Flippy grades its own flips
Every scheduled flip is logged the moment it's announced, then re-priced against the live market every hour for 24 hours: did the projected sell price actually fill? Could you still exit at a profit? These are those results — including the losses, with flips whose market went quiet counted as losses, not quietly dropped. Sites that charge for flip signals don't show you this table. That's the point of this one.
9581 flips logged so far — 116 still inside their 24h grading window. Generated 11/1/2026, 11:06:12 PM.
Promised vs delivered
Quoted spreads shrink between the call and the fill — everyone's do. Most trackers only show the promise; Flippy measures what actually survives, publishes the gap here, and ranks every call by the calibrated number instead of the raw spread.
All flips
| Checkpoint | Flips | Win rate | Target filled by then | Projected ROI | Realized ROI |
|---|---|---|---|---|---|
| 1h after posting | 7603 | 66.9% | 4.3% | 3.14% | 0.77% |
| 2h after posting | 7615 | 65.6% | 9.2% | 3.14% | 0.7% |
| 3h after posting | 7610 | 61.3% | 14.2% | 3.14% | 0.53% |
| 6h after posting | 7598 | 60.5% | 26.8% | 3.14% | 0.51% |
| 9h after posting | 7595 | 61.8% | 34.4% | 3.14% | 0.52% |
| 12h after posting | 7592 | 62.2% | 39.8% | 3.14% | 0.56% |
| 18h after posting | 7582 | 62.6% | 46.9% | 3.14% | 0.49% |
| 24h after posting | 7566 | 63.5% | 51.6% | 3.14% | 0.43% |
Low & high risk tiers
What Flippy is learning
• LOW risk margins are not holding: predicted 3.71% ROI but realized 0.88% at 12h. The announced spreads may be closing before fills — calibration is already discounting this tier's ranking accordingly.
• HIGH risk margins are not holding: predicted 2.64% ROI but realized 0.27% at 12h. The announced spreads may be closing before fills — calibration is already discounting this tier's ranking accordingly.
• Near-misses rejected only by the minimum total-profit gate are winning anyway (64.1% at 12h over 3979 shadow trades, avg 0.76% ROI) — that threshold may be stricter than the market requires. Auto-tune tests a bounded looser setting when the evidence is strong enough to be worth real gp, and reverts it if the flips it admits stop paying.
• Near-misses rejected only by the hourly-volume bar are winning anyway (57.2% at 12h over 1370 shadow trades, avg 0.15% ROI) — that threshold may be stricter than the market requires. Auto-tune tests a bounded looser setting when the evidence is strong enough to be worth real gp, and reverts it if the flips it admits stop paying.
• Near-misses rejected only by the falling-price guard are winning anyway (71.4% at 12h over 210 shadow trades, avg 1.51% ROI) — that threshold may be stricter than the market requires. This one stays a human call — auto-tune never loosens it on its own.
• Across all calls, realized ROI peaks at the 1h checkpoint (0.77% avg) — that is roughly the holding window your filters are tuned for.
Every flip, graded
The raw log behind the tables above — every scheduled flip, newest first. The proof link opens Flippy’s public Discord announcement: Discord timestamps can’t be edited, so a flip can’t be backdated to fit a price move. — means a checkpoint hasn’t come due yet.
| Posted | Item | Risk | Buy → sell | Projected ROI | 1h | 6h | 12h | 24h | Proof |
|---|---|---|---|---|---|---|---|---|---|
| Nov 1, 10:27 PM | Dragon dart | LOW | 1,061 → 1,157 | 6.88% | — | — | — | — | announcement ↗ |
| Nov 1, 10:27 PM | Tormented synapse | HIGH | 30.77m → 31.91m | 1.61% | — | — | — | — | announcement ↗ |
| Nov 1, 10:27 PM | Abyssal bludgeon | HIGH | 13.00m → 13.47m | 1.57% | — | — | — | — | announcement ↗ |
| Nov 1, 10:21 PM | Granite hammer | HIGH | 4.54m → 4.73m | 2.12% | — | — | — | — | announcement ↗ |
| Nov 1, 10:21 PM | Elidinis' ward | HIGH | 2.98m → 3.08m | 1.17% | — | — | — | — | announcement ↗ |