The studies
Every number on this page is a count of what XRP price already did, measured from the same stretch of candles on every read. Team Nothing XRP Pulse Charting runs about fifty live feeds and draws the levels its members watch — so this page grades them in the open rather than asserting that they work. Nothing here is a forecast, and nothing here is advice.
Every call this app has published, counted as it resolved. Long and short are kept as two separate books rather than one blended number, because a rule that works on one side and not the other is a different thing from a rule that works.
One line: a hit reached its target before its stop, a miss hit its stop first, and a timeout is counted in the score rather than dropped out of it — counts of calls already resolved, never a forecast.
For every event a feed called out, how far price moved afterwards, set against the same move measured over every hour in the window. The last number on each line is the one to read: the window's own busy quarter is the line, so a feed marking nothing lands near 25%.
1 hour after
0.28% moved · typical hour 0.27%
36% above the busy quarter (25% by chance)
3 hours after
0.34% moved · typical hour 0.37%
29% above the busy quarter (25% by chance)
the most recent 400 transfers the feed called out. Every large transfer the app posted, timed to the moment it was called out, measured against the same window: how far price moved in the hour and three hours after.
1 hour after
0.30% moved · typical hour 0.27%
35% above the busy quarter (25% by chance)
3 hours after
0.34% moved · typical hour 0.37%
29% above the busy quarter (25% by chance)
the 244 prints of 1M XRP or more among them. The same measurement, restricted to the biggest transfers, on the theory that size is what carries information rather than the count.
Typical hour over the window: 0.27% move at the middle, with the busiest quarter of hours moving more than 0.45%. Three hours: middle 0.37%, busy quarter above 0.77%. Events inside the same hour are counted one by one, so a cluster reads as several. Feeds measured against 299 hours of hourly candles, from 10/23/2026 to 11/5/2026.
The ledger's own escrow activity — XRP locked up, released and handed back — measured the same way as the feeds above: how far price moved in the hour and the three hours after the transaction landed, set against the same move over every hour in the window. Each line is one kind of transaction, because a lockup and a release are not the same event and averaging the two would hide both.
1 hour after
0.13% moved · typical hour 0.27%
21% above the busy quarter (25% by chance)
3 hours after
0.20% moved · typical hour 0.37%
15% above the busy quarter (25% by chance)
33 escrow creates, finishes and cancels inside the candle window. Every escrow transaction on the app’s own ledger record, dated by the ledger itself rather than by when the app noticed it, so the anchor is the moment the transaction landed.
1 hour after
0.33% moved · typical hour 0.27%
20% above the busy quarter (25% by chance)
3 hours after
0.47% moved · typical hour 0.37%
13% above the busy quarter (25% by chance)
15 creates among them. XRP locked up and scheduled to release later. The same measurement as the feed rows: how far price moved in the hour and the three hours after the lockup landed.
1 hour after
0.09% moved · typical hour 0.27%
20% above the busy quarter (25% by chance)
3 hours after
0.20% moved · typical hour 0.37%
13% above the busy quarter (25% by chance)
15 finishes among them. XRP released out of escrow into circulation, which is the transaction people expect to matter most. The same measurement again, so it can be read straight against the lockups rather than argued about.
1 hour after
0.02% moved · typical hour 0.27%
33% above the busy quarter (25% by chance)
3 hours after
0.20% moved · typical hour 0.37%
33% above the busy quarter (25% by chance)
3 cancels among them. XRP handed back to whoever locked it instead of released. The rarest of the three, so this row usually carries the thinnest sample on the page and will report its size rather than a number.
1 hour after
— moved · typical hour 0.27%
— above the busy quarter (25% by chance)
3 hours after
— moved · typical hour 0.37%
— above the busy quarter (25% by chance)
0 of the largest creates on the record. Size is the one thing a ledger row carries about intent, so the biggest lockups are graded on their own rather than blended in with the hundreds of small ones around them.
33 of the 861 escrow events on the app's own record fall inside the 299 hours of hourly candles that a single exchange request reaches, and those are the ones graded here: the rest are older than the tape this study can ask for in one call. Every event is dated by the ledger itself rather than by when the app noticed it, which is what puts the anchor on the moment the transaction landed instead of the moment a sweep found it. Escrows arrive in batches — a scheduled lockup can land as dozens of creates inside a single hour — so a row's graded count is how many transactions were measured, not how many separate moments it stands on, and a cluster reads as several. Under 20 graded events a row is flagged as too thin to read, and its sample size is the number to look at rather than the shares beside it.
The levels the terminal draws — saved order-book walls, the VWAP band, the Fibonacci ladder — walked forward and graded level by level. A level "turned" when price turned at it, and "went through" when price closed past it. Anything that had not answered inside the window is counted as unanswered, never quietly scored as a miss.
654 turned · 805 went through · 20 unanswered inside the window · 600 walls off 200 saved book reads, 21 too new to grade
The deepest cells of every saved book read, each graded on the candles that followed it: a turn is price 0.35% away before a close 0.15% past the wall.
30 turned · 3 went through · 33 pushes outside the ±1σ band, across 4 anchored bands
Anchored VWAP with one-sigma bands, re-anchored at the start of each UTC day and run from that day's first candle. Graded on whether a close outside the band came back inside within 3 hours; a push that stayed out counts against it.
27 turned · 23 went through · 6 unanswered inside the window · 10 ladders, each drawn off the 12 hours of price behind it
The 5 retracements the card prints between the two ends of the swing (23.6 / 38.2 / 50 / 61.8 / 78.6%), each drawn from one stretch of price and graded only on the stretch that followed it: a turn is price 0.35% away from the level before a close 0.15% past it.
Graded over 288 candles (15m back). Price reached a level and then either turned or went through: a turn is 0.35000000000000003% away before a close 0.15% past it, and anything still unanswered after 3h counts as unanswered. Under 20 graded tests a percentage says nothing, so a row reports its sample size instead of a number.
The day after
3.07% moved · usual window 2.09%
33% above the busy quarter (25% by chance)
3 days after
3.55% moved · usual window 3.63%
25% above the busy quarter (25% by chance)
7 days after
6.14% moved · usual window 6.05%
17% above the busy quarter (25% by chance)
24 monthly unlocks graded, from 2024-12-01 to 2026-11-01. Every one of the 32 releases scheduled across Ripple's escrow wallets falls on the 1st of a month, which is why each unlock is measured from that day's close. Unlocks are counted as they come due, so a month is a month here and not a run of separate transactions.
Every reading this terminal makes carries a confidence, and every one of them is graded against what price did. This compares the two. It is the question a scoreboard usually avoids, and the answer is printed here whether it flatters the desk or not. Under 10 graded readings a row prints its sample size and no percentage.
285 readings on file, anchored between 10/25/2026, 2:45:24 AM and 11/4/2026, 11:16:10 AM. This record is young: it grows with every reading the desk makes, and the horizon rows fill in as their own windows come due.
By horizon
Called it
46%
79 graded readings
We said
45%
the average confidence these readings carried
Said vs did
+1
did better than it claimed
The price it projected landed 0.09% of the anchor away from the price at the horizon, averaged over the graded readings, of which 72 were flat reads and 7 directional.
Called it
51%
73 graded readings
We said
46%
the average confidence these readings carried
Said vs did
+5
did better than it claimed
The price it projected landed 0.17% of the anchor away from the price at the horizon, averaged over the graded readings, of which 69 were flat reads and 4 directional.
Called it
64%
64 graded readings
We said
48%
the average confidence these readings carried
Said vs did
+16
did better than it claimed
The price it projected landed 0.21% of the anchor away from the price at the horizon, averaged over the graded readings, of which 61 were flat reads and 3 directional.
Called it
43%
53 graded readings
We said
47%
the average confidence these readings carried
Said vs did
-4
claimed more than it delivered
The price it projected landed 0.36% of the anchor away from the price at the horizon, averaged over the graded readings, of which 48 were flat reads and 5 directional.
Called it
73%
11 graded readings
We said
60%
the average confidence these readings carried
Said vs did
+13
did better than it claimed
The price it projected landed 1.29% of the anchor away from the price at the horizon, averaged over the graded readings, of which 5 were flat reads and 6 directional.
Called it
4 readings
under 10 graded, so no percentage is printed
We said
60%
the average confidence these readings carried
Said vs did
—
not enough graded to compare
The price it projected landed 2.75% of the anchor away from the price at the horizon, averaged over the graded readings.
By the read it made
Called it
4 readings
under 10 graded, so no percentage is printed
We said
50%
the average confidence these readings carried
Said vs did
—
not enough graded to compare
The price it projected landed 0.22% of the anchor away from the price at the horizon, averaged over the graded readings.
Called it
68%
25 graded readings
We said
55%
the average confidence these readings carried
Said vs did
+13
did better than it claimed
The price it projected landed 0.96% of the anchor away from the price at the horizon, averaged over the graded readings.
Called it
52%
255 graded readings
We said
46%
the average confidence these readings carried
Said vs did
+6
did better than it claimed
The price it projected landed 0.2% of the anchor away from the price at the horizon, averaged over the graded readings, of which 255 were flat reads and 0 directional.
By the confidence carried
Called it
49%
225 graded readings
We said
45%
the average confidence these readings carried
Said vs did
+4
did better than it claimed
The price it projected landed 0.18% of the anchor away from the price at the horizon, averaged over the graded readings, of which 216 were flat reads and 9 directional.
Called it
65%
51 graded readings
We said
55%
the average confidence these readings carried
Said vs did
+10
did better than it claimed
The price it projected landed 0.44% of the anchor away from the price at the horizon, averaged over the graded readings, of which 39 were flat reads and 12 directional.
Called it
8 readings
under 10 graded, so no percentage is printed
We said
62%
the average confidence these readings carried
Said vs did
—
not enough graded to compare
The price it projected landed 1.8% of the anchor away from the price at the horizon, averaged over the graded readings.
Read this the way it is written: these are counts of what already happened, over one stretch of price, and a share of past moves tells you nothing about the next one. Educational market analytics, not financial advice, and no part of this page is a recommendation.