Sun Soft
Sun Soft
Where the rules and numbers on these pages came from. Each card is a test on past prices, with the data it used and the date it was checked. Past results do not promise the future.
The prop test on most cards: a $1,000 account fails at -3% ($970) and passes at +12% ($1,120), within a year. It was started fresh every third day from August 2018 to September 2025 (873 starts) on BTC 4-hour candles, with a 0.04% cost per trade.
It is a coin flip. BTC was higher a week later only half the time, and it fell further first more often than not.
| One-day drop | Days | Higher a week later | Fell 5%+ more within a week | Fell 10%+ more |
|---|---|---|---|---|
| Over 5% | 134 | 68 (51%) | 81 (60%) | 41 (31%) |
| Over 8% | 44 | 22 (50%) | 31 (70%) | 19 (43%) |
13 days fell over 5%. A week later 6 were higher; within the week 11 fell another 5% or more, and 4 fell another 10% or more.
| Day | That day | Lowest in the next week | A week later |
|---|---|---|---|
| 2026-06-02 | −6.5% | −11.4% | −7.5% |
| 2026-02-05 | −13.9% | −3.8% | +5.3% |
| 2026-01-29 | −5.2% | −26.3% | −25.6% |
| 2025-11-20 | −5.4% | −6.8% | +5.5% |
| 2025-11-14 | −5.2% | −14.7% | −10.0% |
| 2025-10-10 | −7.1% | −8.3% | −5.8% |
For a prop accountThere is only $30 of room before a fail. Put $600 in after a drop and another 5% fall ends the account, which happened 6 times in 10. So the rule does not buy the bottom; it buys back when the trend turns on again.
DataBTC 4-hour candles including intraday lows, January 2018 to October 13, 2026. Drops on back-to-back days count once.
Open Prop trading →Aggressive + 4-hour trend: passed 64%, failed 0%.
For a prop accountBuying without a maximum failed 3 times in 4. Only the methods that sit out while the signal is off failed 0%.
DataBTC 4-hour candles, 873 starts from August 2018 to September 2025, the same prop test as at the top.
Open Prop trading →With the 4-hour trend, checking twice a day still failed 0%.
| Checks a day | Buy with the Aggressive signal | Aggressive + 4-hour trend |
|---|---|---|
| Every 4 hours (6) | Pass 55% · fail 0% | Pass 64% · fail 0% |
| 4 times | Pass 55% · fail 0% | Pass 65% · fail 0% |
| 3 times | Pass 51% · fail 14% | Pass 63% · fail 0% |
| Twice, 12 hours apart | Pass 52% · fail 14% | Pass 58% · fail 0% |
For a prop accountThe Aggressive rule alone failed 14% when checked 3 times a day or fewer. With the 4-hour trend it did not fail at any of these, but checking more often passed more.
DataBTC 4-hour candles, 873 starts from August 2018 to September 2025, the same prop test as at the top. Check times are counted from the daily close (midnight UTC).
It was picked on the earlier years and then run on later years it had not seen. It still failed 0%.
| Method | Starts Aug 2018 – Feb 2022 (picked here) | Starts Mar 2022 – Sep 2025 (not seen) |
|---|---|---|
| Aggressive + 4-hour trend | Pass 70% · fail 0% | Pass 59% · fail 0% |
| Buy with the Aggressive signal | Pass 62% · fail 0% | Pass 47% · fail 0% |
Change the line length and it still fails 0%, so the result is not tuned to the number 20.
| 4-hour trend line length | Starts Aug 2018 – Feb 2022 (picked here) | Starts Mar 2022 – Sep 2025 (not seen) |
|---|---|---|
| 10 days | Pass 64% · fail 0% | Pass 54% · fail 0% |
| 20 days | Pass 70% · fail 0% | Pass 59% · fail 0% |
| 30 days | Pass 64% · fail 0% | Pass 59% · fail 0% |
| 40 days | Pass 64% · fail 0% | Pass 54% · fail 0% |
For a prop accountThe later years passed less (59%). It is safer to expect less than 64% from here.
DataBTC 4-hour candles, 873 starts from August 2018 to September 2025, the same prop test as at the top.
A bigger account changes nothing. Holding more than the page says, or shorting, made it worse.
| How much to hold | Starts Aug 2018 – Feb 2022 (picked here) | Starts Mar 2022 – Sep 2025 (not seen) |
|---|---|---|
| 20% more than the page says | Pass 64% · fail 19% | Pass 62% · fail 0% |
| What the page says | Pass 70% · fail 0% | Pass 59% · fail 0% |
| 15% less | Pass 63% · fail 0% | Pass 58% · fail 0% |
| 25% less | Pass 62% · fail 0% | Pass 57% · fail 0% |
For a prop accountEvery amount is a share of the account (loss line, target, maximum, alert price, costs), so a $10,000 or $50,000 account gets the same results as $1,000, as long as its rules are the same. Holding just 20% more than the page says failed 1 time in 5 for 2018–22 starts. Holding less kept failures at 0% and passed only a little less. Shorting never failed but passed far less (64% → 50%).
DataBTC 4-hour candles, 873 starts from August 2018 to September 2025, the same prop test as at the top.
About half of all starts had passed by 6 months. Of the passes, the middle one took 114 days.
Share of all starts that had passed by then (Aggressive + 4-hour trend, checked every 4 hours)
For a prop accountThe middle half of passes took 46 to 187 days. Trying to finish in a month or two means holding more, and that is how accounts failed (see the card above). Checking twice a day: 26% by 3 months, 58% by a year.
DataBTC 4-hour candles, 873 starts from August 2018 to September 2025, the same prop test as at the top.
BTC + ETH was best. A third coin added crash risk more than speed.
For a prop accountA third coin sometimes helped a little and sometimes caused failures, because quiet-looking coins broke in a single day (LINK −47% in the COVID crash, BNB −19% at FTX). Picking the calmest coins each day failed 25–29%. On crash days coins fall together (2022-08-19: BTC −10%, XRP −10%), so mixing more does not cut the risk.
DataCoin Metrics daily prices for 13 coins, 829 starts from August 2018 to May 2025, checked once a day at the close, so the pass rates are lower than on the other cards.
Open Prop trading →No. The results were the same.
| Checks a day | No limit | Up to the account size |
|---|---|---|
| Every 4 hours (6) | Pass 64% · fail 0% | Pass 64% · fail 0% |
| Twice, 12 hours apart | Pass 58% · fail 0% | Pass 58% · fail 0% |
For a prop accountIn the past the maximum almost never went above the account. It can now, when BTC is calm and profit has built up.
DataBTC 4-hour candles, 873 starts from August 2018 to September 2025, the same prop test as at the top.