after 1 day, −3.8% after 7 days, −7.9% after 30 days; after 30 days, half the cases −19.4% to +6.0% and 8 in 10 −31.8% to +23.9%.

Explained and tested: What happens to a coin in the 30 days after a Korean premium spike?

10 more: find them with the search above, or from the page of Korean premium spikes

US PREMIUMCoinbase premium spikes3 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (43 cases). Measured on Bitcoin's own price. Coin by coin, it is a likely driver on 52 coins.

Median move +0.7% after 1 day, +1.1% after 7 days, +3.4% after 30 days; after 30 days, half the cases −6.8% to +12.9% and 8 in 10 −12.9% to +27.7%.

US DISCOUNTCoinbase discounts3 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (29 cases). Measured on Bitcoin's own price. Coin by coin, it is a likely driver on 82 coins.

Median move +0.6% after 1 day, +1.0% after 7 days, +1.8% after 30 days; after 30 days, half the cases −7.0% to +10.3% and 8 in 10 −19.5% to +16.5%.

ETF OUTBig spot ETF outflow days8 patterns

Too few to say: 17 separate days of cases, and a call needs 20. Coin by coin, it is a proven driver on 5 coins and a likely one on 217 more coins.

Median move −0.8% after 1 day, −3.6% after 7 days, −17.6% after 30 days; after 30 days, half the cases −22.7% to −10.9% and 8 in 10 −24.6% to +0.8%.

ETF INBig spot ETF inflow days8 patterns

Too few to say: 14 separate days of cases, and a call needs 20. Coin by coin, it is a likely driver on 71 coins.

Median move +0.2% after 1 day, +1.7% after 7 days, −2.9% after 30 days; after 30 days, half the cases −9.0% to +1.5% and 8 in 10 −20.9% to +13.5%.

Leverage

FUNDING LOWFunding extreme lows55 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (1,854 cases on 366 separate days). Coin by coin, it is a likely driver on 5 coins.

Median move −1.1% after 1 day, −2.5% after 7 days, −5.7% after 30 days; after 30 days, half the cases −21.8% to +8.8% and 8 in 10 −36.0% to +28.3%.

Explained and tested: What happens to a coin in the 30 days after an extreme low in funding?

47 more: find them with the search above, or from the page of funding extreme lows

FUNDING HIGHFunding extreme highs37 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (579 cases on 222 separate days). Coin by coin, it is a likely driver on 1 coin.

Median move −0.7% after 1 day, −2.8% after 7 days, −4.3% after 30 days; after 30 days, half the cases −26.0% to +18.7% and 8 in 10 −47.5% to +58.5%.

Explained and tested: What happens to a coin in the week after an extreme high in funding?

28 more: find them with the search above, or from the page of funding extreme highs

SHORTS SQUEEZEDShort liquidation spikes3 cases

Too few to say: 2 separate days of cases, and a call needs 20.

Protocols and networks

USAGE UPUsage surges46 patterns+7 pts 30 days

After usage surges, the price was higher 30 days later in 48% of 558 cases on 432 separate days, against 41% of the same coins' ordinary days: +7 points (95% range +3 to +12; older half +7, newer half +8). Coin by coin, it is a likely driver on 4 coins.

Median move −0.2% after 1 day, −0.9% after 7 days, −1.3% after 30 days; after 30 days, half the cases −19.3% to +19.8% and 8 in 10 −34.1% to +48.8%.

Explained and tested: What happens to a coin in the 30 days after a usage surge?

38 more: find them with the search above, or from the page of usage surges

TVL JUMPTVL surges33 patterns−6 pts 30 days

After TVL surges, the price was higher 30 days later in 37% of 400 cases on 341 separate days, against 43% of the same coins' ordinary days: −6 points (95% range −11 to −1; older half −11, newer half −2). Coin by coin, it is a likely driver on 2 coins.

Median move −0.8% after 1 day, −2.4% after 7 days, −8.5% after 30 days; after 30 days, half the cases −27.0% to +9.9% and 8 in 10 −43.7% to +43.7%.

Explained and tested: What happens to a coin in the 30 days after a TVL surge?

25 more: find them with the search above, or from the page of TVL surges

VOTEGovernance votes42 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (1,310 cases on 895 separate days). Coin by coin, it is a likely driver on 1 coin.

Median move +0.2% after 1 day, −0.8% after 7 days, −4.3% after 30 days; after 30 days, half the cases −19.5% to +14.1% and 8 in 10 −30.7% to +35.8%.

Explained and tested: What happens to a coin in the 30 days after a governance vote?

29 more: find them with the search above, or from the page of governance votes

USAGE DOWNUsage slumps42 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (506 cases on 391 separate days). Coin by coin, it is a likely driver on 5 coins.

Median move −0.7% after 1 day, −2.6% after 7 days, −8.0% after 30 days; after 30 days, half the cases −24.0% to +5.8% and 8 in 10 −35.6% to +23.9%.

Explained and tested: What happens to a coin in the week after a usage slump?

34 more: find them with the search above, or from the page of usage slumps

TVL DROPTVL crashes30 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (419 cases on 315 separate days).

Median move −0.2% after 1 day, −1.7% after 7 days, −9.3% after 30 days; after 30 days, half the cases −23.9% to +4.5% and 8 in 10 −39.1% to +31.0%.

Explained and tested: What happens to a coin in the 30 days after a TVL crash?

22 more: find them with the search above, or from the page of TVL crashes

DIFFICULTYBitcoin difficulty adjustments3 patterns

No clear edge 1, 7 or 30 days later: no 95% range clears the same coins' ordinary days with both halves of history agreeing (208 cases). Measured on Bitcoin's own price. Coin by coin, it is a likely driver on 32 coins.

Median move −0.2% after 1 day, +0.2% after 7 days, +2.1% after 30 days; after 30 days, half the cases −8.6% to +13.0% and 8 in 10 −19.0% to +32.9%.

Explained and tested: What does Bitcoin do the day after a Bitcoin difficulty adjustment?

UPGRADENetwork upgrades12 patterns

Too few to say: 16 separate days of cases, and a call needs 20.

Median move +0.2% after 1 day, −0.8% after 7 days, −0.5% after 30 days; after 30 days, half the cases −19.8% to +37.3% and 8 in 10 −26.4% to +46.3%.

Hacks and funding rounds

HACKHacks22 patterns−17 pts 1 day

After hacks, the price was higher the next day in 29% of 117 cases on 111 separate days, against 46% of the same coins' ordinary days: −17 points (95% range −25 to −8; older half −15, newer half −19).

Median move −5.1% after 1 day, −6.8% after 7 days, −13.1% after 30 days; after 30 days, half the cases −29.6% to +7.8% and 8 in 10 −41.4% to +41.1%.

Explained and tested: What happens to a coin the day after a hack?

13 more: find them with the search above, or from the page of hacks

Source: every past event of these kinds and the price moves measured after it · fetched

How it is counted

How it is counted

as of · methodology

Every number here is a count of dated past cases, measured from the close before each event; it is not financial advice, and past patterns can stop working.

The atlas counts each kind of event against the same coins' ordinary days: for every past case, every window of the same length in the year before it on the same coin. A case needs 90 such days with prices, so one on a newer coin is left out of the atlas but still counts in its kind's fan: that is why the two counts can differ. Cases on one day count as one, so a crash that hits fifty coins is one day of evidence. A cell is coloured only when its whole 95% range is clear of zero and the older and the newer half of the days lean the same way; under 20 separate days it shows no number.

A fan is the moves after a kind's cases, from the close before the event to the close 1, 7 and 30 days later: the median, the middle half and 8 in 10 of them, each horizon drawn once 8 cases are measured that far. It shows how big the moves were, not a projection.

A pattern splits one kind by venue, market-cap size, sector, unlock size and Bitcoin's trend, and is kept once 8 past cases have a measured next-day move; its share higher comes with a 90% range on its own page. "Proven on a coin" and "likely on a coin" come from the nightly test of every coin on its own (what moves each coin). Full write-up on the methodology page.