AI · Berkshire Mirror — Track record
Every elapsed decision is graded against what prices actually did — hit-rates and errors per signal family, plus the cost of the decisions this book chose not to take.
315 graded 0 pending
1W calibration -0.49 How far along the 1-week path a decision typically gets: the fraction of the decided move realized after one week, measured across this book's graded decisions. Feeds a bounded conviction adjustment back into future decisions.
Scorecard
Hit-rates per signal family.
| Signal family | Dir | n | Hit 1W | Hit 4W | Hit 12W | Med err 1W | Stop-first | Bias | Status |
|---|---|---|---|---|---|---|---|---|---|
| FailedBreakout_Down_20D | sell | 101 | 55% | 47% | — | — | 0% | +1 | ● active |
| NoTrade | sell | 65 | 45% | 43% | — | — | 0% | -1 | ● active |
| RallySell_BearTrend | sell | 24 | 70% | 65% | — | — | 0% | +4 | ● active |
| PullbackBuy_VWAP | buy | 15 | 7% | 7% | — | -6.8% | 87% | -9 | ● active |
| FailedBreakdown_Up_20D | sell | 13 | 23% | 15% | — | — | 0% | -5 | ● active |
| NoTrade (L2 stop-break basis) | sell | 12 | 40% | 43% | — | — | 0% | -2 | ● active |
| BreakdownDown_20D | sell | 9 | 44% | 44% | — | — | 0% | -1 | ● active |
| ReversalUp_Oversold (L2 stop-break basis) | sell | 9 | 56% | — | — | — | 0% | +1 | ● active |
| MomentumContinuation_Up | buy | 7 | 43% | 0% | — | -3.6% | 43% | -1 | ● active |
| BreakdownDown_52W | sell | 6 | 33% | 100% | — | — | 0% | -3 | ● active |
| AnticipationUp_VCP | sell | 5 | 60% | 60% | — | — | 0% | +2 | ● active |
| AnticipationUp_VCP | buy | 4 | 0% | 0% | — | -11.9% | 100% | 0 | ○ building sample (4/5) |
| BreakoutUp_20D | sell | 4 | 25% | 25% | — | — | 0% | 0 | ○ building sample (4/5) |
| BreakoutUp_52W | buy | 4 | 0% | 0% | — | -5.5% | 50% | 0 | ○ building sample (4/5) |
| BreakoutUp_20D | buy | 3 | 33% | 100% | — | -2.2% | 0% | 0 | ○ building sample (3/5) |
| BreakoutUp_52W | sell | 3 | 67% | 67% | — | — | 0% | 0 | ○ building sample (3/5) |
| BreakdownDown_20D (L2 stop-break basis) | sell | 2 | 0% | — | — | — | 0% | 0 | ○ building sample (2/5) |
| FailedBreakdown_Up_20D (profit-take armed+c) | sell | 2 | 100% | — | — | — | 0% | 0 | ○ building sample (2/5) |
| FailedBreakdown_Up_20D | buy | 2 | 0% | 0% | — | +0.3% | 50% | 0 | ○ building sample (2/5) |
| NoTrade (profit-take armed+b) | sell | 2 | 0% | — | — | — | 0% | 0 | ○ building sample (2/5) |
| NoTrade (profit-take armed+b+c) | sell | 2 | 50% | 0% | — | — | 0% | 0 | ○ building sample (2/5) |
| NoTrade (profit-take armed+c+d) | sell | 2 | 0% | — | — | — | 0% | 0 | ○ building sample (2/5) |
| OversoldBounce_Up (L2 stop-break basis) | sell | 2 | 100% | — | — | — | 0% | 0 | ○ building sample (2/5) |
| RallySell_BearTrend (L2 stop-break basis) | sell | 2 | 50% | 100% | — | — | 0% | 0 | ○ building sample (2/5) |
| ReversalUp_Oversold | sell | 2 | 50% | 50% | — | — | 0% | 0 | ○ building sample (2/5) |
| FailedBreakdown_Up_20D (profit-take armed+a) | sell | 1 | 0% | — | — | — | 0% | 0 | ○ building sample (1/5) |
| FailedBreakdown_Up_2 | sell | 1 | 0% | 0% | — | — | 0% | 0 | ○ building sample (1/5) |
| FailedBreakout_Down_20D (L2 stop-break basis) | sell | 1 | 0% | 0% | — | — | 0% | 0 | ○ building sample (1/5) |
| MomentumContinuation_Up (profit-take armed+a) | sell | 1 | 100% | — | — | — | 0% | 0 | ○ building sample (1/5) |
| MomentumContinuation_Up | sell | 1 | 100% | 100% | — | — | 0% | 0 | ○ building sample (1/5) |
| NoSignal (profit-take armed+b+c) | sell | 1 | 100% | — | — | — | 0% | 0 | ○ building sample (1/5) |
| NoSignal | sell | 1 | 0% | 0% | — | — | 0% | 0 | ○ building sample (1/5) |
| NoTrade (profit-take armed+b+d) | sell | 1 | 0% | — | — | — | 0% | 0 | ○ building sample (1/5) |
| NoTrade (profit-take armed+c) | sell | 1 | 100% | — | — | — | 0% | 0 | ○ building sample (1/5) |
| NoTrade (stop-break basis) | sell | 1 | 100% | 100% | — | — | 0% | 0 | ○ building sample (1/5) |
| PullbackBuy_VWAP | sell | 1 | 0% | 0% | — | — | 0% | 0 | ○ building sample (1/5) |
| ReversalDown_Overbought (profit-take armed+a) | sell | 1 | 0% | — | — | — | 0% | 0 | ○ building sample (1/5) |
| ReversalDown_Overbought (profit-take armed+c) | sell | 1 | 100% | 0% | — | — | 0% | 0 | ○ building sample (1/5) |
Counterfactuals Counterfactuals grade decisions that were NOT executed — expired queues, no-chase rules, untaken optional trims — against what the market then did. They are kept strictly separate from the calibration cells above.
What the decisions we didn't take would have done.
- Optional trims not taken — 331 cases , median would-have 1W move +0.4% → verdict: saved
- No-chase queue (queued a pullback instead of chasing) — 33 cases , median would-have 1W move -0.6% → verdict: saved
Counterfactuals are advisory evidence for the book's learnings reviews — never blended into the calibration cells.