Notes

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 and 2025 to train, and data from January 1 to September 11, 2026, to test. The experiment used a rule where if the price moved a certain amount, it would immediately close the trade, otherwise it would wait for the price to pull back to a certain level, known as the Average True Range or ATR pullback depth. The experiment also included fees of 25 basis points, which is a measure of the cost of trading, credited to each side of the trade. The results of the experiment are shown in the following list, which includes the return on investment, the maximum amount lost, known as the drawdown or DD, and the percentage of winning trades. The list shows the results for different versions of the experiment, including ones with and without management, and with different levels of slippage, which is the difference between the expected price and the actual price of a trade. The results are as follows: [{"s":"ETH","tf":"1Day","d":-1,"train":false,"slipbp":10,"managed":false,"return":-12.972621042472332,"DD":25.926818266648045,"win":20,"N":5,"open":0,"mean":-2.4766184908829803,"targets":0,"stops":2,"missing":98,"partial":0,"recon":0},{"s":"ETH","tf":"1Day","d":-1,"train":false,"slipbp":10,"managed":true,"return":-12.703528416454557,"DD":25.926818266648045,"win":20,"N":5,"open":0,"mean":-2.4089048036770078,"targets":0,"stops":2,"missing":98,"partial":1,"recon":0},{"s":"ETH","tf":"1Day","d":-1,"train":false,"slipbp":30,"managed":false,"return":-14.695885997531978,"DD":26.810388253475036,"win":20,"N":5,"open":0,"mean":-2.8659345527439117,"targets":0,"stops":2,"missing":98,"partial":0,"recon":-1.1102230246251565e-16},{"s":"ETH","tf":"1Day","d":-1,"train":false,"slipbp":30,"managed":true,"return":-14.43212178537241,"DD":26.810388253475036,"win":20,"N":5,"open":0,"mean":-2.7984911804678503,"targets":0,"stops":2,"missing":98,"partial":1,"recon":-1.1102230246251565e-16}]. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results15.json

A news scan found the following data line: Sep 12 frozen experiment. The training period is from 2024 to 2025, and the holdout period is from Jan 1 to Sep 11, 2026. The decision rule is d = -1 for immediate breakout, otherwise ATR pullback depth. Fees are 25 basis points credited each side. The results include return, drawdown, mean, and win percent. The list of results is as follows: for ETH with a 1-day time frame, the results are: return of 53.34260331587186, drawdown of 26.7951745308839, win percent of 50, and other metrics, with 10 basis points slip and no management; return of 43.21930512266285, drawdown of 26.72555496240102, win percent of 50, and other metrics, with 10 basis points slip and management; return of 51.57747717030936, drawdown of 27.556019642617535, win percent of 43.75, and other metrics, with 30 basis points slip and no management; return of 40.990292928981866, drawdown of 27.470574830622763, win percent of 43.75, and other metrics, with 30 basis points slip and management. Note that ATR means Average True Range, which is a measure of price volatility. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results15.json

A news scan found the headline 'SOL/USD C91204 results cells 93-96'. This is a raw data output. The data shows the results of an experiment on the SOL/USD currency pair, which started on September 12. The training period was from 2024 to 2025, and the holdout period was from January 1 to September 11, 2026. The experiment used a decision rule, where d equals -1 for an immediate breakout, otherwise it used an ATR pullback depth, which is a measure of price volatility. The fees were 25 basis points credited to each side of the trade. The results include return, drawdown, mean, and win percentage. The data is not pristine and prospective out-of-sample, meaning it is not a perfect or real-time test. The data includes four sets of results, each with the following information: symbol, time frame, decision rule, training status, slip basis points, return, drawdown, win percentage, number of trades, open trades, mean, targets, stops, missing trades, entry-exit, and recon. The results are as follows: for the first set, the symbol is SOL, the time frame is 1 hour, the decision rule is d equals 1, the training status is true, the slip basis points is 10, the return is 25.452304897025613, the drawdown is 26.7658218200341, the win percentage is 60, the number of trades is 5, the open trades is 0, the mean is 5.609130745612734, the targets is 1, the stops is 2, the missing trades is 6, the entry-exit is 0, and the recon is 0. The second set has the same symbol, time frame, and decision rule, but the slip basis points is 30, the return is 24.486642785971924, the drawdown is 27.188040848173134, the win percentage is 60, the number of trades is 5, the open trades is 0, the mean is 5.467336217901357, the targets is 1, the stops is 2, the missing trades is 6, the entry-exit is 0, and the recon is 0. The third and fourth sets have the same symbol and time frame, but the training status is false, the slip basis points is 10 and 30 respectively, and all other values are 0 or null. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 to 2025 for training and data from January 1 to September 11, 2026, for testing. The decision rule was to immediately stop the trade if a certain condition was met, otherwise to use a method called ATR pullback depth, which is a way to measure how far a price moves away from a recent high or low. A fee of 25 basis points was added to each side of the trade. The results are shown in a list of four sets of data for the cryptocurrency SOL, which is traded against the US dollar. Each set of data includes the return, which is the profit or loss, the drawdown, which is the largest loss, and the percentage of winning trades. The data also includes the number of trades, the number of open trades, and the average return. The list of data is as follows: [{"s":"SOL","tf":"1Hour","d":0.5,"train":true,"slipbp":10,"return":-6.337931816662357,"DD":37.675288670069904,"win":42.857142857142854,"N":7,"open":0,"mean":-0.07705910000151496,"targets":1,"stops":4,"missing":6,"entryexit":0,"recon":0},{"s":"SOL","tf":"1Hour","d":0.5,"train":true,"slipbp":30,"return":-7.384793951110957,"DD":38.143568854771544,"win":42.857142857142854,"N":7,"open":0,"mean":-0.21957970110344027,"targets":1,"stops":4,"missing":6,"entryexit":0,"recon":-1.1102230246251565e-16},{"s":"SOL","tf":"1Hour","d":0.5,"train":false,"slipbp":10,"return":-3.777931964273773,"DD":20.456624239854047,"win":50,"N":2,"open":0,"mean":-1.5554805987077591,"targets":0,"stops":1,"missing":1,"entryexit":0,"recon":0},{"s":"SOL","tf":"1Hour","d":0.5,"train":false,"slipbp":30,"return":-4.334304756950081,"DD":20.59958926604496,"win":50,"N":2,"open":0,"mean":-1.8495684618483776,"targets":0,"stops":1,"missing":1,"entryexit":0,"recon":0}]. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 and 2025 to train, and data from January 1 to September 11, 2026, to test. The experiment had a rule to immediately stop if a certain condition was met, otherwise it would use a method called ATR pullback depth, which is a way to measure how much a price moves back after a big change. The experiment also included fees of 25 basis points, which is a way to measure a small percentage, credited to each side of a trade. The results of the experiment included return, which is the profit made, drawdown, which is the biggest loss, and mean, which is the average. The results also included win percent, which is the percentage of trades that made a profit. The experiment was not completely open and not pristine prospective out of sample, which means it was not a perfect test. The results were: for SOL, a cryptocurrency, with a time frame of 1 hour, and a parameter d of 0.25, the experiment had different outcomes depending on the slipbp, which is a measure of the difference between the expected and actual price of a trade. The outcomes were: return of 1.1281856724043582, drawdown of 38.39826133614349, win percent of 42.857142857142854, and other metrics, for a slipbp of 10 and train true; return of -1.63510107906516, drawdown of 39.19674843624538, win percent of 42.857142857142854, and other metrics, for a slipbp of 30 and train true; return of -8.722661081098892, drawdown of 21.282368212906434, win percent of 50, and other metrics, for a slipbp of 10 and train false; return of -9.250442586933028, drawdown of 21.423849111630954, win percent of 50, and other metrics, for a slipbp of 30 and train false. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 to 2025 for training and data from January 1 to September 11, 2026, for testing. The decision rule was to immediately break out if d was -1, otherwise use the Average True Range pullback depth, which is a measure of price volatility. Fees of 25 basis points were credited to each side of the trade. The results include return, drawdown, mean, and win percentage. The trades were marked as open and not closed, and the results are not pristine prospective out-of-sample, meaning they are based on a specific set of data. The results are as follows: [{"s":"SOL","tf":"1Hour","d":-1,"train":true,"slipbp":10,"return":33.64409032661262,"DD":37.786239428419776,"win":54.54545454545454,"N":11,"open":0,"mean":3.3942918911147224,"targets":0,"stops":4,"missing":6,"entryexit":0,"recon":-2.220446049250313e-16},{"s":"SOL","tf":"1Hour","d":-1,"train":true,"slipbp":30,"return":29.031729219926206,"DD":38.253026608442894,"win":54.54545454545454,"N":11,"open":0,"mean":3.0534701961129063,"targets":0,"stops":4,"missing":6,"entryexit":0,"recon":0},{"s":"SOL","tf":"1Hour","d":-1,"train":false,"slipbp":10,"return":-8.628140562811094,"DD":21.914734577010197,"win":33.33333333333333,"N":6,"open":1,"mean":-0.6885124421883808,"targets":1,"stops":3,"missing":1,"entryexit":0,"recon":0},{"s":"SOL","tf":"1Hour","d":-1,"train":false,"slipbp":30,"return":-9.9914338246884,"DD":22.512552889387504,"win":33.33333333333333,"N":6,"open":1,"mean":-0.8798975417621036,"targets":1,"stops":3,"missing":1,"entryexit":0,"recon":0}]. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 and 2025 to train, and data from January 1 to September 11, 2026, to test. The decision rule was to immediately stop if the price moved a certain way, otherwise to wait for the price to pull back by a certain amount, which is called the Average True Range or ATR pullback depth. A fee of 25 basis points was added to each trade. The results are shown in a list of four sets of numbers, which are: return, or the profit made, drawdown, or the largest loss, mean, or the average profit, and win percent, or the percentage of trades that made a profit. The list is as follows: for SOL with a 1-day time frame, the results were a return of 15.772344183293496 percent, a drawdown of 16.07520965846756 percent, a win percent of 50 percent, for a trade with a 10 basis point slip, and a return of 15.9970511069621 percent, a drawdown of 16.05617026642362 percent, a win percent of 50 percent, for a trade with a 30 basis point slip, when using training data. When not using training data, the results were a return of -4.205548203202025 percent, a drawdown of 16.54395558478299 percent, a win percent of 50 percent, for a trade with a 10 basis point slip, and a return of -4.750543509115923 percent, a drawdown of 16.686163868857395 percent, a win percent of 50 percent, for a trade with a 30 basis point slip. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 and 2025 to train, and data from January 1 to September 11, 2026, to test. The decision rule was to immediately stop if the price moved a certain way, otherwise to wait for the price to pull back by a certain amount, which is called the Average True Range or ATR pullback depth. A fee of 25 basis points was added to each trade. The results include return, drawdown, mean, and win percentage. The results are for open positions that were not closed, and are not perfect prospective out-of-sample results. The results are as follows: for SOL with a 1-day time frame, a distance of 0.5, and a training period, with a slip of 10 basis points, the return was 13.226880033264333, the drawdown was 19.79081149597316, the win percentage was 50, and the mean was 4.4279119502509126. With a slip of 30 basis points, the return was 12.790128041872073, the drawdown was 19.933949742447588, the win percentage was 50, and the mean was 4.36078289866498. For the test period, with a slip of 10 basis points, the return was -11.009586868160993, the drawdown was 15.7133983491754, the win percentage was 33.33333333333333, and the mean was -3.2758142088458397. With a slip of 30 basis points, the return was -11.680651972623801, the drawdown was 15.856517283775407, the win percentage was 33.33333333333333, and the mean was -3.530065983214551. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 to 2025 for training and data from January 1 to September 11, 2026, for testing. The experiment had a rule to immediately stop trading if a certain condition was met, otherwise it used a method called ATR pullback depth, which is a way to measure how much a price moves away from a recent high or low. The experiment also included fees of 25 basis points, which is a way to measure a small percentage, credited to each side of a trade. The results of the experiment included return, which is the profit or loss, drawdown, which is the largest loss, and mean, which is the average. The results also included the percentage of winning trades. The experiment was not completely open and was not a perfect test of the strategy. The results were as follows: there were four trades for SOL with a time frame of 1 day and a parameter d of 0.25. The first two trades were part of the training data and had a slip of 10 and 30 basis points, which is a way to measure a small percentage. The returns were 13.226880033264333 and 12.790128041872073, the drawdowns were 19.79081149597316 and 19.933949742447588, and the winning percentages were 50. The last two trades were not part of the training data and had a slip of 10 and 30 basis points. The returns were -22.902340420012933 and -23.774255517578336, the drawdowns were 26.87867190229888 and 27.561173435867083, and the winning percentages were 25. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 to 2025 for training and from January 1 to September 11, 2026, for testing. The decision rule was to immediately close a trade if it was not profitable, or to use the Average True Range, a measure of price volatility, to determine the stop-loss depth. A fee of 25 basis points, a measure of 0.25%, was credited to each side of the trade. The results include return, drawdown, mean, and win percentage. The trades were marked as open but not closed, and the results are not considered pristine or prospective out-of-sample. The results are as follows: for SOL with a 1-day time frame and a decision rule of -1, the return was -9.027846920052408%, the drawdown was 36.279436252223405%, and the win percentage was 40%. For the same setup but with a slip of 30 basis points, the return was -11.605234106840424%, the drawdown was 37.48495506570501%, and the win percentage was 40%. For SOL with a 1-day time frame, a decision rule of -1, and not used for training, the return was -13.538212714957698%, the drawdown was 34.70905562974521%, and the win percentage was 33.33333333333333%. For the same setup but with a slip of 30 basis points, the return was -14.314376001493468%, the drawdown was 35.31727593447948%, and the win percentage was 16.666666666666664%. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 and 2025 to train, and data from January 1 to September 11, 2026, to test. The decision rule was to immediately stop if the price moved a certain way, otherwise to wait for the price to pull back to a level determined by the Average True Range, which is a measure of how much prices typically move. A fee of 25 basis points, which is 0.25 percent, was added to each trade. The results are shown in the following list: there were four sets of results, each with the following information: the symbol, which was ETH, the time frame, which was 1 hour, the day, which was 1, whether the results were from training or testing, the amount of slippage, which is the difference between the expected and actual price of a trade, in basis points, the return, which is the profit or loss, the drawdown, which is the largest loss, the percentage of winning trades, the number of trades, the number of open trades, the mean return, the number of targets, which is the number of times a profit goal was reached, the number of stops, which is the number of times a loss limit was reached, the number of missing trades, the number of entry and exit trades, and the number of recon trades. The results were: [{"s":"ETH","tf":"1Hour","d":1,"train":true,"slipbp":10,"return":-45.187596421834066,"DD":54.624430518456954,"win":22.22222222222222,"N":9,"open":0,"mean":-6.314685817720377,"targets":0,"stops":7,"missing":3,"entryexit":0,"recon":0},{"s":"ETH","tf":"1Hour","d":1,"train":true,"slipbp":30,"return":-46.311565356414484,"DD":55.29432832209184,"win":22.22222222222222,"N":9,"open":0,"mean":-6.534827352468351,"targets":0,"stops":7,"missing":3,"entryexit":0,"recon":0},{"s":"ETH","tf":"1Hour","d":1,"train":false,"slipbp":10,"return":11.960544365481063,"DD":15.727493792771941,"win":66.66666666666666,"N":3,"open":0,"mean":3.9028825952207358,"targets":0,"stops":0,"missing":1,"entryexit":0,"recon":0},{"s":"ETH","tf":"1Hour","d":1,"train":false,"slipbp":30,"return":10.625041091620657,"DD":16.063911990460067,"win":66.66666666666666,"N":3,"open":0,"mean":3.4880993869068697,"targets":0,"stops":0,"missing":1,"entryexit":0,"recon":0}]. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 to 2025 for training and data from January 1 to September 11, 2026, for testing. The decision rule was to immediately stop if the price moved in the opposite direction, otherwise to wait for the price to pull back to a level determined by the Average True Range, which is a measure of price volatility. A fee of 25 basis points, or 0.25%, was added to each trade. The results are as follows: return on investment, drawdown, mean, and win percentage. The results are for open positions that have not been closed yet and are not based on a pristine prospective out-of-sample test, which means the test was not done under real-world conditions. The results are: [{"s":"ETH","tf":"1Hour","d":0.5,"train":true,"slipbp":10,"return":59.182117442857795,"DD":34.81209510101794,"win":53.84615384615385,"N":13,"open":1,"mean":5.0707892290636805,"targets":3,"stops":6,"missing":3,"entryexit":0,"recon":0},{"s":"ETH","tf":"1Hour","d":0.5,"train":true,"slipbp":30,"return":21.66107270386273,"DD":35.28360363168789,"win":38.46153846153847,"N":13,"open":1,"mean":2.976234882059139,"targets":3,"stops":7,"missing":3,"entryexit":0,"recon":0},{"s":"ETH","tf":"1Hour","d":0.5,"train":false,"slipbp":10,"return":19.206746821335162,"DD":13.790861403942479,"win":100,"N":3,"open":1,"mean":6.157475017100844,"targets":0,"stops":0,"missing":1,"entryexit":0,"recon":0},{"s":"ETH","tf":"1Hour","d":0.5,"train":false,"slipbp":30,"return":17.549943369892638,"DD":13.790861403942479,"win":100,"N":3,"open":1,"mean":5.733691412890049,"targets":0,"stops":0,"missing":1,"entryexit":0,"recon":0}] https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 and 2025 to train, and data from January 1 to September 11, 2026, to test. The decision rule was to immediately stop if the price moved a certain way, otherwise to wait for the price to pull back by a certain amount, which is called the Average True Range or ATR pullback depth. The fees were 25 basis points, which is a measure of the cost of a transaction, credited to each side of the transaction. The results are shown in the following list, which includes return, drawdown, mean, and win percentage. The list shows results for four different scenarios, including the cryptocurrency ETH, with a time frame of 1 hour, and different slip values, which refer to the difference between the expected price and the actual price of a transaction. The results are as follows: [{"s":"ETH","tf":"1Hour","d":0.25,"train":true,"slipbp":10,"return":10.013423919635489,"DD":41.31982761576022,"win":28.57142857142857,"N":14,"open":0,"mean":2.094268530332681,"targets":3,"stops":8,"missing":3,"entryexit":0,"recon":0},{"s":"ETH","tf":"1Hour","d":0.25,"train":true,"slipbp":30,"return":-1.294712274335763,"DD":41.85169369777062,"win":28.57142857142857,"N":14,"open":0,"mean":1.0719941430414683,"targets":2,"stops":8,"missing":3,"entryexit":0,"recon":0},{"s":"ETH","tf":"1Hour","d":0.25,"train":false,"slipbp":10,"return":4.656312212559133,"DD":20.925981223464106,"win":75,"N":4,"open":1,"mean":1.5971356486825405,"targets":0,"stops":1,"missing":1,"entryexit":0,"recon":0},{"s":"ETH","tf":"1Hour","d":0.25,"train":false,"slipbp":30,"return":3.0091365467297404,"DD":21.07355481162829,"win":75,"N":4,"open":1,"mean":1.2410097500084012,"targets":0,"stops":1,"missing":1,"entryexit":0,"recon":0}]. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The training period was from 2024 to 2025, and the holdout period was from January 1 to September 11, 2026. The decision rule was to immediately breakout if d equals -1, otherwise use ATR pullback depth. Fees of 25 basis points were credited to each side. The results include return, drawdown, mean, and win percentage. The results are based on open positions that were not closed, and the data is not pristine prospective out-of-sample. The results are as follows: for ETH with a 1 Hour time frame, the results were return of 65.60399643679446, drawdown of 29.957079469547566, win percentage of 50, and mean of 3.945519269628471 when the slip was 10 basis points and the training was true. When the slip was 30 basis points and the training was true, the results were return of 60.902564975691575, drawdown of 30.61782191361274, win percentage of 50, and mean of 3.790857757049194. When the training was false and the slip was 10 basis points, the results were return of 13.518439861108945, drawdown of 22.03372917349087, win percentage of 66.66666666666666, and mean of 3.171236983728678. When the training was false and the slip was 30 basis points, the results were return of 11.81695877839546, drawdown of 22.178568166081202, win percentage of 66.66666666666666, and mean of 2.959572781870277. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 and 2025 to train, and data from January 1 to September 11, 2026, to test. The decision rule was to immediately stop if the price moved a certain way, otherwise to wait for the price to pull back by a certain amount, which is called the Average True Range or ATR pullback depth. A fee of 25 basis points was added to each trade. The results include return, drawdown, mean, and win percentage. The trades were not fully closed and the results are not perfect because they are based on old data, which is called out-of-sample or OOS data. The results are as follows: for ETH with a 1-day time frame and a decision delay of 1 day, with a slip of 10 basis points, the return was 6.66376555369077 percent, the drawdown was 13.71663627894577 percent, and the win rate was 50 percent. With a slip of 30 basis points, the return was 16.41555276240574 percent, the drawdown was 29.4987512294249 percent, and the win rate was 50 percent. When the model was not being trained, with a slip of 10 basis points, the return was 2.090775938242717 percent, the drawdown was 5.176639016379648 percent, and the win rate was 100 percent. With a slip of 30 basis points, the return was 1.683226710297947 percent, the drawdown was 5.176639016379648 percent, and the win rate was 100 percent. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 and 2025 to train, and data from January 1 to September 11, 2026, to test. The rules were to immediately stop if the price moved a certain way, or to wait for the price to pull back to a certain level, which is called an ATR pullback depth, where ATR means Average True Range, a measure of price movement. There was a fee of 25 basis points, which is a way to measure a small percentage, credited to both the buyer and seller. The results are shown in a list: [{"s":"ETH","tf":"1Day","d":0.5,"train":true,"slipbp":10,"return":-23.68857142509103,"DD":34.213756214161194,"win":37.5,"N":8,"open":1,"mean":-2.1521207772281796,"targets":0,"stops":5,"missing":0,"entryexit":0,"recon":0},{"s":"ETH","tf":"1Day","d":0.5,"train":true,"slipbp":30,"return":-25.42440140968496,"DD":35.0882231565198,"win":37.5,"N":8,"open":1,"mean":-2.4191004843917105,"targets":0,"stops":5,"missing":0,"entryexit":1,"recon":0},{"s":"ETH","tf":"1Day","d":0.5,"train":false,"slipbp":10,"return":-0.2560795367949842,"DD":19.870208871574757,"win":66.66666666666666,"N":3,"open":0,"mean":0.03620787654811848,"targets":0,"stops":1,"missing":0,"entryexit":0,"recon":-1.1102230246251565e-16},{"s":"ETH","tf":"1Day","d":0.5,"train":false,"slipbp":30,"return":-1.2362885405404311,"DD":20.020377619446094,"win":66.66666666666666,"N":3,"open":0,"mean":-0.29720252039413164,"targets":0,"stops":1,"missing":0,"entryexit":0,"recon":0}]. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline about an experiment that started on September 12. The training period was from 2024 to 2025, and the holdout period was from January 1 to September 11, 2026. The decision rule was to immediately break out if d equals -1, otherwise use the average true range pullback depth. Fees of 25 basis points were credited to each side of the trade. The results include return, drawdown, mean, and win percentage. The trades were not all closed and not all were prospective out-of-sample trades. The results are as follows: for ETH with a 1-day time frame, a distance of 0.25, and training set to true, with a slip of 10 basis points, the return was -13.682997033565847, the drawdown was 30.76627877648086, the win percentage was 44.44444444444444, and there were 9 trades. With a slip of 30 basis points, the return was -15.9831622441457, the drawdown was 31.80913399017061, the win percentage was 44.44444444444444, and there were 9 trades. For the test set, with a slip of 10 basis points, the return was -0.2560795367949842, the drawdown was 19.870208871574757, the win percentage was 66.66666666666666, and there were 3 trades. With a slip of 30 basis points, the return was -1.2362885405404311, the drawdown was 20.020377619446094, the win percentage was 66.66666666666666, and there were 3 trades. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the following data line: Sep 12 frozen experiment. This refers to a test that was stopped on September 12. The test used data from 2024 and 2025 to train, and data from January 1 to September 11, 2026, to hold out. The decision rule was to immediately break out if d was -1, otherwise use the Average True Range pullback depth, which is a measure of price volatility. Fees of 25 basis points were credited to each side of the trade. The results include return, drawdown, mean, and win percentage. The trades were not marked as closed and the results are not pristine prospective out-of-sample, meaning they are not perfectly reliable. The results are as follows: for ETH with a 1-day time frame and d = -1, with 10 basis points slip, the return was 32.19%, the drawdown was 25.56%, and the win percentage was 43.75%. With 30 basis points slip, the return was -4.81%, the drawdown was 43.11%, and the win percentage was 31.25%. For the holdout period, with 10 basis points slip, the return was -12.63%, the drawdown was 24.35%, and the win percentage was 20%. With 30 basis points slip, the return was -13.99%, the drawdown was 24.62%, and the win percentage was 20%. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 to 2025 for training and data from January 1 to September 11, 2026, for testing. The experiment had a rule that if the price moved immediately, it would be considered a breakout, otherwise, it would be considered a pullback. The depth of the pullback was measured using the Average True Range, or ATR, which is a measure of price volatility. The experiment also took into account fees of 25 basis points, or 0.25%, which were added to the results. The results include return, drawdown, mean, and win percentage. The results are for open positions that were not closed, and the data is not pristine or prospective out-of-sample, meaning it is not perfect or based on future data. The results are as follows: [{"s":"BTC","tf":"1Hour","d":1,"train":true,"slipbp":10,"return":19.758755387344216,"DD":20.71605798412265,"win":50,"N":6,"open":0,"mean":3.600540494328003,"targets":1,"stops":2,"missing":3,"entryexit":0,"recon":-2.220446049250313e-16},{"s":"BTC","tf":"1Hour","d":1,"train":true,"slipbp":30,"return":18.398317636637795,"DD":21.352271005150993,"win":50,"N":6,"open":0,"mean":3.420130969837123,"targets":1,"stops":2,"missing":3,"entryexit":0,"recon":-2.220446049250313e-16},{"s":"BTC","tf":"1Hour","d":1,"train":false,"slipbp":10,"return":3.5895172935296538,"DD":8.75596176977782,"win":50,"N":2,"open":0,"mean":2.095590060189961,"targets":0,"stops":1,"missing":1,"entryexit":0,"recon":0},{"s":"BTC","tf":"1Hour","d":1,"train":false,"slipbp":30,"return":2.981083765220194,"DD":8.746361999812047,"win":50,"N":2,"open":0,"mean":1.7869270210402577,"targets":0,"stops":1,"missing":1,"entryexit":0,"recon":-2.220446049250313e-16}] https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 to 2025 for training and data from January 1 to September 11, 2026, for testing. The decision rule was to immediately stop the trade if a certain condition was met, otherwise to use a method called Average True Range pullback depth, which is a way to measure how far a price is likely to move. The fees were 25 basis points, which is a measure of the cost of a trade, credited to each side of the trade. The results include return, which is the profit or loss, drawdown, which is the largest loss, and mean, which is the average return. The results also include the percentage of winning trades. The trades that are still open are marked, and the results are not based on a pristine prospective out-of-sample test, which means the test was not done in a completely realistic way. The results are as follows: [{"s":"BTC","tf":"1Hour","d":0.5,"train":true,"slipbp":10,"return":-17.092346434853635,"DD":31.952780168634444,"win":41.17647058823529,"N":17,"open":0,"mean":-0.6848427321102964,"targets":1,"stops":8,"missing":3,"entryexit":0,"recon":0},{"s":"BTC","tf":"1Hour","d":0.5,"train":true,"slipbp":30,"return":-20.531185021949426,"DD":32.18797818580333,"win":41.17647058823529,"N":17,"open":0,"mean":-0.9292018929952919,"targets":1,"stops":8,"missing":3,"entryexit":0,"recon":-1.1102230246251565e-16},{"s":"BTC","tf":"1Hour","d":0.5,"train":false,"slipbp":10,"return":1.302885890119554,"DD":9.540324449885318,"win":66.66666666666666,"N":3,"open":1,"mean":1.1439406849313736,"targets":0,"stops":1,"missing":1,"entryexit":0,"recon":0},{"s":"BTC","tf":"1Hour","d":0.5,"train":false,"slipbp":30,"return":0.10613080735990543,"DD":9.912252832513568,"win":66.66666666666666,"N":3,"open":1,"mean":0.8061081876898987,"targets":0,"stops":1,"missing":1,"entryexit":0,"recon":0}] https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to a trading experiment that was stopped on September 12. The experiment used data from 2024 and 2025 to train, and data from January 1 to September 11, 2026, to test. The experiment used a rule called 'immediate breakout' if a certain condition, d, was -1, otherwise it used 'ATR pullback depth', which is a way to measure how much a price moves back after a big change. The experiment also included fees of 25 basis points, which is a way to measure a small percentage, credited to each side of a trade. The results include return, which is the profit or loss, drawdown, which is the biggest loss, and mean, which is the average profit or loss. The results also include win percent, which is the percentage of trades that were profitable. The experiment was not pristine, meaning it was not perfect, and was prospective out of sample, meaning it was tested on new data. The results are: a list of trades for BTC, which is a type of currency, with the following details: time frame of 1 hour, d of 0.25, train true or false, slipbp of 10 or 30, return, drawdown, win percent, number of trades, open trades, mean, targets, stops, missing trades, entry exit, and recon. The list includes: [{'s': 'BTC', 'tf': '1Hour', 'd': 0.25, 'train': True,'slipbp': 10,'return': -26.711721943595045, 'DD': 34.64174868129714, 'win': 35.294117647058826, 'N': 17, 'open': 0,'mean': -1.4102467478242195, 'targets': 1,'stops': 9,'missing': 3, 'entryexit': 0,'recon': 0}, {'s': 'BTC', 'tf': '1Hour', 'd': 0.25, 'train': True,'slipbp': 30,'return': -29.59299855587161, 'DD': 37.020475272082166, 'win': 35.294117647058826, 'N': 17, 'open': 0,'mean': -1.6400741454010042, 'targets': 1,'stops': 9,'missing': 3, 'entryexit': 0,'recon': 0}, {'s': 'BTC', 'tf': '1Hour', 'd': 0.25, 'train': False,'slipbp': 10,'return': -1.945755441955288, 'DD': 10.626447602905742, 'win': 33.33333333333333, 'N': 3, 'open': 1,'mean': 0.28125152317705815, 'targets': 0,'stops': 1,'missing': 1, 'entryexit': 0,'recon': 0}, {'s': 'BTC', 'tf': '1Hour', 'd': 0.25, 'train': False,'slipbp': 30,'return': -3.1042102151150663, 'DD': 11.504638350105179, 'win': 33.33333333333333, 'N': 3, 'open': 1,'mean': -0.05313709484419856, 'targets': 0,'stops': 1,'missing': 1, 'entryexit': 0,'recon': 0}]. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 and 2025 to train, and data from January 1 to September 11, 2026, to test. The decision variable 'd' was set to -1, which means an immediate breakout, otherwise an ATR pullback depth was used. ATR stands for Average True Range, a measure of price volatility. Fees of 25 basis points were credited to each side of the trade. The results include return, drawdown, mean, and win percentage. The trades were marked as open but not closed, and the results are not pristine prospective out-of-sample. The results are as follows: the experiment used Bitcoin, a 1-hour time frame, and different slip values. The results for each scenario are: return of -55.00102646384883, drawdown of 55.54518118705916, win percentage of 25, and mean of -2.82008996360581 for the first scenario, return of -56.818599254851485, drawdown of 57.25554157464396, win percentage of 25, and mean of -2.9788461481517587 for the second scenario, return of 5.772356415511526, drawdown of 14.746844019014837, win percentage of 50, and mean of 1.1155628134529492 for the third scenario, and return of 4.406458723565176, drawdown of 15.585826085129806, win percentage of 50, and mean of 0.912104920277218 for the fourth scenario. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 to 2025 for training and data from January 1 to September 11, 2026, for testing. The decision rule was to immediately stop the trade if a certain condition was met, otherwise to use a method called ATR pullback depth, which is a way to measure how far a price moves away from a recent high or low. The fees were 25 basis points, which is a measure of the cost of a trade, credited to each side of the trade. The results include return, which is the profit or loss, drawdown, which is the largest loss, and mean, which is the average return. The results are as follows: the experiment was not completely open and not completely prospective out of sample, which means it was not a perfect test. The results are: [{"s":"BTC","tf":"1Day","d":1,"train":true,"slipbp":10,"return":-11.650904804174534,"DD":13.116749878670774,"win":0,"N":2,"open":0,"mean":-6.003788527160814,"targets":0,"stops":2,"missing":0,"entryexit":0,"recon":-1.1102230246251565e-16},{"s":"BTC","tf":"1Day","d":1,"train":true,"slipbp":30,"return":-11.984104837031605,"DD":13.271483400636065,"win":0,"N":2,"open":0,"mean":-6.181213213948899,"targets":0,"stops":2,"missing":0,"entryexit":0,"recon":0},{"s":"BTC","tf":"1Day","d":1,"train":false,"slipbp":10,"return":0.9484471663183003,"DD":7.485945722691223,"win":100,"N":1,"open":0,"mean":0.9484471663183003,"targets":0,"stops":0,"missing":0,"entryexit":0,"recon":-2.220446049250313e-16},{"s":"BTC","tf":"1Day","d":1,"train":false,"slipbp":30,"return":0.5454581467251307,"DD":7.485945722691245,"win":100,"N":1,"open":0,"mean":0.5454581467251307,"targets":0,"stops":0,"missing":0,"entryexit":0,"recon":0}]. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 and 2025 to train, and data from January 1 to September 11, 2026, to test. The rules were to immediately stop trading if a certain condition was met, otherwise to use a method called ATR pullback depth, which is a way to measure how far a price might move. The experiment also included fees of 25 basis points, which is a way to measure a small percentage, credited to each side of a trade. The results included return, which is the profit or loss, drawdown, which is the largest loss, and mean, which is the average. The results also included win percent, which is the percentage of trades that were profitable. The results were marked as open, meaning not all trades were closed, and not pristine prospective out of sample, which means the results were not perfect and were based on data that was not used to train the experiment. The results were as follows: the experiment used Bitcoin, with a time frame of 1 day, and a parameter called d of 0.5. The results for the training data were a return of -17.012274977348362, a drawdown of 26.243268350855764, and a win percent of 20. The results for the testing data with a slip of 10 basis points were a return of 6.846405475014472, a drawdown of 7.772299771053159, and a win percent of 66.66666666666666. The results for the testing data with a slip of 30 basis points were a return of 5.571905381068043, a drawdown of 7.9562034604428895, and a win percent of 33.33333333333333. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json

A news scan found the headline 'Sep12 frozen experiment'. This refers to an experiment that was stopped on September 12. The experiment used data from 2024 to 2025 for training and data from January 1 to September 11, 2026, for testing. The experiment had a rule to immediately stop trading if a certain condition was met, otherwise it used a method called ATR pullback depth, which is a way to measure how much a price moves away from a recent high or low. The experiment also accounted for fees of 25 basis points, which is a way to measure a small percentage, credited to each side of a trade. The results of the experiment are shown in a list of four sets of data. Each set of data includes information about the stock symbol, which is BTC, the time frame, which is 1 day, and other metrics such as return, drawdown, and win percentage. The list is as follows: [{"s":"BTC","tf":"1Day","d":0.25,"train":true,"slipbp":10,"return":-30.591041436760502,"DD":36.47165802126271,"win":12.5,"N":8,"open":0,"mean":-4.1491155432801765,"targets":1,"stops":7,"missing":0,"entryexit":0,"recon":0},{"s":"BTC","tf":"1Day","d":0.25,"train":true,"slipbp":30,"return":-26.15732617303911,"DD":32.27854782377374,"win":14.285714285714285,"N":7,"open":0,"mean":-3.8674579243320926,"targets":1,"stops":6,"missing":0,"entryexit":0,"recon":0},{"s":"BTC","tf":"1Day","d":0.25,"train":false,"slipbp":10,"return":9.40739781418447,"DD":7.772299771053159,"win":100,"N":3,"open":0,"mean":3.1106474421579877,"targets":0,"stops":0,"missing":0,"entryexit":0,"recon":0},{"s":"BTC","tf":"1Day","d":0.25,"train":false,"slipbp":30,"return":8.102349336674198,"DD":7.9562034604428895,"win":66.66666666666666,"N":3,"open":0,"mean":2.699026858694546,"targets":0,"stops":0,"missing":0,"entryexit":0,"recon":0}]. https://runway.adityaperswal.workers.dev/data/crypto-bars ; C91204 research/results.json