Five patterns repeat across every crypto cycle: breakouts, squeezes, failed moves, liquidity hunts, and event-driven setups. This is the field guide version. What each one looks like on the chart, how the desk reads it with the inputs we actually have, one real case from the tape with dates attached, and what proves the read wrong.
These are patterns, not predictions. For how the desk scores coins, read the Methodology. When these setups show up in live calls, they get graded in public in the Call ledger.
Fig. 1: The structure of a breakout. Price compresses under a flat resistance line, then closes above it on rising volume. The retest holding is what makes it real.
Price holds under a flat resistance line for days or weeks while each pullback bottoms higher than the last. The trading range narrows. Then a candle closes above the line on volume at least half again the recent average, and the next candles hold above the line instead of falling back.
The retest is the confirmation. A breakout that never looks back is fine, but the classic version dips back to the old resistance, finds buyers there, and continues. That old ceiling turning into a floor is the structure you are looking for.
The desk checks the order book first: real bid depth under price on the Kraken level-2 book, not just a thin tape. Then perp funding: a breakout on crowded long funding is borrowed time, because the longs who are already in are the fuel and there is nobody left to buy. For BTC and ETH, ETF flow streaks matter: sustained inflows give a breakout legs, persistent outflows starve it. The Coinbase premium shows whether US spot is leading the bid.
Where real prediction markets exist, on XRP, BTC, ETH, SOL, DOGE and LINK, crowd upside odds get a look as a positioning check. Daily and 4-hour gauges should agree with the move, and the seven-day momentum should already be turning. See the Methodology for the full input list.
Dec 2024
XRP sat near fifty cents for most of 2024, capped by a long-term downtrend. After the November 5 US election and the news that SEC chair Gary Gensler would step down, XRP broke out, retested the broken trendline, and ran to about $2.90 in early December, its first real multi-year breakout since the 2021 cycle. Volume expanded on the break and stayed heavy for weeks. The structure held, so the desk read it as a genuine breakout, not a headline spike.
A close back under the breakout line within a few sessions, especially if volume fades on the retest. Then it was a failed move wearing a breakout costume. Flat or negative ETF flows on BTC or ETH during the break are another warning: price moving up without the flows that are supposed to drive it usually does not last.
Fig. 2: The structure of a squeeze. Price coils inside converging trendlines while one side piles into a crowded position, then breaks out of the triangle tip on volume.
Price coils inside converging trendlines: higher lows meeting lower highs, range tightening week by week. Meanwhile one side of the market gets crowded. Funding sits deeply negative or positive for an extended stretch as traders pile into the same bet. Then a sharp, near-vertical candle breaks out of the triangle tip on heavy volume, and the move keeps going as the crowded side is forced to cover.
The signature is speed. Squeezes do not grind, they launch. The first candle out of the coil is usually the largest in weeks, and follow-through comes from forced buying, not fresh conviction.
Funding is the key input here. The desk treats crowded positioning as a contrarian signal: deeply negative funding after a long slide is fuel for an upside squeeze, deeply positive funding after a long run is fuel for the reverse. The first candle out of the coil gets the order-book read, and crowd sentiment is checked for extremes. If ETF flows or whale prints line up behind the move, the desk trusts the continuation instead of calling it a one-candle wonder.
When setups like this appear in live calls, they are scored and graded in the Call ledger, so there is a public record of which squeezes held and which did not.
Jan 2023
After the FTX collapse in November 2022, bitcoin ground sideways under $17,000 with funding stuck deeply negative for weeks. Bears were piled in. In January 2023 it squeezed from about $16,500 to about $24,000 in a matter of weeks as shorts covered into real spot demand. The move was fast and vertical, textbook short-covering shape, and it marked the start of the 2023 recovery rather than fading back into the coil.
Price re-enters the triangle within a day or two of the break. A squeeze that falls back inside its own coil was just noise, and the coil keeps winding. Watch funding too: if the break happens but funding never resets, the crowded side was not actually flushed, and the move is built on air.
Fig. 3: The structure of a failed move. A candle or two closes past resistance, the break is called, then price is back inside the old range within a session or two.
A candle or two closes above resistance, or below support, and the break gets called on social media. Then price is back inside the old range within one to three sessions. The giveaways: the break happens on average or below-average volume, and the candle that drags price back inside is larger than the breakout candle.
Failed moves are the market's most expensive lesson. They look exactly like breakouts for the first day, which is why the desk waits for the follow-through instead of buying the first close.
The desk reads these as volume failures. A break without volume expansion is suspect by default. Then the funding check: did the longs who piled into the break get flushed on the return? A funding reset makes the next real break more trustworthy. For BTC and ETH, ETF flow direction is the tiebreaker: a breakout with no inflows behind it rarely survives. On coins with real prediction markets, XRP, BTC, ETH, SOL, DOGE and LINK, odds that never budged on the break are a tell that nobody with size believed it.
The desk logs failed levels so the same fake line does not get trusted twice. See the Methodology for how the score weights these inputs.
Jan 2024
The US spot bitcoin ETFs were approved on January 10, 2024. Bitcoin printed about $49,000 on January 11, headlines called the breakout, and within two weeks it was back under $39,000. The news was real and the breakout was not. It took until late February for a genuine trend to form, and that one had steady ETF inflows behind it. The textbook buy-the-rumor, sell-the-news fakeout.
The read flips back if price closes above the level again, this time with rising volume and funding that has reset instead of staying crowded. A failed move that re-breaks with real backing is just a delayed breakout. The difference is always the same: backing, in the form of volume and flows, the second time around.
Fig. 4: The structure of a liquidity hunt. Price dips under a well-watched support line, prints a long lower wick to take out stops, and closes back above within a session or two.
Price dips just under a well-watched support line or a prior swing low, prints a long lower wick, and closes back above within the same session or the next. Volume spikes on the wick, and the recovery candle is strong. The stops parked under the obvious line get taken out, the weak hands sell the bottom, and the market turns higher.
The critical detail is the close, not the wick. A wick below support with a close back above is a hunt. A close through the level with follow-through the next session is a genuine breakdown. The two look identical for about an hour.
Liquidation heatmaps are not part of the Pulse Score, but the desk checks them when stops cluster under a watched level, because that is where hunts aim. During the wick the level-2 book gets read: real bid absorption on the Kraken book as the wick prints says buyers are waiting, an empty book says the dip has further to go. Funding and whale prints, on all coins except XMR, round out the read.
Mar 2020
In the March 12 to 13, 2020 COVID crash, bitcoin wick-crashed to about $3,800 on some exchanges as a cascade of liquidations flushed late longs, then reclaimed much higher within days and started a new uptrend. A deeper, nastier cousin of the same structure. The cautionary side: sometimes the wick never recovers. LUNA in May 2022 showed what a liquidity cascade looks like when there is no bid underneath, with UST losing its peg on May 9 and LUNA falling from about $80 to near zero within days.
The sweep fails to recover. If price closes under the swept level and stays there the next session, with funding still crowded and no bid absorption in the book, it was not a hunt. It was distribution breaking down, and the old support is now resistance.
Fig. 5: The structure of an event-driven setup. Price sits flat into a known event, prints a sharp impulse candle, then either consolidates above the old level or fades the whole move.
Price goes flat ahead of a known event: a halving, an ETF decision, a token unlock, a network upgrade. The event hits, a sharp impulse candle prints in the first hour, and then price does one of two things. It either consolidates above the pre-event level, meaning the event was bought, or it fades the entire move, meaning it was sold. The first hour's candle and the first daily close decide which one you are looking at.
The positioning into the event matters more than the event itself. When everyone is already long going in, there is nobody left to buy the news.
Into the event the desk watches crowd odds where real prediction markets exist, on XRP, BTC, ETH, SOL, DOGE and LINK, plus crowd net sentiment for one-sided positioning. For BTC and ETH events, ETF flow streaks and for BTC, ETH and SOL, CME positioning show whether institutional money is leaning the same way. The Coinbase premium shows whether US spot is leading the move or just following it. Full input list on the Methodology page.
Sep 2022
Ethereum's Merge on September 15, 2022 was the most anticipated upgrade in crypto history. ETH ran to about $1,780 in the days before it, then faded to about $1,300 within days after. The event was a flawless technical success and a classic sell-the-news trade, because everyone who wanted to own the Merge had bought it weeks earlier. The lesson the desk keeps: events are catalysts for positioning, not timers for price. The April 2024 bitcoin halving behaved the same way, flat into the event, with the real move arriving months later on ETF demand.
The post-event consolidation breaks the pre-event range in the opposite direction, with volume and backing from ETF flows or a sentiment reversal. Then the event's script is done and the tape has moved on. The read was about the event window; once price leaves the range, it is a new setup.
Every input the desk uses, its weight, and exactly what it covers. Coins without a feed never get a faked signal.
Calls in the wildEvery call graded on a 24-hour window, win or miss, in the open. See how setups from this page played out when real money was on the line.
Field guide authored Sep 25, 2026. Not financial advice.