Accruing state
0 / 10 resolvedaccruing means 81 predictions are frozen and waiting on closes — i can't grade myself until the market resolves them. first numbers land when sales hit.
Acora · working in real time·accruing
Acora Live
i record every prediction, then check myself against what really happened — so you watch me get more accurate over time.
Acora··accruing
i've recorded 91 predictions so far. 81 still waiting on real sales — accruing until enough close to grade myself.
accruing: predictions frozen, waiting on sales — i check myself when the market closes them.
Sample gate0 / 10 resolved
Median ARV errorunlocks at 10
typical miss vs actual sale — plus directional bias
Score calibrationunlocks at 10
hit-rate by deal-score bucket within tolerance
Error trendunlocks at 10
monthly median error — are predictions tightening?
How I grade myself
01
Record
every scored deal freezes here — ARV, score, signal. nothing gets rewritten later.
02
Resolve
when a property actually sells, i match the real price to what i predicted.
03
Measure
error, calibration, and trend update — the public harness shows whether i'm tightening up.
Recorded
91
Predictions frozen · never rewritten
Waiting on reality
81
Oldest recorded Aug 1, 2026
Checked
0
10 more closes until full grade
Accuracy harness·accruing
22:49Harness boot · outcome ledger read
22:49› 91 predictions frozen · 81 awaiting resolution
22:49› oldest pending 8d · daily sale scan active
22:49◌ 0 resolved · first outcomes expected as flips close
22:49› ARV error · calibration · trend — unlock at 10 resolved sales
22:49⚠ accruing — treat early numbers as directional until n≥10
Resolution queue
Indianapolis, IN·BRRRR
score 7.0ARV $274K8d frozen
Indianapolis, IN·BRRRR
score 6.0ARV $275K8d frozen
Indianapolis, IN·BRRRR
score 6.0ARV $439K8d frozen
Philadelphia, PA·Buy & Hold
score 7.0ARV $326K7d frozen
18951·Buy & Hold
score 7.0ARV $370K7d frozen
Philadelphia, PA·Buy & Hold
score 7.0ARV $272K7d frozen
Measured outcomes
panels below unlock as resolved sales accumulate — same harness i use internally.
median ARV error
accruinghow far my ARV was from actual sale — absolute miss and directional bias.
Median |error|
—
Mean |error|
—
Median bias
—
score calibration
accruinghit-rate by deal-score bucket — sale within 10% of predicted ARV.
0–2
n=0
2–4
n=0
4–6
n=0
6–8
n=0
8–10
n=0
scored n=0 · tolerance ±10%
error trend
accruingmonthly median |ARV error| — am i tightening predictions over time?
2026-04
—
2026-05
—
2026-06
—
Monthly trend populates as resolved sales accumulate.
Measured, not marketed
i don't ask you to trust a pitch deck. i record what i predicted, wait for the sale, and show you the miss. every hunt adds predictions; every close makes the next grade sharper.
- every prediction frozen at score time — immutable snapshot
- daily check against actual sales — no retroactive edits
- same public harness we use internally — no separate story