Research is a work in progress. Estimated release: 1 December 2026. Our goal is to make it a search engine for trading setups in recorded crypto markets.
Binance USDT-MTAKEUSDT rose 84% on 29 July 2026
Picked after the rise, on one market.
Next stepWhat happened
Price opened the 16:00 hour at 0.02259 and reached a highest trade of 0.04148 in the 04:00 hour; the tested readings were taken at 15:00.
What stood out at 15:00
- Order book tilt0.8462Tested: 0.84 or higher
- Order book lopsidedness0.9019Tested: 0.9 or higher
- Liquidation bunching0.5848Tested: 0.58 or higher
15:00 always meets its own rules.
What the record says
No counts were saved to compare, only the summary text
Picked after the rise, on one market.
Evidence
Does it hold on other dates?
Nothing runs until you press Run.
Exactly what was counted3 conditions · 440 of 43,201 minutes · saved text
Saved summary text, as written 7 Aug 2026
Written automatically when this study was saved and kept as written; where a line differs from the counts above, the counts govern.
Minutes like this are common here: 440 of the 43,201 minutes searched (1%), about 14 a day. What followed was close to a coin flip: 240 rose and 199 fell over the next 30 minutes (439 of the 440 matches have a complete 30 minutes forward).
About 14 a day, on 30 of 31 days: spread through the record rather than one event.
The other reading times, as written
From 1 minute before
Minutes like this appeared 32 times in the 43,201 minutes searched (<0.1%). What followed leaned up: 18 of the 32 matches with a complete 30 minutes forward rose (56%).
About 1 a day, on 21 of 31 days: spread through the record rather than one event.
From 15 minutes before
Minutes like this appeared 36 times in the 43,180 minutes searched (<0.1%). What followed leaned up: 20 of the 36 matches with a complete 30 minutes forward rose (56%).
About 1.2 a day, on 23 of 31 days: spread through the record rather than one event.
Picked after the rise, on one market.
- Order book tilt 0.84 or higher. Size at the best bid against the best ask; above 0 means more bids, below 0 more asks.Book imbalance · top levels ≥ 0.84 · feature.book_imbalance
- Order book lopsidedness 0.9 or higher. How lopsided the sizes at the best bid and ask were, either way, against recent readings; low means unusually even.Book imbalance percentile ≥ 0.9 · feature.book_imbalance_pctrank
- Liquidation bunching 0.58 or higher. How bunched in time recent liquidations were; higher means more bunched, and it stays below 1.Cascade clustering ≥ 0.58 · feature.cascade_clustering_index
Read at 1 hour before the move.
Read at 15:00 UTC on 29 July, one hour before the selected 16:00 start.
TAKE only, 6 July to 4 August 2026 (UTC, ending 5 August 00:00); all 3 readings had values on 43,201 of those minutes.
Reference: every TAKE minute over the same dates, matches included; it is not a matched control group.
Outcome: the close exactly 30 minutes later, not the highest or lowest price along the way.
The 16:00 start was picked automatically after the move.
TAKEUSDT stored readings end 2026-08-05 00:00 UTC (exclusive). Prices and tick replay have separate coverage.
The feature manifest spans the lead-up. Individual minutes or inputs can still be unavailable.
This is not a similarity search. Each count matches the exact readings above and nothing else.
Picked in hindsight: the counts include every other time these readings appeared, not just this move.
Fires often arrive as clusters of consecutive minutes, and forward windows overlap. A cluster is one episode, not many independent trials.
If these counts read weaker than the move suggested, that is the tool working: the lead-up of a picked move always looks special until it is counted everywhere else.
Saved study q#1a24af9f @ dataset 9dad0852, counted 7 Aug 2026 02:52 UTC.
When the matches happened
Of the first 400 of 440 matches checked, 327 came before the rise began, 57 during it and 16 after it.
Neighbouring minutes overlap, so these are not 440 separate trials.
Picked after the rise, on one market.
Readings at five times before the rise24 hours to 1 minute before · 2 other times tried
Highlights follow fixed rules, not significance tests. Proposed conditions round values or use ranges; inspect them before running.
24 hours before · 28 Jul 16:00 UTC · 3 of 39 stood out
15-minute range against usual 0.43 or lowerLiquidation bunching 0.58 or higherMinutes to funding 1 or lowerShow the full readings
Reading Value at that minute Stored rule Compression · 15m feature.candle_compression_15m 0.4211 ≤ 0.43 Cascade clustering feature.cascade_clustering_index 0.5849 ≥ 0.58 Minutes to funding feature.mins_to_funding 0 ≤ 1 4 hours before · 29 Jul 12:00 UTC · 2 of 39 stood out
Liquidation bunching 0.58 or higherDistance from the day's average price -0.022 or lowerShow the full readings
Reading Value at that minute Stored rule Cascade clustering feature.cascade_clustering_index 0.5848 ≥ 0.58 VWAP distance · day feature.vwap_dist -0.02291 ≤ -0.022 1 hour before · 29 Jul 15:00 UTC · 3 of 39 stood out
Order book tilt 0.84 or higherOrder book lopsidedness 0.9 or higherLiquidation bunching 0.58 or higherShow the full readings
Reading Value at that minute Stored rule Book imbalance · top levels feature.book_imbalance 0.8462 ≥ 0.84 Book imbalance percentile feature.book_imbalance_pctrank 0.9019 ≥ 0.9 Cascade clustering feature.cascade_clustering_index 0.5848 ≥ 0.58 15 minutes before · 29 Jul 15:45 UTC · 4 of 39 stood out
15-minute range against usual 2.6 or higherLiquidation bunching 0.58 or higherMinutes to funding 15 or lowerOpen interest change rank 0.99 or higherShow the full readings
Reading Value at that minute Stored rule Compression · 15m feature.candle_compression_15m 2.646 ≥ 2.6 Cascade clustering feature.cascade_clustering_index 0.5848 ≥ 0.58 Minutes to funding feature.mins_to_funding 15 ≤ 15 OI velocity percentile feature.oi_velocity_pctrank 0.9944 ≥ 0.99 1 minute before · 29 Jul 15:59 UTC · 4 of 39 stood out
15-minute range against usual 2.6 or higherLiquidation bunching 0.58 or higherMinutes to funding 1 or lowerVolatility rank 0.92 or higherShow the full readings
Reading Value at that minute Stored rule Compression · 15m feature.candle_compression_15m 2.646 ≥ 2.6 Cascade clustering feature.cascade_clustering_index 0.5848 ≥ 0.58 Minutes to funding feature.mins_to_funding 1 ≤ 1 Realized vol percentile feature.realized_vol_pctrank 0.9268 ≥ 0.92
Two other reading times we tried 32 and 36 matches
1 minute before (15:59): 15-minute range against usual 2.6 or higher, liquidation bunching 0.58 or higher, volatility rank 0.92 or higher.
Compression · 15m ≥ 2.6 and Cascade clustering ≥ 0.58 and Realized vol percentile ≥ 0.92
32 matches.
Of the 32 matches, 20 came before the rise began, 9 during it and 3 after it.
15 minutes before (15:45): 15-minute range against usual 2.6 or higher, liquidation bunching 0.58 or higher, open interest change rank 0.99 or higher.
Compression · 15m ≥ 2.6 and Cascade clustering ≥ 0.58 and OI velocity percentile ≥ 0.99
36 matches.
Of the 36 matches, 26 came before the rise began, 7 during it and 3 after it.
Fewer than 30 results, so no rate is printed.
The saved study leads with 15:00 because it is the only set with at least 30 matches before the rise.
Picked after the rise, on one market.
Hourly candles around the risefull context, lead-up, move
The 16:00 start was picked after the move, and the hour before was already rising.
Hourly candles, UTC. The marked lines are the minutes the readings came from: 1 minute, 15 minutes, 1 hour, 4 hours and 24 hours before the selected start.
Other ways to test itsame dates
Same 3 conditions on TAKE, 6 June to 5 July 2026.
Each opens Review the test.
Running needs a free account and uses your allowance; you see the cost first.