TYO Research Log
Log Riset
Sistem yang selesai hanyalah bagian terkecil dari riset. Log ini mencatat sisanya: hipotesis, perubahan terkontrol, hasil, dan keputusan — termasuk yang ditolak. Bila metadata historis tidak pernah dicatat, log mengatakannya alih-alih mengarangnya.
- 26Eksperimen
- 16Diterima
- 1Ditolak
- 3Tak meyakinkan
- 6Diarsipkan
- 14Sistem tercakup
Setiap entri mengikuti satu bentuk: Observasi → Hipotesis → Perubahan terkontrol → Backtest → Validasi → Keputusan. "Ubah parameter sampai terlihat bagus" bukan eksperimen dan tidak dicatat di sini.
Perbaikan backtest tidak berarti perbaikan masa depan. Makin banyak eksperimen, makin besar peluang sebagian berhasil kebetulan — log ini ada sebagian agar risiko multiple-testing tetap terlihat.
Protokol pengembangan
Strategi baru di bawah protokol evidence-by-design: data tak terlihat disegel SEBELUM pengembangan, anggaran pencarian dibatasi dan dipublikasikan, dan holdout dibuka tepat satu kali — setelah kandidat dibekukan.
RES-XAU-001
EVIDENCE BY DESIGN
TYO XAU PYRAMID
Kualitas prospektif: Prospektif proses
Prospektif proses: pengembang telah mengalami pasar ini dan MT5 dapat membaca seluruh riwayat, jadi isolasinya prosedural (manifes tersegel, perkakas terjaga, log audit) — bukan ketidaktahuan kriptografis. Dinyatakan apa adanya.
Lapisan data
| Pengembangan | 2016-02-01 → 2023-12-31 | |
|---|---|---|
| Validasi internal | 2024-01-01 → 2024-12-31 | |
| OOS prospektif | 2025-01-01 → 2025-12-31 | |
| HOLDOUT AKHIR | 2026-01-01 → 2026-06-30 | 🔒Tersegel |
Anggaran pencarian
Varian logika 0/12
Eksperimen 15/30
Putaran optimasi 0/10
Kandidat beku 0
Makin banyak dicoba, makin mungkin ada konfigurasi yang kebetulan tampak bagus — maka jumlah pencarian itu sendiri dipublikasikan sebagai bukti.
Tonggak
Baseline: selesaiValidasi internal: tertundaMonte Carlo: tertunda
Status
Kategori
Sumber
Tidak ada eksperimen yang cocok.
-
EXP-000005 ✓Diterima
Risk/return map: simple annual average vs CAGR
- Hipotesis
- CAGR compares unequal test lengths more honestly than total-return-divided-by-years.
- Keputusan
- The simple average produced figures no reader interprets correctly and compressed twelve systems into an unreadable band behind two outliers. CAGR puts 13 of 14 inside 1–28%.
-
EXP-000006 ✓Diterima
Correlation minimum-overlap rule
- Hipotesis
- Pairwise correlations computed over few shared months are noise and should not be displayed.
- Keputusan
- The 13 suppressed pairs involve short-window systems whose r estimates would swing wildly with 1–2 months of data. Surviving range: −0.56 to +0.49, median 0.07.
-
EXP-000201 ✓Diterima
COSMOS — 70/30 split stability of the monthly profile
- Hipotesis
- The monthly return profile of the first 70% of the backtest (positive-month share, mean monthly return) persists in the final 30%.
- Keputusan
- Late window kept a positive mean month (0.22%) and its positive-month share (100%) stayed within 10pp of the early window (100%).
Bulan positif 0pp Rerata bulanan +0.06pp Bulan terburuk +0.09ppBaca entri -
EXP-000202 ✓Diterima
EVERGREEN — 70/30 split stability of the monthly profile
- Hipotesis
- The monthly return profile of the first 70% of the backtest (positive-month share, mean monthly return) persists in the final 30%.
- Keputusan
- Late window kept a positive mean month (0.48%) and its positive-month share (84.93%) stayed within 10pp of the early window (85.71%).
Bulan positif -0.78pp Rerata bulanan -0.37pp Bulan terburuk +17.94ppBaca entri -
EXP-000203 ✓Diterima
GALOA — 70/30 split stability of the monthly profile
- Hipotesis
- The monthly return profile of the first 70% of the backtest (positive-month share, mean monthly return) persists in the final 30%.
- Keputusan
- Late window kept a positive mean month (0.89%) and its positive-month share (96.88%) stayed within 10pp of the early window (96.6%).
Bulan positif +0.28pp Rerata bulanan -0.36pp Bulan terburuk +10.35ppBaca entri -
EXP-000204 ✓Diterima
GODSPEED — 70/30 split stability of the monthly profile
- Hipotesis
- The monthly return profile of the first 70% of the backtest (positive-month share, mean monthly return) persists in the final 30%.
- Keputusan
- Late window kept a positive mean month (0.31%) and its positive-month share (72.15%) stayed within 10pp of the early window (82.07%).
Bulan positif -9.92pp Rerata bulanan -0.72pp Bulan terburuk +11.16ppBaca entri -
EXP-000205 ≈Tak meyakinkan
GRANDSLAM — 70/30 split stability of the monthly profile
- Hipotesis
- The monthly return profile of the first 70% of the backtest (positive-month share, mean monthly return) persists in the final 30%.
- Keputusan
- Late window shorter than 12 months.
Bulan positif 0pp Rerata bulanan -1.01pp Bulan terburuk -0.39ppBaca entri -
EXP-000206 ≈Tak meyakinkan
JAYRO — 70/30 split stability of the monthly profile
- Hipotesis
- The monthly return profile of the first 70% of the backtest (positive-month share, mean monthly return) persists in the final 30%.
- Keputusan
- Late window shorter than 12 months.
Bulan positif 0pp Rerata bulanan -4.2pp Bulan terburuk -0.53ppBaca entri -
EXP-000207 ✓Diterima
JOKER — 70/30 split stability of the monthly profile
- Hipotesis
- The monthly return profile of the first 70% of the backtest (positive-month share, mean monthly return) persists in the final 30%.
- Keputusan
- Late window kept a positive mean month (0.26%) and its positive-month share (86.67%) stayed within 10pp of the early window (85.51%).
Bulan positif +1.16pp Rerata bulanan +0.19pp Bulan terburuk +0.5ppBaca entri -
EXP-000208 ✓Diterima
KOKOMO — 70/30 split stability of the monthly profile
- Hipotesis
- The monthly return profile of the first 70% of the backtest (positive-month share, mean monthly return) persists in the final 30%.
- Keputusan
- Late window kept a positive mean month (0.49%) and its positive-month share (98.28%) stayed within 10pp of the early window (89.63%).
Bulan positif +8.65pp Rerata bulanan -1.09pp Bulan terburuk +15.2ppBaca entri -
EXP-000209 ✓Diterima
METRO — 70/30 split stability of the monthly profile
- Hipotesis
- The monthly return profile of the first 70% of the backtest (positive-month share, mean monthly return) persists in the final 30%.
- Keputusan
- Late window kept a positive mean month (0.35%) and its positive-month share (95.95%) stayed within 10pp of the early window (84.3%).
Bulan positif +11.65pp Rerata bulanan -0.9pp Bulan terburuk +21.54ppBaca entri -
EXP-000210 ✓Diterima
MIRUMO — 70/30 split stability of the monthly profile
- Hipotesis
- The monthly return profile of the first 70% of the backtest (positive-month share, mean monthly return) persists in the final 30%.
- Keputusan
- Late window kept a positive mean month (0.93%) and its positive-month share (90%) stayed within 10pp of the early window (88.32%).
Bulan positif +1.68pp Rerata bulanan -0.3pp Bulan terburuk +15.35ppBaca entri -
EXP-000211 ✓Diterima
NEXUS — 70/30 split stability of the monthly profile
- Hipotesis
- The monthly return profile of the first 70% of the backtest (positive-month share, mean monthly return) persists in the final 30%.
- Keputusan
- Late window kept a positive mean month (1.76%) and its positive-month share (93.33%) stayed within 10pp of the early window (86.96%).
Bulan positif +6.37pp Rerata bulanan +1.04pp Bulan terburuk +0.93ppBaca entri -
EXP-000212 ≈Tak meyakinkan
RINA — 70/30 split stability of the monthly profile
- Hipotesis
- The monthly return profile of the first 70% of the backtest (positive-month share, mean monthly return) persists in the final 30%.
- Keputusan
- Positive-month share fell by 14.54pp — inside the rejection bound but outside the acceptance bound.
Bulan positif -14.54pp Rerata bulanan -4.21pp Bulan terburuk +1.53ppBaca entri -
EXP-000213 ✓Diterima
SENA — 70/30 split stability of the monthly profile
- Hipotesis
- The monthly return profile of the first 70% of the backtest (positive-month share, mean monthly return) persists in the final 30%.
- Keputusan
- Late window kept a positive mean month (0.73%) and its positive-month share (87.18%) stayed within 10pp of the early window (93.18%).
Bulan positif -6pp Rerata bulanan -1.32pp Bulan terburuk -4.55ppBaca entri -
EXP-000214 ✓Diterima
SUPREMACY — 70/30 split stability of the monthly profile
- Hipotesis
- The monthly return profile of the first 70% of the backtest (positive-month share, mean monthly return) persists in the final 30%.
- Keputusan
- Late window kept a positive mean month (0.93%) and its positive-month share (100%) stayed within 10pp of the early window (100%).
Bulan positif 0pp Rerata bulanan -0.2pp Bulan terburuk +0.31ppBaca entri -
EXP-000001 ✓Diterima
Equity-basis vs balance-basis drawdown across the catalogue
- Hipotesis
- Balance-basis drawdown (closed trades only) materially understates peak-to-trough risk for systems that hold losing positions, and must not be ranked against equity-basis figures silently.
- Keputusan
- The gap on the worst case was a factor of ~4 (12.44% vs 48.82%). Publishing the balance figure as "the" drawdown would have materially understated risk.
-
EXP-000002 ✓Diterima
Stop-loss share denominator — data-integrity check
- Hipotesis
- An observed stop-loss share above 100% indicates the denominator excludes some closed trades.
- Keputusan
- Share can no longer exceed 100%; extracted totals now match each report header exactly.
-
EXP-000003 ✕Ditolak
Positive-month share as a consistency score input
- Hipotesis
- Share of profitable months is a usable consistency measure for these systems as-is.
- Keputusan
- Grid- and martingale-style exits win almost every month by construction and give it back in one. Rewarding the raw share ranks the riskiest builds highest — the opposite of what a consistency score is for. Replaced by share × worst-month factor.
-
EXP-000004 ✓Diterima
TYO SCORE: worst-month factor and drawdown cap
- Hipotesis
- Scaling consistency by the worst month, and capping the total by drawdown, restores discrimination without hand-tuning per system.
- Keputusan
- RINA (2nd-highest profit factor, 56% DD) lands at 55 and JAYRO at 40 — the ordering now reflects what the score claims to measure. Caps are disclosed on every page they bind.
-
EXP-000101 ▤Diarsipkan
GALOA — SUNS build vs current build, full 2003–2026 window
- Hipotesis
- Tidak tercatat
- Keputusan
- Tidak tercatat
Profit factor +5.27 Drawdown maks -6.09pp Transaksi +86 Win rate -2.12ppBaca entri -
EXP-000102 ▤Diarsipkan
GALOA — recent-window run (2024–2026)
- Hipotesis
- Tidak tercatat
- Keputusan
- Tidak tercatat
-
EXP-000103 ▤Diarsipkan
GRANDSLAM — 2022-start vs 2024-start dataset window
- Hipotesis
- Tidak tercatat
- Keputusan
- Tidak tercatat
-
EXP-000104 ▤Diarsipkan
JOKER — MT4 build vs MT5 rebuild
- Hipotesis
- Tidak tercatat
- Keputusan
- Tidak tercatat
-
EXP-000105 ▤Diarsipkan
NEXUS — two builds on the identical dataset
- Hipotesis
- Tidak tercatat
- Keputusan
- Tidak tercatat
Profit factor +0.02 Drawdown maks +2.96pp Transaksi 0 Win rate 0ppBaca entri -
EXP-000106 ▤Diarsipkan
GODSPEED — cross-market run on GOLD
- Hipotesis
- Tidak tercatat
- Keputusan
- Tidak tercatat
Sistem yang sudah kami rilis.
Setiap Expert Advisor didokumentasikan lengkap dengan konsep, logika, model risiko, dan kondisi pengujiannya. Tanpa kotak hitam, tanpa hasil yang tak bisa dijelaskan.