# Market Flow Research

> Independent research on crypto market microstructure: second-by-second Binance order-book data for six coins since February 2026, 83 published studies and trading bots tested live. AI-assisted, human-reviewed. Research, not financial advice.

Every number in the articles comes from research logs, and losing results are published as they are. Many studies come with a research log (Markdown, Ukrainian) and a reproduction kit.

## Start here

- [How to Read This Site in Five Minutes — and Ask an AI About Every Study](https://marketflowresear.ch/articles/how-to-read-this-site-in-five-minutes/): This site holds 83 studies of second-by-second order-book data and the trading bots built from them.
- [Line 7: The Best of All Our Bots — Live Trading Report, Every Trade](https://marketflowresear.ch/articles/line-7-live-trading-report/): We have built and tested more than a dozen trading bots on the research published here.
- [AI digest — every study in one Markdown file](https://marketflowresear.ch/mfr-digest.md): titles, links, summaries and the key sentence of every section; made to be read by an AI in one go.

## The road to Bot 7

One question at a time: from zigzag legs to a network that reads 160 hours of order book as one canvas.

- [The Pair That Behaves the Same on Three Stretches of Data — How the Live Bot's Networks Were Chosen](https://marketflowresear.ch/articles/the-pair-that-behaves-the-same-on-three-stretches-of-data/): Ranking 440 buy-and-sell pairs by money on the month they were picked on gives a list of lucky pairs.
- [One Account, Two Networks — 440 Buy-and-Sell Pairs on Validation, and Why the Best Still Trails the Market](https://marketflowresear.ch/articles/one-account-two-networks-440-buy-and-sell-pairs-on-validation/): A buy network and a sell network were put on one $1000 account under one set of rules: one trade per 2.3% zigzag leg, take-profit 1.9%, stop 2.5%,….
- [Picking a Network by the Width of Its Good Zone, Not Its Best Point — 42 "Potentially Best" Networks](https://marketflowresear.ch/articles/picking-a-network-by-the-width-of-its-good-zone-not-its-best-point/): Every trained network becomes a signal through two knobs: the window W over which its probabilities are summed and the threshold θ that sum must reach.
- [Teaching the Network to See Tops — Sell Pivots on BTC, One Change and Eighteen Training Runs](https://marketflowresear.ch/articles/teaching-the-network-to-see-tops-sell-pivots-on-btc/): The bottom-finding network that powers our buy-only bot was rebuilt for the other side of the market with exactly one change: the label is now a….
- [When a "Fake" Buy Signal Is Really a Deeper Bottom — Rule 7.1, Tested on Validation and Unseen Days](https://marketflowresear.ch/articles/when-a-fake-buy-signal-is-really-a-deeper-bottom-rule-7-1/): Our buy-only bot throws away a buy signal if price has not risen 2.3% since the previous signal — a repeat call inside the same dip is treated as fake.
- [A Buy-Only Bot on Nine Days It Never Saw — Three Take-Profits and a Draw With Buy & Hold](https://marketflowresear.ch/articles/a-buy-only-bot-on-nine-days-it-never-saw/): The buy-only bot from the previous study had one weakness we wrote down plainly: its window, threshold and stop were chosen on the same 38….
- [From a Canvas of the Whole Leg to a Buy-Only Bot — What Survived, and What It Earned](https://marketflowresear.ch/articles/from-a-canvas-of-the-leg-to-a-buy-only-bot-what-survived/): The last study of the series put a whole leg — and 50 rows on either side — onto one canvas and asked a network to mark pivots and corrections on it.
- [How Full Is This Leg? A Network That Finally Beats the Counter — by Reading Its Last Rows](https://marketflowresear.ch/articles/how-full-is-this-leg-a-network-that-finally-beats-the-counter/): Give a network the unfinished leg — everything from the last confirmed pivot to now, as one fixed-size canvas — and ask how complete it is.
- [Where a Leg Really Ends — Four Events Earlier Than the Zigzag Says](https://marketflowresear.ch/articles/where-a-leg-really-ends-four-events-earlier-than-the-zigzag-says/): We went looking for a unique, universal signature of a pivot in 531 legs of six coins.
- [Can Our Networks' Outputs Become a Buy/Sell Signal? A Simple 1.2% Zigzag Says No](https://marketflowresear.ch/articles/can-the-networks-outputs-become-a-buy-sell-signal-a-simple-zigzag-says-no/): Our networks each say something about a leg: it is ending, it is 80% done, this is a reversal zone.
- [A Loss Ladder for Five Trained Networks — and the Biggest Win That Was a Dropout Bug](https://marketflowresear.ch/articles/a-loss-ladder-for-five-trained-networks-and-the-win-that-was-a-dropout-bug/): We took five approved networks from this series and tried to improve them without retraining: change the loss function between epochs, raise the….
- [A Correction and a Reversal Are One Event of Different Strength — What the Book Shows Inside a Leg](https://marketflowresear.ch/articles/correction-and-reversal-are-one-event-of-different-strength/): Every leg of a zigzag holds its own common level in the order book, and removing it would make the unusual events stand out — if only it could be….
- [The Book Sees the Edge of a Leg, Not Its End — Event by Event on BTC](https://marketflowresear.ch/articles/the-book-sees-the-edge-of-a-leg-not-its-end/): Event by event on BTC, every feature of our order-book class carries its place in the leg: a U-shaped profile that holds inside individual legs,….
- [The Order Book Along a Leg — Slow Legs, Thin Edges, and a Signal at the End That Beats Chance ×4.6](https://marketflowresear.ch/articles/the-order-book-along-a-leg-slow-legs-thin-edges-and-a-signal-at-the-end/): We laid the seven features of our best order-book class along 554 zigzag legs of six coins.
- [A Leg Moves at Its Edges — and You See the New One Only After It Starts](https://marketflowresear.ch/articles/a-leg-moves-at-its-edges-and-you-see-the-new-one-only-after-it-starts/): Split every 2.3% zigzag leg of six coins into tenths and measure where the price actually travels: 39% of the move happens in the first and last….
- [The ±0.15% Pivot Zone — Seventeen Ways to Describe an Event, One Honest Ceiling of 0.70](https://marketflowresear.ch/articles/the-pivot-zone-plus-minus-015-percent-seventeen-ways-to-describe-an-event/): Near every 2.3% pivot, price lingers within 0.15% of the turning level for a while — a median of 11 events, about 38 minutes, 2% of the whole series.
- [The Filter That Would Lift Precision From 51% to 93% — and Why We Cannot Build It Yet](https://marketflowresear.ch/articles/the-filter-that-would-lift-precision-to-93-percent-and-why-we-cannot-build-it/): Half of the reversal-confirmation signals are false, and they share one trait: they point into the middle of a leg.
- [Can a Memory Network Rebuild Context From the Raw Sequence? Not From the Order Book](https://marketflowresear.ch/articles/can-a-memory-network-rebuild-context-from-the-raw-sequence/): Trees won the reversal ladder with hand-made aggregates: five numbers about recent extremes over four look-back windows.
- [Pivot Confirmation on Three Unseen Weeks — Precision Holds, the Threshold Does Not](https://marketflowresear.ch/articles/pivot-confirmation-on-three-unseen-weeks/): The reversal ladder of the previous article was scored on a held-out coin — but in the same weeks it trained on.
- [A Reversal Is a Place in the Move, Not a Shape in the Book — Five Causal Numbers Beat 62](https://marketflowresear.ch/articles/a-reversal-is-a-place-in-the-move-not-a-shape-in-the-book/): The edge of a leg really is different from its middle — in 85% of 600 legs, by a factor of about 3.8, with a coin-flip control at 0.51.
- [A Reversal Has No Signature of Its Own — One in Four Ordinary Stretches Looks Stronger](https://marketflowresear.ch/articles/a-reversal-has-no-signature-of-its-own-in-the-order-book/): Around a 2.3% pivot the order book changes in multiples: flow ratios two to five times bigger, book-per-flow ratios three to seventeen times smaller.
- [Derivatives of the Order Book — Reversals Leave a Trace, Not a Warning](https://marketflowresear.ch/articles/derivatives-of-the-order-book-reversals-leave-a-trace-not-a-warning/): We rebuilt the basic event of this series from scratch: 210 seconds cut into ten 21-second blocks, book levels divided by the volume of their own….
- [Change the Question, Not the Network — "Will the Leg End Soon?" Scores 0.73 Where "Which Tenth?" Scored Nothing](https://marketflowresear.ch/articles/change-the-question-not-the-network-leg-end-hazard/): The memory network of the previous study learned to count instead of reading the market.
- [A Memory Network Asked "Is the Leg Over?" Learned to Count, Not to Read](https://marketflowresear.ch/articles/a-memory-network-that-learned-to-count-not-to-read/): An LSTM with attention walked each 2.3% zigzag leg event by event and, at every step, said how far along the leg was.
- [Four Numbers per Event and a 28-Cell Answer — Where the Signal Really Came From](https://marketflowresear.ch/articles/four-numbers-per-event-and-a-28-cell-answer/): The previous networks read each book feature through 13 exponential averages and topped out at a correlation of 0.29 with the leg target.
- [Same Data, Different Readout — How a Transformer Fails Differently From an Attention Query](https://marketflowresear.ch/articles/same-data-different-readout-transformer-vs-attention-query/): We repeated the previous study exactly — same 2.3% zigzag target, same pivot zone, same loss, split and seed — and changed one thing: how the….
- [A Network That Found the Pivot Zone — Only Where It Was Trained](https://marketflowresear.ch/articles/a-network-that-found-the-pivot-zone-only-where-it-was-trained/): We asked an attention network to do one specific thing: tell when price is near the end of a 2.3% zigzag leg — within 10% of the pivot and no more….
- [Six Networks Tried, the Winner Has 53 Parameters — Teaching a Model Where Price Sits in Its Leg](https://marketflowresear.ch/articles/six-networks-one-winner-with-53-parameters/): Correlations said the slow order book knows roughly where price sits inside its zigzag leg.
- [Twelve Hours of Order-Book State and the Direction of a Move — From 0.16 to 0.42](https://marketflowresear.ch/articles/twelve-hours-of-order-book-state-and-the-direction-of-a-move/): Does the order book, averaged over about half a day, know which way the current move is going? We smoothed four book features with exponential….
- [Predicting the Next Phase of a Move From the Order Book — 832 Window Sizes, One Honest Ceiling](https://marketflowresear.ch/articles/predicting-the-next-phase-of-a-move-from-the-order-book/): Every event of BTC's order book sits in one of four phases of a zigzag: rising, a pullback inside a rise, falling, or a pullback inside a fall.
- [A Leg Is the Sum of Its Events — and Size Beats Count](https://marketflowresear.ch/articles/a-leg-is-the-sum-of-its-events-size-beats-count/): We cut the price of six coins into 210-second events and into 2.3% zigzag legs, then asked the simplest question a direction model has to answer:….

## Live bots — every result

Bots built from this research and run live: winners and losers, every trade.

- [Line 7: The Best of All Our Bots — Live Trading Report, Every Trade](https://marketflowresear.ch/articles/line-7-live-trading-report/): We have built and tested more than a dozen trading bots on the research published here.
- [Bot 7 — From an Order-Book Canvas to Live Trading: Every Stretch, Every Result, Every Real Trade](https://marketflowresear.ch/articles/bot-7-from-an-order-book-canvas-to-live-trading-every-result-every-trade/): Bot 7 is the end of a long research line: a transformer reads 30 hours of BTC order book as one canvas and says how likely it is that this moment….
- [Bot 6, Line 61: Nine Trades, Eight Winners, and Why We Are Not Celebrating](https://marketflowresear.ch/articles/bot-6-line-61-nine-trades-eight-winners/): Line 61 of Bot 6 holds one position at a time and has been live since 17 July 2026: 31 days, 9 closed trades, +11.81 USD on a 100 USD slot, win….
- [Bot 6, Line 60: A Bot That Mostly Waits — and What the Clock Costs It](https://marketflowresear.ch/articles/bot-6-line-60-a-bot-that-mostly-waits/): Line 60 of Bot 6 holds one position at a time and has been live since 15 July 2026: 33 days, 9 closed trades, +0.68 USD on a 100 USD slot, win….
- [Bot 4: What Forty Days of Live Paper Trading Actually Look Like](https://marketflowresear.ch/articles/bot-4-forty-days-of-live-paper-trading/): Bot 4 has been trading live on paper since 09 July 2026 — 40 days, 49 closed trades, +29.53 USD on 600 USD of capital, a 80% win rate, against….
- [Gates: The Part of a Trading Bot That Decides Not to Trade](https://marketflowresear.ch/articles/gates-the-part-of-a-bot-that-decides-not-to-trade/): A trading bot's model produces a number; its gate decides whether that number is worth a fee.
- [Typhoon+: A Signal That Finds the Perfect Trader's Zones — and Still Loses Money](https://marketflowresear.ch/articles/typhoon-plus-a-signal-that-finds-the-zones-and-still-loses-money/): We drew the zones a perfect trader would use — 0.9% zigzag turns, a 0.15% band around each one — and found that the price sits inside them barely….

## Order-book research

What second-by-second order-book data can and cannot tell you: walls, drains, volume, market states, reversals.

- [A Detector for Corrections Inside a Move — and Why It Took a New Way of Measuring](https://marketflowresear.ch/articles/a-detector-for-corrections-inside-a-move/): We stopped measuring how loud the order book is and started measuring how much wall stands per unit of aggression — eight features collapsed into….
- [The Order Book Tells You How Far, Never Which Way](https://marketflowresear.ch/articles/the-order-book-tells-you-how-far-never-which-way/): We asked eight order-book features the same two questions at 16 horizons from one second to a day, on six coins: how far will the price move, and….
- [We Taught a Network to See a Move. It Only Heard How Loud It Was.](https://marketflowresear.ch/articles/we-taught-a-network-to-see-a-move-and-it-only-heard-how-loud-it-was/): We trained three networks on the same 56,263 order-book windows and learned two things we did not want to learn.
- [A Recipe, Not a Result: How We Clean an Order-Book Dataset](https://marketflowresear.ch/articles/a-recipe-not-a-result-how-we-clean-an-order-book-dataset/): This is the least glamorous and most useful study of the series: not a finding, but a procedure.
- [Throwing Away a Third of the Data and Keeping 98% of the Reversals](https://marketflowresear.ch/articles/throwing-away-a-third-of-the-data-and-keeping-the-reversals/): If a third of your data provably contains almost nothing you are looking for, it should not be in the training set.
- [Twenty-Five Weeks, Twenty-Five Rebuilds: Which Clustering Survives](https://marketflowresear.ch/articles/twenty-five-weeks-twenty-five-rebuilds-which-clustering-survives/): A clustering that only exists once is a story, not a result.
- [Tops Hide, Bottoms Don't: an Asymmetry We Did Not Expect](https://marketflowresear.ch/articles/tops-hide-bottoms-do-not-the-asymmetry-we-did-not-expect/): This is the mirror of the previous study: the fall from a top to a bottom, 27,180 windows, the same pipeline to the letter.
- [The Rising Half of a Move: What the Order Book Looks Like Between a Bottom and a Top](https://marketflowresear.ch/articles/the-rising-half-of-a-move-what-the-order-book-looks-like/): Half of every price move is a climb from a bottom to a top.
- [One Axis, Again: Clustering 56,000 Order-Book Windows Without a Teacher](https://marketflowresear.ch/articles/one-axis-again-clustering-order-book-windows-without-a-teacher/): Take 56,263 windows of the order book, 210 seconds each, eight features per second.
- [One Network for Three Market States: What a Pooled Reversal Model Learns](https://marketflowresear.ch/articles/one-network-for-three-market-states/): We trained one network on the events of three different market states at once — 55,929 events, 3,758 of which sit on a real turning point — and….
- [Three Classes, One Question: What It Takes to Separate a Reversal](https://marketflowresear.ch/articles/three-classes-one-question-what-it-takes-to-separate-a-reversal/): Three market states, one question: can an order-book event that sits on a turning point be told apart from the ordinary events of its own class?….
- [Class 2: the Middle State Where No Single Feature Wins](https://marketflowresear.ch/articles/class-2-the-middle-state-where-no-single-feature-wins/): The class 2 state fires on 21,704 windows of this dataset, and 1,961 of them (9.0%) sit on a turning point of price.
- [Reversals in a Quiet Market: Where 2.5% of Events Hide](https://marketflowresear.ch/articles/reversals-in-a-quiet-market-where-two-percent-of-events-hide/): The calm state fires on 29,303 windows of this dataset, and 739 of them (2.5%) sit on a turning point of price.
- [The Storm Class: Which Order-Book Feature Marks a Reversal](https://marketflowresear.ch/articles/the-storm-class-which-order-book-feature-marks-a-reversal/): The storm state fires on 5,249 windows of this dataset, and 1,085 of them (20.7%) sit on a turning point of price.
- [We Went Looking for the DNA of a Reversal. We Found a Ceiling Instead.](https://marketflowresear.ch/articles/we-went-looking-for-the-dna-of-a-reversal-we-found-a-ceiling-instead/): An earlier study on this dataset (175 days of BTC) found 271 order-book events that sit exactly on the turning points of price, and showed that a….
- [We Found the Reversal Signal. Machine Learning Could Not Tell It Apart.](https://marketflowresear.ch/articles/we-found-the-reversal-signal-machine-learning-could-not-tell-it-apart/): A market moves in legs: a few percent one way, a turning point, then a few percent the other.
- [Sixteen Networks, One Working Answer — and the Price Benchmark That Nearly Killed It](https://marketflowresear.ch/articles/sixteen-networks-one-working-answer/): We trained sixteen one-class autoencoders on Bitcoin order-book windows: four market states, and inside each state one network per phase of the….
- [Does the Phase of a Move Change What the Order Book Looks Like?](https://marketflowresear.ch/articles/does-the-phase-of-a-move-change-what-the-order-book-looks-like/): We have four detectors of market state and a map of where price is in its move — reversal zones and the legs between them.
- [Eight Ways to Compress 210 Seconds: the Autoencoder That Lost to PCA](https://marketflowresear.ch/articles/the-autoencoder-that-lost-to-pca/): We cut 170 days of per-second Bitcoin order-book data into 50,143 back-to-back windows and pushed them through eight different compressions — PCA,….
- [Two Poles of Market Activity, Found Without a Teacher on Six Coins](https://marketflowresear.ch/articles/two-poles-of-market-activity-found-on-six-coins/): We cut 122–141 days of per-second order-book data on 6 coins into 164,342 windows of 210 seconds, showed a clustering pipeline nothing but the….
- [Local tops and bottoms in the order book: what four trained models actually see](https://marketflowresear.ch/articles/local-tops-and-bottoms-what-four-trained-models-actually-see/): We marked every local top and bottom on 171 days of per-second Bitcoin order-book data — 734 of them, defined by a rule rather than by hand — and….
- [Four Market States, Found Twice by Two Independent Pipelines](https://marketflowresear.ch/articles/four-market-states-found-twice-by-two-independent-pipelines/): Four market states, found twice by two pipelines that share no code.
- [There Are No Classes in the Order Book — Only One State That Survives](https://marketflowresear.ch/articles/no-classes-in-the-order-book-only-one-state/): We locked a 210-second window to its own 1,680 numbers — eight order-book features, no norm, no price, no clock — and asked whether the market….
- [Event Classes Found Without a Teacher: 3 Survive, 60% Are Noise](https://marketflowresear.ch/articles/event-classes-found-without-a-teacher/): We asked the order book to sort its own events into classes — 2,786 events, eight features, no price and no human labels.
- [The Alphabet of the Order Book: 6,561 Possible States, 333 Real Ones](https://marketflowresear.ch/articles/the-alphabet-of-the-order-book-signal-states/): We described the order book as a single word: eight features, three states each.
- [From Calm to Calm: a Standard for What Counts as a Signal (BTC)](https://marketflowresear.ch/articles/signal-standard-from-calm-to-calm-btc/): We stopped defining market signals with a stopwatch.
- [Side Against Side: Resistance and Support Compared](https://marketflowresear.ch/articles/resistance-against-support/): We put all six order-book features grouped into the resistance side and the support side into one study and asked how the pieces relate, which one….
- [Flow Against the Wall: When Volume Meets Resting Liquidity](https://marketflowresear.ch/articles/flow-against-the-wall/): We put executed volume hitting the resting wall on the opposite side into one study and asked how the pieces relate, which one leads, and what the….
- [The Life of a Sell Wall: Standing, Growing, Thinning](https://marketflowresear.ch/articles/lifecycle-of-a-sell-wall/): We put the three phases of the same sell-side wall — how much stands there, what arrives, what leaves into one study and asked how the pieces….
- [Executed Sell Volume, Measured Alone](https://marketflowresear.ch/articles/executed-sell-volume-alone/): We isolated executed sell volume — one feature, no partner, no composite — and measured what it predicts second by second across six coins.
- [Executed Buy Volume, Measured Alone](https://marketflowresear.ch/articles/executed-buy-volume-alone/): We isolated executed buy volume — one feature, no partner, no composite — and measured what it predicts second by second across six coins.
- [Drain Out of the Buy Wall: Liquidity Leaving Before the Move](https://marketflowresear.ch/articles/drain-out-of-the-buy-wall/): We isolated drain out of the buy wall — one feature, no partner, no composite — and measured what it predicts second by second across six coins.
- [Build-Up Into the Buy Wall: What Arriving Liquidity Predicts](https://marketflowresear.ch/articles/build-up-into-the-buy-wall/): We isolated build-up into the buy wall — one feature, no partner, no composite — and measured what it predicts second by second across six coins.
- [Drain Out of the Sell Wall: Liquidity Leaving Before the Move](https://marketflowresear.ch/articles/drain-out-of-the-sell-wall/): We isolated drain out of the sell wall — one feature, no partner, no composite — and measured what it predicts second by second across six coins.
- [Build-Up Into the Sell Wall: What Arriving Liquidity Predicts](https://marketflowresear.ch/articles/build-up-into-the-sell-wall/): We isolated build-up into the sell wall — one feature, no partner, no composite — and measured what it predicts second by second across six coins.
- [The Resting Buy Wall, Measured Alone](https://marketflowresear.ch/articles/the-resting-buy-wall-alone/): We isolated the resting buy wall — one feature, no partner, no composite — and measured what it predicts second by second across six coins.
- [The Resting Sell Wall, Measured Alone](https://marketflowresear.ch/articles/the-resting-sell-wall-alone/): We isolated the resting sell wall — one feature, no partner, no composite — and measured what it predicts second by second across six coins.
- [The Support Cycle: Liquidity Arriving vs Leaving](https://marketflowresear.ch/articles/support-cycle-build-up-versus-drain/): We measured the same buy-side wall growing and thinning second by second across six coins, and asked the only question that matters: does it tell….
- [The Resistance Cycle: Liquidity Arriving vs Leaving](https://marketflowresear.ch/articles/resistance-cycle-build-up-versus-drain/): We measured the same sell-side wall growing and thinning second by second across six coins, and asked the only question that matters: does it tell….
- [Liquidity Arriving: What a Growing Wall Says Next](https://marketflowresear.ch/articles/wall-build-up-liquidity-arriving/): We measured fresh resting liquidity arriving into the book second by second across six coins, and asked the only question that matters: does it….
- [Liquidity Pulled: What Happens When a Wall Drains](https://marketflowresear.ch/articles/wall-drain-liquidity-pulled-before-the-move/): We measured resting liquidity being pulled away from the book second by second across six coins, and asked the only question that matters: does it….
- [The Wall That Goes Quiet: What Resting Liquidity Predicts](https://marketflowresear.ch/articles/resting-walls-what-thick-liquidity-predicts/): We measured resting limit liquidity sitting on both sides of the book second by second across six coins, and asked the only question that matters:….
- [About Market Research Lab — What We Collect and Why](https://marketflowresear.ch/articles/about-market-research-lab/): Most market commentary is storytelling.
- [Volume Is the Fuel — Not the Steering Wheel](https://marketflowresear.ch/articles/volume-is-fuel-not-steering-wheel/): We recorded the Binance order book every second for six coins over five months and ran eighteen tests on what volume really does.
- [Order flow predicts how much, not which way](https://marketflowresear.ch/articles/order-flow-predicts-volatility/): Across six coins, order-flow surges line up with bigger price moves — but not their direction.

## Optional

- [Full text of every article](https://marketflowresear.ch/llms-full.txt): the complete articles in Markdown.
- [Open dataset](https://marketflowresear.ch/#dataset): 12.1 million rows of BTC/USDT order book, one per second, 14 Feb – 5 Aug 2026.
- [RSS](https://marketflowresear.ch/feed.xml)