Fortress toolchain - predicted vs realized

Calibration Report

Scored 2026-10-03 cc_scanner 10,047 picks logged - 8,577 matured fortress_fight 2996 graded - span 92d fortress_rebuild 448 resolved wall 0m 29s (28.5s)

Predicted vs realized, across every calibration ledger

12,021 graded predictions across cc_scanner, fortress_fight, and fortress_rebuild, scored against actual daily highs and closes. Bars and dots compare what the models claimed to what the market did.

cc_scanner - survival
86% 90%
predicted vs realized, 8,577 matured picks across all sigma bands.
fortress_fight - touch
37% 33%
raw factor 0.90 over 2996 expired contracts.
fortress_fight - breach
441 breaches
raw breach factor 0.82 (predicted rate vs realized) over 2996 graded contracts.
fortress_rebuild - OTM at expiry
82% 84%
375 of 448 resolved. Touch: 36% predicted, 38% realized.

fortress_fight: touch calibration curve

Each dot is a predicted-touch bucket of expired contracts (dot size = count, hover for exact rates). On the dashed line, prediction equals reality. Dots below the line = the engine over-warned; above = under-warned. Also on the books: 25 in progress (censored), 1 unexpired early touches (preview only).

perfectly calibrated 0-10% bucket - n=422 - predicted 4.0% - realized 2.6% 10-20% bucket - n=626 - predicted 16.4% - realized 13.4% 20-30% bucket - n=379 - predicted 24.2% - realized 24.3% 30-40% bucket - n=347 - predicted 34.6% - realized 31.1% 40-50% bucket - n=399 - predicted 44.3% - realized 39.9% 50-60% bucket - n=192 - predicted 55.0% - realized 45.8% 60-70% bucket - n=115 - predicted 64.9% - realized 55.6% 70-80% bucket - n=218 - predicted 75.8% - realized 67.9% 80-90% bucket - n=190 - predicted 84.5% - realized 79.5% 90-100% bucket - n=108 - predicted 96.5% - realized 89.8% 0 25 50 75 100 0 25 50 75 100 Predicted touch % Realized touch %

Correction-factor arming gates

Raw factors are written every grading pass; the engine only applies them when all four gates pass.

✓
Graded contracts ≥ 30
2996 / 30
✓
Breach events ≥ 10
441 / 10
✓
Expiry weeks ≥ 6 (one cycle teaches one regime)
13 / 6
✓
History span ≥ 60d (must see vol vary)
92d / 60d
✕
Challenger beats raw out of sample (champion check)
raw
MetricNPredictedRealizedRaw factor
Touch299637.3%33.4%0.90
Breach2996realized 441 breach event(s)0.82
IV haircutcorrect the input: touch model on iv × h0.92
Champion/challenger (rolling-origin walk-forward, 12 folds, 2683 pooled out-of-sample rows): flat Brier 0.1742 / ECE 0.0686; haircut Brier 0.1728 / ECE 0.0723; logit Brier 0.1715 / ECE 0.0661; raw Brier 0.1670 / ECE 0.0334; trend Brier 0.1722 / ECE 0.0805 (experimental, loses to raw) → winner raw. Admission needs Brier no worse than raw AND ECE better; ties go to raw; a correction only arms while it wins. Experimental shapes are scored but never armed.

fortress_fight: week by week

Each row is one expiry week of fully graded contracts. Ratio = realized ÷ predicted touch: below 1.0 the model over-warned that week, above 1.0 it under-warned. A correction that would have looked right on the first rows is judged by the later ones.

1.0 = calibrated2026-W28: n=313, predicted 34.3%, realized 24.0%, ratio 0.70W282026-W29: n=574, predicted 39.6%, realized 20.2%, ratio 0.51W292026-W30: n=530, predicted 41.8%, realized 38.7%, ratio 0.93W302026-W31: n=257, predicted 34.8%, realized 36.6%, ratio 1.05W312026-W32: n=368, predicted 43.4%, realized 50.3%, ratio 1.16W322026-W33: n=212, predicted 36.2%, realized 40.6%, ratio 1.12W332026-W34: n=120, predicted 27.3%, realized 43.3%, ratio 1.59W342026-W35: n=114, predicted 29.6%, realized 24.6%, ratio 0.83W352026-W36: n=102, predicted 33.8%, realized 38.2%, ratio 1.13W362026-W37: n=83, predicted 32.5%, realized 34.9%, ratio 1.07W372026-W38: n=155, predicted 34.7%, realized 32.9%, ratio 0.95W382026-W39: n=86, predicted 34.3%, realized 46.5%, ratio 1.36W392026-W40: n=82, predicted 30.7%, realized 2.4%, ratio 0.08W40
WeekNPred touchReal touchRatioPred breachReal breach
2026-W2831334.3%24.0%0.7016.5%6.7%
2026-W2957439.6%20.2%0.5119.1%1.6%
2026-W3053041.8%38.7%0.9320.2%10.0%
2026-W3125734.8%36.6%1.0516.8%18.7%
2026-W3236843.4%50.3%1.1620.3%38.9%
2026-W3321236.2%40.6%1.1217.2%19.3%
2026-W3412027.3%43.3%1.5913.2%31.7%
2026-W3511429.6%24.6%0.8314.3%7.9%
2026-W3610233.8%38.2%1.1316.3%24.5%
2026-W378332.5%34.9%1.0715.8%14.5%
2026-W3815534.7%32.9%0.9517.0%16.1%
2026-W398634.3%46.5%1.3616.7%19.8%
2026-W408230.7%2.4%0.0814.9%0.0%

Touch by prior-20-day trend (ex-ante, experimental)

Rows bucket each graded contract by its ticker’s return over the 20 trading days BEFORE the forecast, the one trend fact knowable at pick time. If realized splits by bucket while predicted does not, the model is trend-blind. Walk-forward verdict on the trend challenger: loses to raw out of sample.

Prior 20d trendNPredicted touchRealized touchGap
down >10%140939%36%-4 pts
flat +/-10%136436%33%-3 pts
up >10%21332%21%-10 pts

Scoring-run history

One row per weekly grading pass (cumulative sample at that date). Watch the raw factors converge and the winner column stay on raw.

RunFight NTouch pred / realRaw touch factorRaw breach factorWinnerScanner NScanner surv pred / realRebuild NRebuild touch pred / real
2026-09-03248938% / 34%0.890.80raw663986% / 88%30035% / 35%
2026-09-05259038% / 34%0.890.82raw694086% / 89%31535% / 35%
2026-09-12267338% / 34%0.900.82raw710386% / 89%33334% / 35%
2026-09-19282838% / 33%0.890.81raw762487% / 90%37636% / 38%
2026-09-26291438% / 34%0.910.84raw800386% / 90%40636% / 39%
2026-10-03299637% / 33%0.900.82raw857786% / 90%44836% / 38%

cc_scanner: sigma bands, then ticker by ticker

8,577 matured picks graded against daily highs (10,047 logged since 20260703). Survival = finishing OTM. Red touch bars: realized ran hotter than predicted.

Predicted Realized

Survival % by sigma band

0.0-0.5σ - n=682
64% → 81%
0.5-1.0σ - n=2630
78% → 86%
1.0-1.5σ - n=2317
88% → 92%
>1.5σ - n=2948
97% → 95%

Touch % by sigma band

0.0-0.5σ - n=682
56% → 68%
0.5-1.0σ - n=2630
36% → 32%
1.0-1.5σ - n=2317
23% → 18%
>1.5σ - n=2948
8% → 8%

Predicted vs realized touch by ticker (min 3 matured picks)

Sorted by how hot reality ran vs the model. Red connectors: realized above predicted (model too relaxed). Gray connectors: realized below predicted (model too scared).

Ticker
0% — predicted • realized — 100%
Gap (real - pred)
ETHA
+74 ▲ n=34
NEM
+44 ▲ n=224
AAPL
+39 ▲ n=18
BMNR
+25 ▲ n=279
MSTR
+22 ▲ n=100
IBIT
+20 ▲ n=12
META
+18 ▲ n=294
AMD
+14 ▲ n=75
DELL
+13 ▲ n=317
NVDA
+13 ▲ n=227
SNDK
+12 ▲ n=81
AMZN
+10 ▲ n=266
IGV
+8 ▲ n=273
MDB
+8 ▲ n=252
SPY
+8 ▲ n=259
RIOT
+5 ▲ n=192
NOW
+2 ▲ n=291
GOOG
0 n=875
COPX
-2 n=256
SOFI
-3 n=152
GLD
-3 n=213
RKLB
-4 n=27
HIMS
-5 n=317
INTC
-10 n=307
COIN
-11 n=538
CRWV
-12 n=85
MU
-14 n=689
QCOM
-16 n=204
ENPH
-16 n=151
SPCX
-17 n=198
APP
-18 n=194
GLXY
-19 n=157
IREN
-22 n=677
MARA
-23 n=150
CLSK
-23 n=163
CRCL
-40 n=15
HOOD
-40 n=15

fortress_rebuild: CC-survival forecasts

Rebuild-local ledger (same calibration machinery as fight): 527 predictions since 2026-07-03, 448 resolved. Buckets group picks by the survival odds the model claimed.

Predicted survival Realized OTM rate
0-60% - n=26
51% → 73%
60-70% - n=33
67% → 73%
70-80% - n=88
76% → 76%
80-90% - n=184
85% → 84%
90-100% - n=117
94% → 95%
MetricValue
Predictions logged527
Resolved (expiry passed)448
Predicted touch (avg)36.3%
Realized touch37.6%
Touch grades from exact daily highs434 / 447

Diversification complements

Out-of-sample search for tickers that zig when the $1M model book zags — low correlation, positive behaviour on the book's worst days, and enough volatility to sell covered calls against. 70 nights logged (2026-07-24 → 2026-10-03).

No confident add yet — 71 days of history, needs ≥60 nights at 70% persistence and 60% out-of-sample delivery. Candidates are ranked below; a complement moves opposite the model book (corr ≤ 0.30) AND still pays sellable CC premium (vol ≥ 25%).
TickerGroupNightsPersistMed corrCrisis dayOOS deliverVerdict
HCCCoal (ENERGY)4098%+0.23-1.51%35%WATCH
VKTXGLP-1 (HEALTH)4098%+0.20-0.65%5%WATCH
BILLFintech (TECH)3998%-0.25+1.98%85%WATCH
CAVARestaurants (CONS)3998%+0.09-0.61%10%WATCH
EXPETravel (CONS)3998%-0.21+2.15%95%WATCH
LDOSDef Tech (INDL)3998%-0.04+0.01%100%WATCH
LLYGLP-1 (HEALTH)3998%-0.32+0.79%100%WATCH
NFLXInternet (TECH)3998%-0.12+1.26%90%WATCH
NVOGLP-1 (HEALTH)3998%-0.21+0.97%80%WATCH
PARRRefiners (ENERGY)3998%+0.02+1.66%90%WATCH
SNOWSoftware (TECH)3995%+0.19+0.67%5%WATCH
TOSTFintech (TECH)3998%-0.14+1.20%74%WATCH
UBERInternet (TECH)3998%-0.05+0.45%100%WATCH
VGMidstream (ENERGY)3998%-0.19+1.26%100%WATCH
XOPOil E&P (ENERGY)3998%-0.21+0.76%100%WATCH
ABNBInternet (TECH)3897%+0.04+1.10%95%WATCH
DKNGCasinos (CONS)3897%-0.07+0.16%22%WATCH
FLUTCasinos (CONS)3897%-0.19+0.31%79%WATCH
ZSCyber (TECH)3897%+0.03+0.83%21%WATCH
OIHOil Svcs (ENERGY)3688%+0.23-0.78%72%WATCH
AMGNGLP-1 (HEALTH)3588%-0.14+0.86%100%WATCH
IHIMedTech (HEALTH)3588%-0.33+1.24%100%WATCH
ZBRARobotics (INDL)3380%+0.25+0.51%38%WATCH
AMRCoal (ENERGY)3280%+0.24-2.31%56%WATCH
HONRobotics (INDL)3176%+0.19-0.98%8%WATCH
FTNTCyber (TECH)2972%+0.15-0.17%50%WATCH
NRGPower (ENERGY)2492%+0.24-2.57%0%WATCH
MDBSoftware (TECH)2496%+0.18+0.21%25%WATCH
SQMRare Earth (MATL)2396%+0.24-1.67%0%WATCH
TXSteel (MATL)2396%+0.23-0.51%0%WATCH
WFRDOil Svcs (ENERGY)2396%+0.22-1.44%0%WATCH
XYZFintech (TECH)2296%+0.16+0.58%0%WATCH
XHBBuilders (CONS)2181%+0.23-0.41%0%WATCH
ROKRobotics (INDL)1771%+0.28-1.23%33%WATCH
AFRMFintech (TECH)1275%+0.28-1.05%youngWATCH
CIBRCyber (TECH)1292%+0.27-0.46%youngWATCH
BTUCoal (ENERGY)37%+0.22-2.69%0%noise
PLTRDef Tech (INDL)2768%+0.22-0.40%9%noise
PANWCyber (TECH)1639%+0.23-0.73%0%noise
CRWDCyber (TECH)1639%+0.23-0.66%0%noise
CNRCoal (ENERGY)2359%+0.23-1.95%21%noise
JETSAirlines (CONS)1053%+0.24+0.04%0%noise
AVAVDef Tech (INDL)615%+0.26-2.29%33%noise
VIKTravel (CONS)1562%+0.26-0.56%0%noise
“Crisis day” = mean candidate return on the model composite's worst days (average correlation lies in a crash; this is the all-weather test). “OOS deliver” = share of matured claims whose low correlation held in the 30 days AFTER the flag. 36 WATCH.
Generated by calibration_report.py (weekly cron, Sat 08:30 SGT). Sources: runs/cc_scanner/PORTFOLIO/ledger_calibration.json, runs/fight/calib_factors.json, runs/rebuild/calibration_stats.json. Graders grade only fully expired windows; duplicate nightly re-forecasts of the same contract are deduped to one bet. This page mirrors the graders' published numbers and derives nothing of its own.