On this page (8 sections)
We took every full hour from 18 August to 16 November as a starting point on each coin, 256,460 in all, and followed price for up to three days.
Measure the distance in daily ranges
A 3 percent target is a different bet on every coin. We use the average daily range (ADR) of the last 14 days: each daily candle’s high minus its low, divided by its close, then averaged. Over 5 to 18 November it was 4.1 percent on BTC, 5.8 on ETH, 6.0 on SOL and 10.6 on ZEC, so 3 percent was about three quarters of a day’s range on BTC and under a third on ZEC.
| Per 100 starts (up / down) | Touched within 4h | Within 24h | Within 72h |
|---|---|---|---|
| A quarter ADR away | 24.4 / 25.0 | 58.7 / 64.7 | 74.1 / 82.9 |
| Half an ADR | 7.4 / 7.0 | 33.8 / 37.2 | 54.0 / 65.1 |
| One ADR | 1.5 / 1.4 | 12.6 / 12.2 | 28.7 / 36.1 |
| Two ADR | 0.3 / 0.2 | 3.7 / 1.8 | 11.3 / 7.3 |
Half an ADR away, a level was touched within a day about once in three starts; a full ADR away, about once in eight. Price that got there within three days took its time, a median of about 17 hours for half an ADR up and 28 for a full one, 20 and 34 going down.
We expected altcoins to reach far levels more often than BTC. In ADR terms they didn’t.
| Within 24h, per 100 (up / down) | Half an ADR | One ADR |
|---|---|---|
| BTC | 34.7 / 41.8 | 12.7 / 14.7 |
| ETH | 35.7 / 41.7 | 11.8 / 15.8 |
| 118 altcoins | 33.8 / 37.1 | 12.6 / 12.2 |
Resampling whole days put the all-coin upward figure at one ADR between 10.6 and 15.0 in 95 of 100 resamples, a range BTC and ETH both fall inside.
The same distances in percent
Switch back to percent and the coins look like different markets.
| Above the start, per 100 (within 24h / 72h) | BTC | 118 altcoins |
|---|---|---|
| 1 percent | 49.9 / 72.1 | 78.5 / 86.6 |
| 2 percent | 23.0 / 46.9 | 61.7 / 75.3 |
| 5 percent | 6.3 / 15.9 | 30.8 / 49.8 |
| 10 percent | 0.7 / 3.0 | 12.1 / 27.0 |
A 2 percent target came within a day 23 times in 100 on BTC and 62 on the altcoins. Most of that gap is the ADR: over the window the median altcoin range was 8.6 percent a day, BTC’s 2.8 and ETH’s 4.0.
Which comes first, target or stop
Reaching the target only helps if the stop isn’t hit first. Here a long’s stop sits a quarter ADR below the start and the target one, two or three times as far above (1R, 2R, 3R), and shorts mirror it. A case is settled once one of the two is hit, and within 72 hours fewer than 6 in 100 touched neither.
If every small step were a coin toss, a target k times as far as the stop would come first once in 1 + k cases.
| Target first, per 100 settled cases | Long | Short | Long and short together | Coin toss |
|---|---|---|---|---|
| 1R | 47.2 | 52.8 | n/a | 50.0 |
| 2R | 30.4 | 36.8 | 33.6 | 33.3 |
| 3R | 21.7 | 27.9 | 24.8 | 25.0 |
Long and short together, which cancels the window’s drift, matched the coin toss, and resampling whole days kept 2R between 32.9 and 34.3 in 95 of 100 resamples. At 1R a long’s target is a short’s stop, so the two always add up to 100; the 47 and 53 mostly reflect a window in which price was higher a day later 45.6 times in 100.
A random hour, long or short, reached 2R before the stop about once in three. That is the rate at which a 2R plan breaks even before fees (the risk/reward calculator shows how often the plan must work at any R), so a random entry gets you to break-even and no further.
Trend and weekends
Split the window at 2 October and the long and short columns trade places. In the first half BTC rose from about 64,500 to 84,800 and the 2R long came first 35.8 times in 100 settled cases, the short 29.6. In the second half BTC fell back to 65,300 and it was 25.0 against 43.8. One ADR within 24 hours went from 16.5 up and 8.6 down to 8.6 up and 15.9 down.
On 10 November at 00:00 UTC BTC opened at 75,327 with an ADR of 3.08 percent, about 2,320 points. The level one ADR below, 73,007, was touched before 16:00 UTC that day. The one above, 77,647, never was: the next 72 hours topped at 75,411. SOL and ETH did the same, touching the lower level before 14:00 UTC.
We didn’t test whether a chart’s trend catches this drift in advance; the nearest thing you have is the structure on your timeframe and the one above it.
Weekends were quieter. From a weekday start, a level one ADR away was touched within 24 hours 13.9 times in 100 going up and 13.5 going down; from a Saturday or Sunday start, 9.2 and 9.0. These are UTC days (see the candle close table), so 19:00 on a Friday in New York in winter already counts as Saturday.
Moving your stop to break-even after 1R
Here the stop is half an ADR away. We took the 109,177 longs that reached 1R before the stop and ran each one twice.
With the stop moved to the entry, 42.7 in 100 went on to 2R and 45.5 came back to the entry first, plus 2.6 that touched it in the same candle as 1R or 2R, too close to call. Left where it was, the stop was hit 23.8 times and 2R came 51.8 times. Counting only trades that ended one way or the other, that is 48.4 against 68.5, close to the coin toss’s 50 and 66.7.
Break-even didn’t make 2R more likely. It turned the 23.8 stop-outs into flat exits, and did the same to about 9 in 100 that touched the entry on their way to 2R and about 15 that would otherwise still have been open after three days. Only 9.1 touched neither the entry nor 2R in that time. The invalidation level guide covers where the original stop belongs.
Checking your next target
- Divide the distance to your target by the ADR. SOL’s was about 6 percent in mid-November, so a target 3 percent away is half an ADR. Across our 120 coins, levels that far up were touched within 24 hours about 34 times in 100, and 54 within three days.
- Give it a deadline. Our numbers stop at 72 hours. A full ADR, when it came, took a median of 28 hours going up and 34 going down.
- Put it next to the stop. With the stop a quarter ADR away, about 1.5 percent, that target is 2R and must come first more than once in three, before fees. Size from the stop with the position size calculator and keep liquidation beyond it.
- Then look at the trend and the day. In the falling half a 2R long came first 25 times in 100 settled cases, and weekend starts reached one ADR about a third less often.
Doing this in the bot
Send a pair and timeframe to the bot, say SOL 1h, and the read gives an entry zone, the targets and the invalidation level. If even the farthest target sits closer to the entry than the invalidation does, the read is set up to say “No clean entry right now” and name what would create a plan, rather than force one. A plan like that would need its target first more than half the time just to break even.
Watch levels pings you when price touches one of those levels and, on a 1h read, keeps watching for three days, the span of our numbers. It counts a level as broken only on a candle close, and a target alert won’t tell you whether your entry filled. 📍 I’m in this trade adds your own entry and stop, and its stop alert shows the result in R (see how Watch works).
How we measured
- Window: hourly starts from 18 August to 16 November 2026 (90 days, UTC), on 15-minute candles from 4 August to 19 November, none missing.
- Universe: Binance USDT-M perpetuals of COIN type (PAXG, a gold token, included), stablecoins excluded, listed before 4 August and cut at 120 by average daily volume over 11 to 17 August (the last about 11.6 million USDT a day). EPIC and STORJ count until trading stopped; contracts already dropped from Binance’s list are missing.
- Starts and touches: each start is priced at the open of the hour’s first 15-minute candle (2,160 starts each for BTC and ETH, 252,140 for the altcoins), with the ADR from the 14 UTC days before. A level is touched when a 15-minute high or low reaches it, start candle included; target and stop in one candle (at most 0.2 in 100) count as unknown.
- Checks: 1,000 resamples of whole coins and 2,000 of whole days (ranges quoted are the middle 95 in 100 of the day resamples), coins weighted equally (one ADR within 24 hours: 12.6 up, 12.4 down), the ten most volatile altcoins left out (12.3, 12.5), the two halves, weekdays against weekends.
- Limits: one exchange’s last price, 15-minute candles, no fees, slippage or funding; hourly starts overlap in time, hence the resampling by day. Base rates for a random hour, not the odds of a particular setup or a tested trading rule.
- Reproduce: Binance’s public futures endpoints exchangeInfo and klines (1d for the ranking, 15m for the rest).
Today’s trend and key levels: Bitcoin (BTC) · Solana (SOL) · Ethereum (ETH)