Classic pivot points on GBPUSD, 15-minute bars (spec piv_spec.md). Run 2026-11-11T12:16:15.016Z; 500 control runs per variant data: 8293892 minutes 2004-01-01 01:05 .. 2026-09-28 15:46 New York; forex days 5905: {"no previous day":1,"short previous day":4,"traded day":5871,"no 08:15 London minute":29}; cost 1.5 pips a round trip previous-day range (H - L), median pips by year: 2004 168 | 2005 133 | 2006 131 | 2007 120 | 2008 184 | 2009 200 | 2010 154 | 2011 132 | 2012 91 | 2013 100 | 2014 80 | 2015 113 | 2016 128 | 2017 92 | 2018 101 | 2019 85 | 2020 116 | 2021 87 | 2022 114 | 2023 93 | 2024 73 | 2025 84 | 2026 75 === PIV-REV === DEV n= 1329 net 814.67 pips avg 0.613 PF 1.03 t 0.38 VAL n= 980 net -1107.67 pips avg -1.130 PF 0.94 t -0.87 HOLDOUT n= 974 net -2495.33 pips avg -2.562 PF 0.85 t -2.15 ALL n=3283; net -2788.33 pips, avg -0.849 (gross 0.651, cost 1.5); PF 0.96 t -1.02; ended by {"stop":899,"target":1202,"close":1182} in R: median R 54.97 pips; gross 0.005 R, net -0.025 R a trade; cost 0.030 R a trade cost scenarios (not gates): 1 pips: avg -0.349, net -1146.83 | 2.5 pips: avg -1.849, net -6071.33 win rate 1761/3283 = 53.6% (95% Wilson 51.9-55.3%); the 10 best trades 1920.03; drawdown on closed trades -4641.43 longs n=1626 net -1840.97 gross 598.03 | shorts n=1657 net -947.37 gross 1538.13 by year (n, net pips): 2004 133 -471 | 2005 159 -975 | 2006 135 129 | 2007 147 -123 | 2008 142 521 | 2009 157 2,269 | 2010 154 -107 | 2011 150 -42 | 2012 152 -387 | 2013 135 -405 | 2014 136 -162 | 2015 146 -544 | 2016 133 128 | 2017 145 306 | 2018 146 -245 | 2019 139 -185 | 2020 158 -283 | 2021 150 -244 | 2022 146 -317 | 2023 155 -727 | 2024 136 -416 | 2025 130 -346 | 2026 99 -160 control (the same trades moved to other days at the same London time, 500 runs, 67 s): mean trades 3,283, median avg -1.542, 5-95% -2.626 .. -0.393; real avg -0.849 at percentile 81.60 [FAIL] 1 PF >= 1.2 in validation: VAL PF 0.94 [PASS] 2 >= 50 trades in each period: 1329 / 980 / 974 [FAIL] 3 t >= 2.5 on development: DEV t 0.38 [FAIL] 4 above 95th pct of the control: percentile 81.60 [FAIL] 5 positive without top-10 trades: -4708.37 pips [FAIL] 6 holdout net positive (already used): HOLDOUT net -2495.33 pips PF 0.85 VERDICT Buried (1/6 gates) rule 3: 26 trades open across a suspect gap; without them avg -0.821, gates passed 1/6 -> Buried (verdict unchanged) === PIV-BRK === DEV n= 1390 net -4606.20 pips avg -3.314 PF 0.86 t -2.17 VAL n= 1034 net 208.37 pips avg 0.202 PF 1.01 t 0.17 HOLDOUT n= 1028 net -1465.00 pips avg -1.425 PF 0.91 t -1.25 ALL n=3452; net -5862.83 pips, avg -1.698 (gross -0.198, cost 1.5); PF 0.91 t -2.16; ended by {"target":1554,"close":1131,"stop":767} in R: median R 60.80 pips; gross 0.009 R, net -0.019 R a trade; cost 0.028 R a trade cost scenarios (not gates): 1 pips: avg -1.198, net -4136.83 | 2.5 pips: avg -2.698, net -9314.83 win rate 1925/3452 = 55.8% (95% Wilson 54.1-57.4%); the 10 best trades 1437.73; drawdown on closed trades -7523.93 longs n=1711 net -1781.67 gross 784.83 | shorts n=1741 net -4081.17 gross -1469.67 by year (n, net pips): 2004 152 313 | 2005 152 406 | 2006 157 133 | 2007 148 -666 | 2008 146 -937 | 2009 159 -2,212 | 2010 155 -1,460 | 2011 167 -44 | 2012 154 -139 | 2013 155 721 | 2014 135 128 | 2015 155 330 | 2016 135 170 | 2017 141 -629 | 2018 160 -166 | 2019 153 -345 | 2020 171 -654 | 2021 161 -224 | 2022 146 -265 | 2023 163 3 | 2024 139 -307 | 2025 141 283 | 2026 107 -300 control (the same trades moved to other days at the same London time, 500 runs, 65 s): mean trades 3,452, median avg -1.696, 5-95% -2.743 .. -0.578; real avg -1.698 at percentile 49.60 [FAIL] 1 PF >= 1.2 in validation: VAL PF 1.01 [PASS] 2 >= 50 trades in each period: 1390 / 1034 / 1028 [FAIL] 3 t >= 2.5 on development: DEV t -2.17 [FAIL] 4 above 95th pct of the control: percentile 49.60 [FAIL] 5 positive without top-10 trades: -7300.57 pips [FAIL] 6 holdout net positive (already used): HOLDOUT net -1465.00 pips PF 0.91 VERDICT Buried (1/6 gates) rule 3: 27 trades open across a suspect gap; without them avg -1.615, gates passed 1/6 -> Buried (verdict unchanged) === PIV-LDN === DEV n= 1943 net -3876.03 pips avg -1.995 PF 0.90 t -1.78 VAL n= 1619 net -3372.05 pips avg -2.083 PF 0.86 t -2.38 HOLDOUT n= 1510 net -1655.07 pips avg -1.096 PF 0.92 t -1.23 ALL n=5072; net -8903.15 pips, avg -1.755 (gross -0.255, cost 1.5); PF 0.89 t -3.04; ended by {"target+break-even":1218,"target+target":589,"stop+stop":2329,"target+close":458,"close+close":478} in R: median R 33.53 pips; gross 0.012 R, net -0.043 R a trade; cost 0.054 R a trade cost scenarios (not gates): 1 pips: avg -1.255, net -6367.15 | 2.5 pips: avg -2.755, net -13975.15 win rate 2375/5072 = 46.8% (95% Wilson 45.5-48.2%); the 10 best trades 2346.28; drawdown on closed trades -9826.00 longs n=2575 net -5131.32 gross -1268.82 | shorts n=2497 net -3771.83 gross -26.33 by year (n, net pips): 2004 231 -1 | 2005 217 404 | 2006 225 -36 | 2007 205 -130 | 2008 207 -1,197 | 2009 193 -246 | 2010 210 -1,971 | 2011 226 -250 | 2012 229 -447 | 2013 237 -432 | 2014 228 -1,296 | 2015 242 82 | 2016 227 28 | 2017 226 -555 | 2018 232 -678 | 2019 227 -521 | 2020 223 748 | 2021 219 -720 | 2022 222 -444 | 2023 234 -130 | 2024 230 -931 | 2025 212 247 | 2026 170 -425 control (the same trades moved to other days at the same London time, 500 runs, 96 s): mean trades 5,072, median avg -1.149, 5-95% -1.938 .. -0.362; real avg -1.755 at percentile 11.00 [FAIL] 1 PF >= 1.2 in validation: VAL PF 0.86 [PASS] 2 >= 50 trades in each period: 1943 / 1619 / 1510 [FAIL] 3 t >= 2.5 on development: DEV t -1.78 [FAIL] 4 above 95th pct of the control: percentile 11.00 [FAIL] 5 positive without top-10 trades: -11249.43 pips [FAIL] 6 holdout net positive (already used): HOLDOUT net -1655.07 pips PF 0.92 VERDICT Buried (1/6 gates) rule 3: 35 trades open across a suspect gap; without them avg -1.918, gates passed 1/6 -> Buried (verdict unchanged)