{"approach": "Retains the base's tick probes, sub-tick belief, marginal sizing and cheap crossing stock. In the low-price spread-1 case, one-cent selling remains the default. Retail share demand is estimated only after a main one-cent ask survives; a persistent two-cent queue is enabled when YES stock is below one third of projected remaining demand and more than 100 steps remain. Surviving one-cent bids and competitor moves govern mode switches. The queue is capped at 6000 shares, preserves earned FIFO priority without replenishing behind the competitor, suppresses extra YES purchases, and permits limited earlier cash collateral in the final 1000 steps. A 100-share stock-backed two-cent overflow ask is also used below a 0.3-cent belief. Broad replacement of the base's one-cent ask lost substantially on tuning and was rejected.", "blocked": null, "python_version": "3.10.12", "base_mean_edge": 102.3319790889927, "score": {"mean_edge": 107.47676082579723, "retail_edge": 135.99677430634952, "arb_edge": -28.520013480566334, "simulations": 600, "seed_start": 1000, "failures": 0}, "holdout": {"mean_edge": 96.03782167039321, "simulations": 600, "seed_start": 30000, "failures": 0}, "code": "\"\"\"Cheap crossing stock with selective two-cent queues.\n\nOne-cent asks remain the default in low-price spread-one markets. Retail share\nflow is estimated only while a main one-cent ask survives, which excludes ask\narbitrage in the prior step. Scarce stock activates a two-cent queue; surviving\none-cent bids, competitor moves, and the final 100 steps control mode changes.\nExisting two-cent FIFO priority is retained without replenishing behind a\ncompetitor. A small, covered two-cent backup is allowed at very low beliefs.\nWhile selectively joining, extra YES buying is disabled and limited cash can\ncollateralize uncovered asks earlier in the run. All other regimes keep the\nbase's probes, grid belief, marginal sizing and cheap crossing-stock policy.\n\"\"\"\nimport math\nfrom orderbook_pm_challenge.strategy import BaseStrategy\nfrom orderbook_pm_challenge.types import CancelOrder, PlaceOrder, Side\n\nP = dict(lam_r=0.25, m=5.0, sig_n=1.2, lam_j=0.0019, sig_j=0.35, sig=0.02,\n         qmax=3000.0, rmin=2, kA=0.7, kL=0.15, pL=0.05,\n         ammo=1.0, ydem=0.4, W=200.0, fy=0.6, ammo_q=60.0, ammo_pmax=0.05, ammo_maxpay=0.02,\n         fb=0.9, fu=0.7, xu=0.1, sticky=1.0, min_ahead=2.0, keep_hi=1.5, keep_lo=0.7, Hu=300.0, bins=8.0, filt=1.0, uni=0.01, jumpinfo=1.0, dmin=0.0, kLb=0.08, kLf=0.08, pinf=1.0, pre=0.0, H1=100.0,\n         ydf=0.6, Wf=2000.0, g_hi=0.15, g_lo=0.15, qf=400.0, pmaxf=0.05, pre_gap=1.0,\n         ydf1=0.6, g1=5.0, f1=0.6, a1=100.0, nb1=0.0, gs=1.0, p1t=1.0, m1=1.0, zd=1.0, pfl=0.0005)\nSQ2 = math.sqrt(2.0)\nSQ2PI = math.sqrt(2.0 * math.pi)\ndef cdf(x): return 0.5 * (1.0 + math.erf(x / SQ2))\ndef pdf(x): return math.exp(-0.5 * x * x) / SQ2PI\ndef inv(p):\n    p = min(max(p, 1e-9), 1 - 1e-9)\n    t = math.sqrt(-2.0 * math.log(min(p, 1 - p)))\n    x = t - (2.515517 + 0.802853 * t + 0.010328 * t * t) / (1 + 1.432788 * t + 0.189269 * t * t + 0.001308 * t ** 3)\n    x = -x if p < 0.5 else x\n    for _ in range(2):\n        d = pdf(x)\n        if d < 1e-12:\n            break\n        x -= (cdf(x) - p) / d\n    return x\ndef lpm(k):\n    return pdf(k) + k * cdf(k)\ndef fl(q):\n    return math.floor(q * 100 + 1e-9) / 100\n\nNODES = [(-2.2, .06), (-1.3, .17), (-0.6, .27), (0.6, .27), (1.3, .17), (2.2, .06)]\nPROBE = 0.01\n\n\nclass Strategy(BaseStrategy):\n    def __init__(self):\n        self.n = 0\n        self.ahead = set()     # main orders placed strictly ahead of any competitor order at their tick\n        self.res = {}          # order id -> uncovered quantity at placement\n        self.bc = None         # belief: sub-tick grid centres (ticks)\n        self.bw = None         # belief: weights\n        self.st = 0.5          # last one-step sd in ticks\n        self.pcb = self.pca = None\n        self.pb = set()\n        self.pa = set()\n        self.tb = 0\n        self.join_mode = False\n        self.retail_qsum = 0.0\n        self.retail_n = 0\n        self.retail_qrate = 60.0\n        self.s = None          # competitor spread parameter\n\n    def _oid(self, tag):\n        self.n += 1\n        return tag + str(self.n)\n\n    def on_step(self, state):\n        H = max(1, state.steps_remaining)\n        cb = state.competitor_best_bid_ticks or 0\n        ca = state.competitor_best_ask_ticks or 100\n        if self.s is None:\n            self.s = max(1, (ca - cb + 1) // 2)\n        lo, hi = cb, ca\n        for o in state.own_orders:\n            if o.side is Side.BUY:\n                lo = max(lo, o.price_ticks)\n            else:\n                hi = min(hi, o.price_ticks)\n        if P['pinf'] > 0:\n            sb = {o.price_ticks for o in state.own_orders if o.side is Side.BUY and o.remaining_quantity <= PROBE + 1e-9}\n            sa = {o.price_ticks for o in state.own_orders if o.side is Side.SELL and o.remaining_quantity <= PROBE + 1e-9}\n            ea = self.pa - sa\n            eb = self.pb - sb\n            if sa and ea:\n                lo = max(lo, min(max(ea), hi))\n            if eb and (sb or min(eb) <= P['p1t']):\n                hi = min(hi, max(min(eb), lo))\n        if P['jumpinfo'] > 0 and self.pcb is not None:\n            if ca > self.pca + 1:\n                lo = max(lo, ca - 1)\n            if cb < self.pcb - 1:\n                hi = min(hi, cb + 1)\n        self.pcb, self.pca = cb, ca\n        if lo > hi:\n            lo, hi = hi, lo\n        # sub-tick belief over p_(t-1): predict with one-step noise, truncate to [lo, hi]\n        wid = max(hi - lo, 1e-6)\n        nb = max(2, min(48, int(round(wid * P['bins']))))\n        cs = [lo + (i + 0.5) * wid / nb for i in range(nb)]\n        if self.bc is None or P['filt'] <= 0:\n            ws = [1.0 / nb] * nb\n        else:\n            sg = max(self.st, 0.6 * wid / nb)\n            k = -0.5 / (sg * sg)\n            ws = []\n            uni = P['uni'] * sg\n            for c in cs:\n                t = uni\n                for ci, wi in zip(self.bc, self.bw):\n                    d = c - ci\n                    d2 = d * d * k\n                    if d2 > -12.5:\n                        t += wi * math.exp(d2)\n                ws.append(t)\n            tot = sum(ws)\n            ws = [w / tot for w in ws]\n        self.bc, self.bw = cs, ws\n        mean_t = sum(c * w for c, w in zip(cs, ws))\n        var_t = max(sum(c * c * w for c, w in zip(cs, ws)) - mean_t * mean_t, 0.0)\n        phat = min(max(mean_t / 100.0, 0.002), 0.998)\n        var = P['sig'] ** 2 + P['lam_j'] * P['sig_j'] ** 2\n        sd = math.sqrt(H * var)\n        self.tb = self.tb + 1 if hi <= 1 else 0\n        if self.tb > 0 and P['zd'] > 0:\n            # below the last tick there is no probe: the score keeps walking away from it\n            phat = min(phat, max(P['pfl'], cdf(inv(0.01) - P['zd'] * 0.02 * math.sqrt(self.tb) / sd)))\n        z = inv(phat)\n        Z = z * sd\n        sp = pdf(z) / sd * P['sig']\n        self.st = sp * 100.0\n        s = math.sqrt(sp * sp + var_t / 1e4 + 1e-8)\n        sd1 = math.sqrt(max(1.0, H - 1.0) * var)\n        jp = [(cdf((Z + n * P['sig_j']) / sd1), w) for n, w in NODES]\n        lj = P['lam_j']\n\n        def AB(x):\n            k = (x - phat) / s\n            a = s * lpm(k)\n            b = a - (x - phat)\n            ja = sum(w * max(0.0, x - p) for p, w in jp)\n            jb = sum(w * max(0.0, p - x) for p, w in jp)\n            return (1 - lj) * a + lj * ja, (1 - lj) * b + lj * jb\n\n        if H == 1:\n            def AB(x):\n                return (1 - phat) * x, phat * (1 - x)\n        rho = P['lam_r'] * 0.5\n        mu = math.log(P['m']) - 0.5 * P['sig_n'] ** 2\n        kA = P['kL'] if phat < P['pL'] else P['kA']\n\n        def qsize(loss, gain, px, limit=None):\n            loss *= kA\n            if gain <= 0:\n                return 0.0\n            tail = loss / (rho * gain)\n            if tail >= 1.0:\n                return 0.0\n            return min(P['qmax'] if limit is None else limit, math.exp(mu + P['sig_n'] * inv(1 - max(tail, 1e-6))) / px)\n\n        ptk = phat * 100\n        def qbid(t):\n            if not 1 <= t <= 99:\n                return 0.0\n            a, b = AB(t / 100.0)\n            if phat * 100 - t < P['dmin']:\n                return 0.0\n            if P['nb1'] > 0 and self.s == 1 and t == 1 and ptk < 1.0:\n                return 0.0\n            if phat < P['pL']:\n                ytg = P['ydem'] * min(H, P['W']) / max(ca - 1, ptk + 1) * 100.0\n                a *= (P['kLb'] if state.yes_inventory < ytg else P['kLf']) / P['kL']\n            return qsize(a, b, max(phat, 0.05))\n\n        def qask(t, limit=None):\n            if not 1 <= t <= 99:\n                return 0.0\n            a, b = AB(t / 100.0)\n            if t - phat * 100 < P['dmin']:\n                return 0.0\n            return qsize(b, a, t / 100.0, limit)\n\n        bids = [o for o in state.own_orders if o.side is Side.BUY and o.remaining_quantity > PROBE + 1e-9]\n        asks = [o for o in state.own_orders if o.side is Side.SELL and o.remaining_quantity > PROBE + 1e-9]\n        # A surviving main 1c ask rules out ask-side arbitrage in the prior step.\n        if self.s == 1 and any(o.price_ticks == 1 for o in asks):\n            self.retail_qsum += state.sell_filled_quantity\n            self.retail_n += 1\n            self.retail_qrate = (3000.0 + self.retail_qsum) / (50.0 + self.retail_n)\n        if self.s == 1 and ca > 2:\n            for o in asks:\n                if o.price_ticks == 2:\n                    self.ahead.add(o.order_id)\n        sticky = P['sticky'] > 0\n        def top(orders, comp, step):\n            if sticky and sum(o.remaining_quantity for o in orders if o.price_ticks == comp and o.order_id in self.ahead) >= P['min_ahead']:\n                return comp\n            return comp + step\n        bid_t = top(bids, cb, 1)\n        ask_t = top(asks, ca, -1)\n        qb = qbid(bid_t)\n        qa = qask(ask_t)\n        Y = state.yes_inventory\n        low_survives = any(o.side is Side.SELL and o.price_ticks == 1 for o in state.own_orders)\n        high_survives = any(o.side is Side.BUY and o.price_ticks == 1 for o in state.own_orders)\n        if high_survives or ca > 2 or H <= 100:\n            self.join_mode = False\n        elif self.s == 1 and low_survives and phat < 0.008 and Y < self.retail_qrate * H / 3.0:\n            self.join_mode = True\n        join_low = self.join_mode and self.s == 1 and ask_t == 1 and H > 100 and Y < self.retail_qrate * H / 3.0\n        front_low = self.s == 1 and ask_t == 2 and ca >= 2 and H > 100 and Y < self.retail_qrate * H / 3.0 and phat < 0.01 and not high_survives and any(o.price_ticks == 2 and o.order_id in self.ahead for o in asks)\n        join_low = join_low or front_low\n        if join_low:\n            ask_t = 2\n            qa = qask(2, 6000.0)\n        dual = self.s == 1 and ask_t == 1 and H > 100 and phat < 0.003\n        q2 = min(qa, qask(2), 100.0) if dual else 0.0\n        if dual:\n            qa = min(qa, 3000.0)\n\n        Y = state.yes_inventory\n        if ask_t == 1 and H > P['H1'] and P['pre'] > 0:\n            qa = qa if Y >= P['a1'] > 0 else 0.0          # let the competitor's 2-tick ask erode instead of fronting it for nothing\n        # ammunition: hold YES so that low asks are covered\n        ptick = phat * 100\n        amt = 0\n        qam = 0.0\n        if P['ammo'] > 0 and not join_low and ask_t >= 2 and phat < max(P['ammo_pmax'], P['pmaxf']):\n            wealth = state.cash + Y * phat\n            x = max(ask_t, ptick + 1) / 100.0\n            capw = P['fy'] * wealth / max(phat, 0.005)\n            target = min(P['ydem'] * min(H, P['W']) / x, capw)\n            full = min(P['ydf'] * min(H, P['Wf']) / x, capw) if phat < P['pmaxf'] else 0.0\n            t = max(1, min(99, int(math.ceil(ptick + 1e-9))))\n            g = t - ptick\n            if Y < full and g <= P['g_hi'] and bid_t <= t < ask_t:\n                amt, qam = t, min(P['qf'], full - Y)     # a tick p has just crossed: nearly free\n            elif Y < target and phat < P['ammo_pmax'] and bid_t <= t < ask_t and t / 100.0 - phat < P['ammo_maxpay']:\n                amt, qam = t, min(P['ammo_q'], target - Y)\n            elif Y < full and t >= 2 and 1.0 - g <= P['g_lo']:\n                amt, qam = t - 1, min(P['qf'], full - Y)  # rest just under p: filled only on the crossing, or by retail\n\n        if P['pre'] > 0 and ca == 2 and phat < 0.01 and H > P['H1'] and Y < P['pre'] and not qam and 1.0 - ptick < P['pre_gap']:\n            amt, qam = 1, min(P['ammo_q'], P['pre'] - Y)\n        q1 = 0.0\n        if not join_low and P['ydf1'] > 0 and P['m1'] < ptick < P['g1'] + 1.0 and amt != 1:\n            q1 = max(0.0, P['ydf1'] * min(H, P['Wf']) / (self.s / 100.0) - Y)\n        # order management\n        actions = []\n        free = state.free_cash - 2.0\n        keep_ask = 0.0\n        new_b = []\n        new_a = []\n        def cancel(o):\n            nonlocal free\n            actions.append(CancelOrder(o.order_id))\n            x = o.price_ticks / 100.0\n            if o.side is Side.BUY:\n                free += x * o.remaining_quantity\n            else:\n                free += (1 - x) * min(self.res.get(o.order_id, 0.0), o.remaining_quantity)\n            self.ahead.discard(o.order_id)\n            self.res.pop(o.order_id, None)\n        for o in state.own_orders:\n            if o.remaining_quantity <= PROBE + 1e-9:\n                cancel(o)\n        kept_ids = set()\n        def manage(orders, t, q, comp):\n            kept = 0.0\n            for o in orders:\n                if o.price_ticks == t and q > 0 and kept + o.remaining_quantity <= max(q * P['keep_hi'], 1.0):\n                    kept += o.remaining_quantity\n                    kept_ids.add(o.order_id)\n                else:\n                    cancel(o)\n            return kept\n        kb = manage(bids, bid_t, qb, cb)\n        second_asks = [o for o in asks if dual and o.price_ticks == 2]\n        first_asks = [o for o in asks if o not in second_asks]\n        ka = manage(first_asks, ask_t, qa, ca)\n        k2 = manage(second_asks, 2, q2, ca)\n        add_b = (qb - kb) if (kb < qb * P['keep_lo'] and bid_t != cb) else 0.0\n        add_a = (qa - ka) if (ka < qa * P['keep_lo'] and (ask_t != ca or (join_low and not front_low))) else 0.0\n\n        budget = P['fb'] * max(free, 0.0)\n        cost = add_b * bid_t / 100.0 + qam * amt / 100.0\n        if cost > budget:\n            f = budget / cost\n            add_b *= f\n            qam *= f\n            cost = budget\n        out = []\n        q1 = min(q1, P['f1'] * max(budget - cost, 0.0) * 100.0)\n        if q1 >= 0.01:\n            out.append(PlaceOrder(side=Side.BUY, price_ticks=1, quantity=fl(q1)))\n            free -= fl(q1) / 100.0 + 1e-6\n        if amt and qam >= 0.01:\n            q = fl(qam)\n            out.append(PlaceOrder(side=Side.BUY, price_ticks=amt, quantity=q))\n            free -= q * amt / 100.0 + 1e-6\n        if bid_t >= 1 and add_b >= 0.01:\n            q = fl(add_b)\n            oid = self._oid('b')\n            out.append(PlaceOrder(side=Side.BUY, price_ticks=bid_t, quantity=q, client_order_id=oid))\n            free -= q * bid_t / 100.0 + 1e-6\n            if bid_t > cb:\n                self.ahead.add(oid)\n        if ask_t <= 99 and add_a >= 0.01:\n            x = ask_t / 100.0\n            covk = sum(o.remaining_quantity - min(self.res.get(o.order_id, 0.0), o.remaining_quantity)\n                       for o in asks if o.order_id in kept_ids)\n            cov = max(0.0, Y - covk - 0.1)\n            unc = max(0.0, add_a - cov)\n            if unc > 0:\n                ub = max(0.0, free - (1 - P['fu']) * max(state.cash, 0.0)) if x >= P['xu'] else 0.0\n                if H < P['Hu'] and phat < 0.05:\n                    ub = max(0.0, free * (1 - H / P['Hu']))\n                if join_low and phat < 0.01 and H < 1000:\n                    ub = max(ub, max(0.0, free * 0.5 * (1 - H / 1000.0)))\n                unc = min(unc, ub / (1 - x))\n            q = fl(min(add_a, cov + unc))\n            if q >= 0.01:\n                oid = self._oid('a')\n                out.append(PlaceOrder(side=Side.SELL, price_ticks=ask_t, quantity=q, client_order_id=oid))\n                r = max(0.0, q - cov)\n                self.res[oid] = r\n                free -= r * (1 - x) + 1e-6\n                if ask_t < ca:\n                    self.ahead.add(oid)\n\n        if dual and k2 < q2 * P['keep_lo']:\n            covk2 = sum(o.remaining_quantity - min(self.res.get(o.order_id, 0.0), o.remaining_quantity)\n                        for o in asks if o.order_id in kept_ids)\n            # Primary ask is posted first; its covered part has priority.\n            newly_covered = sum(o.quantity for o in out if o.side is Side.SELL and o.quantity > PROBE + 1e-9)\n            cov2 = max(0.0, Y - covk2 - newly_covered - 0.1)\n            q = fl(min(max(0.0, q2 - k2), cov2))\n            if q >= 0.01:\n                oid = self._oid('a')\n                out.append(PlaceOrder(side=Side.SELL, price_ticks=2, quantity=q, client_order_id=oid))\n                self.res[oid] = 0.0\n\n        # probes: tiny orders that the arbitrageur removes when p crosses them\n        f = int(math.floor(ptick + 1e-9))\n        r = max(int(P['rmin']), int(math.ceil(2.0 * s * 100)))\n        self.pb = set()\n        self.pa = set()\n        if free > 1.0:\n            for t in range(f, f + r + 1):\n                if cb < t < ca and 1 <= t <= 99:\n                    out.append(PlaceOrder(side=Side.BUY, price_ticks=t, quantity=PROBE))\n                    self.pb.add(t)\n            for t in range(f + 1 - r, f + 2):\n                if cb < t < ca and 1 <= t <= 99:\n                    out.append(PlaceOrder(side=Side.SELL, price_ticks=t, quantity=PROBE))\n                    self.pa.add(t)\n        return actions + out\n", "code_sha256": "3d11f8011ce248338b7474891e8da9b952235089d169927796cdf4d81f42774c", "tried": ["Public simulator commit b244530ee2a6247b5f930830eeca9dd950da0e7e; Python 3.10.12. Base reproduced on seeds 0-599: mean_edge=102.3319790889927, retail_edge=127.31995100289954, arb_edge=-24.98797191391217, failures=0. No searches used seeds 0-599.", "Tuning used only seeds 1000-1599. Base full tuning: mean_edge=107.2457521863612, retail_edge=135.80165652288818, arb_edge=-28.555904336541. Most searches ran all 144 actual spread-1 seeds in that range, because the graft is restricted to inferred spread 1. Other spreads keep the base policy.", "The following trial means, retail and arb contributions are conditional on those 144 spread-1 runs, not full-600 scores; delta is paired versus base. Endgame trials used a deterministic base-state eligibility screen (47 possible seeds); screened-out runs have identical strategy actions until any activation, so their base rows were reused. The screen reproduced every base result exactly. Final scores come from direct unmodified CLI runs, without screening.", "Trial names: joinH/Lean/Y vary whether to queue at 2c, final switch time and stock threshold; Lean disables additional ammunition purchases while joining. pressureX queues only when stock is below X*remaining_steps/0.01. lateH joins with stock below 1500 until H remaining steps; lateSafe limits size to the original 1c allowance. Empty trials restrict joining to the final 300 steps.", "Trial names: dualQ splits 1c depth capped at Q with a stock-backed 2c ask; Safe caps the second ask at the original 1c sizing allowance. largeQT waits for below-tick dwell T and increases 2c depth to Q. hyst retains joining until a surviving 1c bid or competitor move shows a rise. idleQ quotes at 2c only when the base would place no 1c ask. priorityQ splits total size and remembers that old 2c orders precede later competitor replacements; Extra allows extra size. backupQ keeps full base 1c desired depth and adds stock-backed 2c overflow of at most Q.", "Additional trial names: stockSafe suppresses large tick-1 crossing bids while joining; stockProbe permits them only when an own tick-1 bid survived. queueRisk subtracts assumed competitor queue notional from the marginal-size quantile; Deep requires phat below 0.003. combo combines stock-sensitive joining with stock-backed overflow.", "finance trials permit a time-dependent fraction of free cash to collateralize uncovered 2c asks earlier than the base final-300-step window, only during selective joining; Safe uses the original 1c quantity limit.", "earnedPriority preserves existing 2c priority after the competitor leaves that tick, restricted to low stock; earnedNoRefill also avoids replenishing behind a competitor while spending that already-earned priority. Equal training means were resolved in favor of the smaller/simpler policy.", "recycleAmmo restores ordinary tick-1 YES purchases while joining at 2c, recycling sales cash rather than relying only on covered stock or minted shorts; recycleWindow targets inventory only for the remaining H-100 two-cent quoting window.", "observedDemand estimates retail share demand from prior sell fills only when a substantial 1c ask survives, then gates scarce-stock joining at one third of estimated remaining demand, using a 50-step prior at 60 shares/step.", "Retail-flow observation audit on 144 spread-1 tuning runs: 22182 qualifying steps, 987036.9099999998 observed sell shares, zero arbitrage ask fills misclassified as retail; every instrumented simulation result field matched the unmodified scorer.", "hyst3000: spread1_mean=18.584973765236825, paired_delta=-34.88275733872777, paired_SE=7.546173658111172, retail=49.85892222051669, arb=-31.27394845523054, failures=0.", "joinLean0: spread1_mean=19.748577042322914, paired_delta=-33.71915406164168, paired_SE=7.430027710163298, retail=56.089971507509745, arb=-36.34139446511736, failures=0.", "join0: spread1_mean=20.25755305224583, paired_delta=-33.21017805171877, paired_SE=7.420727247751663, retail=56.089523114301414, arb=-35.831970061986794, failures=0.", "hyst12000: spread1_mean=21.36769823078891, paired_delta=-32.10003287317569, paired_SE=6.817152022588632, retail=52.76629730337848, arb=-31.39859907253818, failures=0.", "joinWait10: spread1_mean=21.835021821488212, paired_delta=-31.632709282476394, paired_SE=7.040749773223527, retail=52.22149635847363, arb=-30.386474536936802, failures=0.", "large3000T10: spread1_mean=21.835021821488212, paired_delta=-31.632709282476394, paired_SE=7.040749773223527, retail=52.22149635847363, arb=-30.386474536936802, failures=0.", "hyst6000: spread1_mean=23.69296087686391, paired_delta=-29.774770227100696, paired_SE=6.034226999426565, retail=55.07775558538126, arb=-31.384794708462486, failures=0.", "leanH25: spread1_mean=24.228328982109037, paired_delta=-29.239402121855548, paired_SE=7.2582550798056085, retail=60.5584285316736, arb=-36.33009954950417, failures=0.", "dualSafe10: spread1_mean=25.028056024227784, paired_delta=-28.4396750797368, paired_SE=6.587637195320397, retail=54.857422735259746, arb=-29.829366710989568, failures=0.", "joinSame: spread1_mean=25.05167383172572, paired_delta=-28.416057272238888, paired_SE=6.657629383466901, retail=54.871249545049324, arb=-29.81957571328263, failures=0.", "large12000T10: spread1_mean=26.03566257750488, paired_delta=-27.43206852645973, paired_SE=5.820382298997572, retail=56.57905552000071, arb=-30.543392942442356, failures=0.", "large24000T10: spread1_mean=26.03566257750488, paired_delta=-27.43206852645973, paired_SE=5.820382298997572, retail=56.57905552000071, arb=-30.543392942442356, failures=0.", "large6000T10: spread1_mean=26.38896153993752, paired_delta=-27.078769564027088, paired_SE=5.710934851782713, retail=56.93016156836112, arb=-30.54120002837152, failures=0.", "leanH50: spread1_mean=26.48129535003265, paired_delta=-26.98643575393194, paired_SE=7.186338994865346, retail=62.817362817214565, arb=-36.33606746711458, failures=0.", "dualSafe50: spread1_mean=27.97118124008544, paired_delta=-25.496549863879167, paired_SE=6.3459182186333765, retail=57.82823124321388, arb=-29.85705000305555, failures=0.", "joinSafe100: spread1_mean=29.947536571441674, paired_delta=-23.520194532522922, paired_SE=5.618949194536845, retail=58.688379222176394, arb=-28.740842650686798, failures=0.", "dual10: spread1_mean=30.73898437084931, paired_delta=-22.728746733115273, paired_SE=6.985503766745084, retail=67.16333822080276, arb=-36.424353849913864, failures=0.", "joinRich: spread1_mean=31.010669062472928, paired_delta=-22.457062041491678, paired_SE=7.051472628072517, retail=67.40773603157501, arb=-36.397066969043735, failures=0.", "joinLean: spread1_mean=31.012441284695154, paired_delta=-22.455289819269453, paired_SE=7.051369887956959, retail=67.40950825379724, arb=-36.397066969043735, failures=0.", "joinWait50: spread1_mean=31.12593150777987, paired_delta=-22.341799596184725, paired_SE=5.49498677932088, retail=62.03767665824234, arb=-30.911745150427766, failures=0.", "large3000T50: spread1_mean=31.12593150777987, paired_delta=-22.341799596184725, paired_SE=5.49498677932088, retail=62.03767665824234, arb=-30.911745150427766, failures=0.", "join100: spread1_mean=31.715747988481958, paired_delta=-21.75198311548265, paired_SE=7.045757363757082, retail=67.54718082187155, arb=-35.83143283333193, failures=0.", "dualSafe100: spread1_mean=33.192346890616676, paired_delta=-20.275384213347923, paired_SE=5.008716386893405, retail=63.10146852988888, arb=-29.909121639226377, failures=0.", "dual50: spread1_mean=33.3321638047757, paired_delta=-20.135567299188885, paired_SE=6.834948533469169, retail=69.79115795653405, arb=-36.458994151699976, failures=0.", "joinSafe500: spread1_mean=33.6787373741521, paired_delta=-19.788993729812507, paired_SE=5.080663661276916, retail=62.746299874161814, arb=-29.067562499960406, failures=0.", "large6000T50: spread1_mean=34.16557863516528, paired_delta=-19.30215246879931, paired_SE=4.353166164192773, retail=65.37756537284166, arb=-31.211986737624986, failures=0.", "large12000T50: spread1_mean=34.196887830380554, paired_delta=-19.270843273584035, paired_SE=4.3435075320374095, retail=65.40407795729236, arb=-31.20719012686179, failures=0.", "large24000T50: spread1_mean=34.196887830380554, paired_delta=-19.270843273584035, paired_SE=4.3435075320374095, retail=65.40407795729236, arb=-31.20719012686179, failures=0.", "dual100: spread1_mean=35.18977359480557, paired_delta=-18.277957509159027, paired_SE=6.6326671013266365, retail=71.69762911806323, arb=-36.50785552320832, failures=0.", "joinSafe1000: spread1_mean=35.265095660925695, paired_delta=-18.202635443038886, paired_SE=4.622401083473103, retail=64.7134368785458, arb=-29.448341217574296, failures=0.", "joinHalf: spread1_mean=35.29215375746736, paired_delta=-18.17557734649722, paired_SE=4.662495446078672, retail=64.48712540994305, arb=-29.194971652431246, failures=0.", "leanH200: spread1_mean=36.48858206107013, paired_delta=-16.97914904289444, paired_SE=6.680542904005166, retail=73.30615200735136, arb=-36.817569946222214, failures=0.", "dualSafe250: spread1_mean=36.80803250303124, paired_delta=-16.65969860093334, paired_SE=4.62116592614649, retail=66.87923263742361, arb=-30.071200134361803, failures=0.", "joinWait100: spread1_mean=37.14468189659028, paired_delta=-16.323049207374307, paired_SE=3.857884790076872, retail=68.72190589959928, arb=-31.577224002980543, failures=0.", "large3000T100: spread1_mean=37.14468189659028, paired_delta=-16.323049207374307, paired_SE=3.857884790076872, retail=68.72190589959928, arb=-31.577224002980543, failures=0.", "dual250: spread1_mean=38.25035626812986, paired_delta=-15.217374835834729, paired_SE=6.422044653371166, retail=75.00324967081734, arb=-36.75289340265901, failures=0.", "priority100: spread1_mean=38.7459194107007, paired_delta=-14.72181169326389, paired_SE=4.236588569242968, retail=68.37897923965137, arb=-29.63305982890347, failures=0.", "large12000T100: spread1_mean=39.26336814160001, paired_delta=-14.204362962364588, paired_SE=3.7957174541823195, retail=70.31477921183955, arb=-31.051411070194433, failures=0.", "large24000T100: spread1_mean=39.26336814160001, paired_delta=-14.204362962364588, paired_SE=3.7957174541823195, retail=70.31477921183955, arb=-31.051411070194433, failures=0.", "large6000T100: spread1_mean=39.4137515926507, paired_delta=-14.053979511313894, paired_SE=3.7501580853079104, retail=70.47000543315345, arb=-31.056253840460403, failures=0.", "join300: spread1_mean=39.86776960476041, paired_delta=-13.599961499204166, paired_SE=6.509616935883677, retail=75.71440026423954, arb=-35.84663065941804, failures=0.", "stockSafe1p0: spread1_mean=40.45473318316529, paired_delta=-13.01299792079931, paired_SE=6.313434392163326, retail=77.71879993702774, arb=-37.2640667538111, failures=0.", "stockProbe1p0: spread1_mean=41.462946793777775, paired_delta=-12.004784310186814, paired_SE=6.2130401341962225, retail=78.6284593336847, arb=-37.16551253985346, failures=0.", "dual500: spread1_mean=41.59874835413195, paired_delta=-11.868982749832638, paired_SE=6.115992230190364, retail=78.87317791440762, arb=-37.27442956024444, failures=0.", "priorityExtra1000: spread1_mean=42.5467027673243, paired_delta=-10.921028336640282, paired_SE=6.07847228921429, retail=80.9509984640111, arb=-38.40429569665762, failures=0.", "leanH400: spread1_mean=42.86817458931458, paired_delta=-10.599556514649999, paired_SE=6.1653131808034685, retail=78.53299149065207, arb=-35.66481690128748, failures=0.", "pressure1p0: spread1_mean=42.897368434417366, paired_delta=-10.57036266954723, paired_SE=6.052653939746769, retail=79.3519908186958, arb=-36.45462238422847, failures=0.", "dualSafe500: spread1_mean=43.44699810345276, paired_delta=-10.020733000511804, paired_SE=3.1840463871821814, retail=73.89269947411597, arb=-30.445701370629852, failures=0.", "joinWait200: spread1_mean=43.799365146073605, paired_delta=-9.668365957890977, paired_SE=2.5657695562748857, retail=75.47196853810622, arb=-31.672603391999296, failures=0.", "dual1000: spread1_mean=44.012422273945134, paired_delta=-9.455308830019447, paired_SE=5.889569819895178, retail=82.39991629233958, arb=-38.38749401836942, failures=0.", "stockSafe0p6: spread1_mean=44.771965551520125, paired_delta=-8.695765552444447, paired_SE=5.785126872902464, retail=82.69987060394787, arb=-37.92790505239928, failures=0.", "priorityExtra2000: spread1_mean=45.332921947107636, paired_delta=-8.134809156856948, paired_SE=5.7987694511931, retail=84.85773197906873, arb=-39.52481003194513, failures=0.", "dual2000: spread1_mean=45.6080575927993, paired_delta=-7.859673511165277, paired_SE=5.7886449108474585, retail=85.27526230723055, arb=-39.66720471440415, failures=0.", "priority500: spread1_mean=46.071369621163875, paired_delta=-7.396361482800696, paired_SE=2.9021204836003918, retail=76.07686119201179, arb=-30.005491570812485, failures=0.", "stockProbe0p6: spread1_mean=46.53215508675901, paired_delta=-6.935576017205556, paired_SE=5.653812733646664, retail=83.59623189836107, arb=-37.0640768115715, failures=0.", "pressure0p6: spread1_mean=47.90454083236526, paired_delta=-5.563190271599306, paired_SE=5.542349600469368, retail=84.63791576904234, arb=-36.73337493664512, failures=0.", "dualSafe1000: spread1_mean=48.02864358555068, paired_delta=-5.4390875184138885, paired_SE=1.6953822297431314, retail=79.88516535255349, arb=-31.856521766979156, failures=0.", "priority1000: spread1_mean=49.523932888135406, paired_delta=-3.9437982158291667, paired_SE=1.9442866823795135, retail=80.57611291463122, arb=-31.052180026470825, failures=0.", "lateEmpty3000: spread1_mean=50.77199715475276, paired_delta=-2.6957339492118053, paired_SE=0.9196141399292683, retail=82.71449204007014, arb=-31.942494885289573, failures=0.", "stockSafe0p4: spread1_mean=50.80957708013888, paired_delta=-2.658154023825695, paired_SE=2.0606185491574798, retail=84.04807880174371, arb=-33.23850172157499, failures=0.", "late0: spread1_mean=51.845750624303456, paired_delta=-1.6219804796611128, paired_SE=0.49802885180050666, retail=83.30493527955831, arb=-31.45918465522985, failures=0.", "lateSafe0: spread1_mean=52.014721240453454, paired_delta=-1.4530098635111128, paired_SE=0.5427672635072723, retail=83.95630681280622, arb=-31.94158557232499, failures=0.", "priority2000: spread1_mean=52.16481513851875, paired_delta=-1.3029159654458335, paired_SE=0.881549708761233, retail=83.64126031157637, arb=-31.47644517303958, failures=0.", "joinY500: spread1_mean=52.45450421117221, paired_delta=-1.0132268927923613, paired_SE=1.0434101787794832, retail=84.74833666596454, arb=-32.29383245476804, failures=0.", "idle500: spread1_mean=52.539768256842365, paired_delta=-0.9279628471222229, paired_SE=0.9109399921534842, retail=84.57882933294857, arb=-32.039061076081246, failures=0.", "stockSafe1500: spread1_mean=52.617467573293744, paired_delta=-0.8502635306708335, paired_SE=1.1182050919013784, retail=84.23137670560483, arb=-31.613909132282622, failures=0.", "stockSafeLate75: spread1_mean=52.63845464163749, paired_delta=-0.8292764623270837, paired_SE=1.1204110888633714, retail=84.24585969347356, arb=-31.60740505180971, failures=0.", "priority2500: spread1_mean=52.75502164177359, paired_delta=-0.7127094621909733, paired_SE=0.3488504056879002, retail=83.82682401717635, arb=-31.07180237537638, failures=0.", "joinY1500: spread1_mean=52.75742250809027, paired_delta=-0.7103085958743058, paired_SE=0.9988439875308401, retail=84.95634782564997, arb=-32.19892531753263, failures=0.", "backup3000: spread1_mean=52.781441082197894, paired_delta=-0.6862900217666672, paired_SE=0.9293417427977307, retail=85.97743546524649, arb=-33.19599438302706, failures=0.", "lateEmpty500: spread1_mean=52.9637456905708, paired_delta=-0.5039854133937504, paired_SE=0.2562988680216482, retail=84.35234327789927, arb=-31.388597587303455, failures=0.", "idle250: spread1_mean=53.006373342369436, paired_delta=-0.461357761595139, paired_SE=0.45560130838356677, retail=84.57458807778747, arb=-31.568214735394438, failures=0.", "backupPriority3000: spread1_mean=53.01423724501248, paired_delta=-0.4534938589520842, paired_SE=0.8742950163679665, retail=85.19997678223608, arb=-32.185739537205535, failures=0.", "stockProbe0p4: spread1_mean=53.05161647901943, paired_delta=-0.4161146249451391, paired_SE=1.2849726586204124, retail=84.30203812583608, arb=-31.25042164678749, failures=0.", "recycleAmmo: spread1_mean=53.07518494029999, paired_delta=-0.3925461636645839, paired_SE=0.8512248192175701, retail=85.91505540133745, arb=-32.83987046100762, failures=0.", "recycleWindow: spread1_mean=53.16397952070485, paired_delta=-0.3037515832597229, paired_SE=0.8651143333537106, retail=85.68697336478955, arb=-32.522993844059705, failures=0.", "lateEmpty100: spread1_mean=53.20840588234789, paired_delta=-0.259325221616667, paired_SE=0.1562652182030696, retail=84.36257533718053, arb=-31.15416945480693, failures=0.", "idle100: spread1_mean=53.28239651675694, paired_delta=-0.18533458720763948, paired_SE=0.1822395853496809, retail=84.56651171164509, arb=-31.284115194865265, failures=0.", "idleH100Q100: spread1_mean=53.28459703371806, paired_delta=-0.1831340702465286, paired_SE=0.18223578987734676, retail=84.56603512211801, arb=-31.281438088377072, failures=0.", "idle50: spread1_mean=53.376091402280544, paired_delta=-0.09163970168402816, paired_SE=0.0911257289053039, retail=84.56364843830691, arb=-31.18755703600137, failures=0.", "idleH100Q50: spread1_mean=53.377193272488185, paired_delta=-0.09053783147638943, paired_SE=0.09112379171122749, retail=84.5631350394687, arb=-31.185941766956237, failures=0.", "late25: spread1_mean=53.40268471404721, paired_delta=-0.06504638991736157, paired_SE=0.3225336219228082, retail=84.85045718990415, arb=-31.447772475831243, failures=0.", "idle25: spread1_mean=53.422000600899295, paired_delta=-0.04573050306527867, paired_SE=0.045566137062366986, retail=84.56265421384441, arb=-31.140653612921522, failures=0.", "idle10: spread1_mean=53.449424717320134, paired_delta=-0.018306386644445085, paired_SE=0.018225935818414144, retail=84.56126099861733, arb=-31.111836281272897, failures=0.", "base: spread1_mean=53.46773110396457, paired_delta=0.0, paired_SE=0.0, retail=84.56016191939995, arb=-31.0924308154111, failures=0.", "priority3000: spread1_mean=53.46773110396457, paired_delta=0.0, paired_SE=0.0, retail=84.56016191939995, arb=-31.0924308154111, failures=0.", "backup1000: spread1_mean=53.485842263707625, paired_delta=0.018111159743054946, paired_SE=0.7436543291113952, retail=85.60431133867775, arb=-32.11846907494443, failures=0.", "backupPriority500: spread1_mean=53.493099365538875, paired_delta=0.02536826157430521, paired_SE=0.3777429184875098, retail=85.09808219120552, arb=-31.60498282564304, failures=0.", "queueRisk24Q500: spread1_mean=53.49701536357012, paired_delta=0.029284259605555165, paired_SE=0.3782715240965342, retail=85.07951307341108, arb=-31.582497709815257, failures=0.", "backup250: spread1_mean=53.4980051148986, paired_delta=0.030274010934028213, paired_SE=0.19002479128892227, retail=84.83284192747703, arb=-31.33483681255554, failures=0.", "queueRisk48Q100: spread1_mean=53.4987997771854, paired_delta=0.031068673220832583, paired_SE=0.07974186129007206, retail=84.67926560637288, arb=-31.180465829165264, failures=0.", "backup100: spread1_mean=53.50885495410901, paired_delta=0.0411238501444438, paired_SE=0.08074716061392678, retail=84.70215745410134, arb=-31.19330249997013, failures=0.", "queueRisk24Q100: spread1_mean=53.50885495410901, paired_delta=0.0411238501444438, paired_SE=0.08074716061392678, retail=84.70215745410134, arb=-31.19330249997013, failures=0.", "backup500: spread1_mean=53.50907499582984, paired_delta=0.04134389186527739, paired_SE=0.37966525926037725, retail=85.09342318659719, arb=-31.58434819074235, failures=0.", "pressure0p1: spread1_mean=53.51217366332569, paired_delta=0.044442559361110905, paired_SE=0.03904718139258883, retail=85.75124384006037, arb=-32.2390701767104, failures=0.", "queueRisk48Q500: spread1_mean=53.53273629494582, paired_delta=0.06500519098124968, paired_SE=0.3056569614758019, retail=84.99394698155899, arb=-31.461210686588174, failures=0.", "backupDeep100: spread1_mean=53.534533138406935, paired_delta=0.0668020344423604, paired_SE=0.033695381596412666, retail=84.63576227717357, arb=-31.101229138742344, failures=0.", "late150: spread1_mean=53.60362286511805, paired_delta=0.1358917611534719, paired_SE=0.13411885461889322, retail=84.61549884938884, arb=-31.01187598424443, failures=0.", "late50: spread1_mean=53.608147576686804, paired_delta=0.14041647272222174, paired_SE=0.26364789015496476, retail=84.86872039108538, arb=-31.2605728143736, failures=0.", "late200: spread1_mean=53.60907846424444, paired_delta=0.14134736027986083, paired_SE=0.13957399607631016, retail=84.55437467217425, arb=-30.94529620790415, failures=0.", "pressure0p3: spread1_mean=53.6446042387368, paired_delta=0.17687313477222202, paired_SE=0.2043882267529639, retail=86.10349943484788, arb=-32.45889519608818, failures=0.", "joinLeanY500: spread1_mean=53.65833576041666, paired_delta=0.19060465645208297, paired_SE=0.15465644824297822, retail=84.89383523999442, arb=-31.235499479553457, failures=0.", "joinY3000: spread1_mean=53.66341808903056, paired_delta=0.195686985065972, paired_SE=0.3495517758251114, retail=85.11136729833467, arb=-31.447949209280537, failures=0.", "joinLeanY3000: spread1_mean=53.66548640182985, paired_delta=0.19775529786527724, paired_SE=0.18805304326676428, retail=86.306763238259, arb=-32.64127683640415, failures=0.", "hyst3000Scarce: spread1_mean=53.67056483229999, paired_delta=0.2028337283354162, paired_SE=0.18324901852937772, retail=86.07159789285897, arb=-32.40103306053818, failures=0.", "pressure0p2: spread1_mean=53.70591718705138, paired_delta=0.23818608308680536, paired_SE=0.18988384249429988, retail=86.0774899251701, arb=-32.3715727380979, failures=0.", "comboHyst: spread1_mean=53.71241871449582, paired_delta=0.24468761053124932, paired_SE=0.20233735918856124, retail=86.17931327527356, arb=-32.46689456075762, failures=0.", "joinLeanY1500: spread1_mean=53.7161137370243, paired_delta=0.2483826330597219, paired_SE=0.21107865691633465, retail=84.82651447392982, arb=-31.110400736878457, failures=0.", "late100: spread1_mean=53.7161137370243, paired_delta=0.2483826330597219, paired_SE=0.21107865691633465, retail=84.82651447392982, arb=-31.110400736878457, failures=0.", "pressure0p05: spread1_mean=53.7161137370243, paired_delta=0.2483826330597219, paired_SE=0.21107865691633465, retail=84.82651447392982, arb=-31.110400736878457, failures=0.", "pressure0p4: spread1_mean=53.722584591519436, paired_delta=0.25485348755486065, paired_SE=0.28376327094004, retail=86.04982852452636, arb=-32.327243932980544, failures=0.", "late75: spread1_mean=53.73829269724652, paired_delta=0.2705615932819442, paired_SE=0.22293586421123265, retail=84.84745891646523, arb=-31.109166219193046, failures=0.", "financeSafe600: spread1_mean=53.74164170580415, paired_delta=0.27391060183958316, paired_SE=0.20843608545094516, retail=86.0871888148694, arb=-32.34554710903817, failures=0.", "backupDeep500: spread1_mean=53.74501953000416, paired_delta=0.2772884260395833, paired_SE=0.13735651843980914, retail=84.89669522220552, arb=-31.15167569217568, failures=0.", "lateSafe50: spread1_mean=53.76749755127778, paired_delta=0.29976644731319424, paired_SE=0.3271943390081323, retail=85.50258120472357, arb=-31.735083653418737, failures=0.", "comboLate75: spread1_mean=53.826353236389565, paired_delta=0.3586221324249992, paired_SE=0.2954693671249304, retail=85.15879637595066, arb=-31.332443139536096, failures=0.", "lateSafe100: spread1_mean=53.82637982184513, paired_delta=0.35864871788055525, paired_SE=0.2527835865924272, retail=85.38807540235274, arb=-31.56169558047915, failures=0.", "comboLate50Safe: spread1_mean=53.84999540324652, paired_delta=0.38226429928194366, paired_SE=0.3780227758582863, retail=85.79807041409856, arb=-31.948075010827072, failures=0.", "hyst12000Scarce: spread1_mean=53.91000662894652, paired_delta=0.4422755249819443, paired_SE=0.3149848963916218, retail=85.81453906509995, arb=-31.904532436131237, failures=0.", "hyst6000Scarce: spread1_mean=53.91000662894652, paired_delta=0.4422755249819443, paired_SE=0.3149848963916218, retail=85.81453906509995, arb=-31.904532436131237, failures=0.", "finance600F0p5: spread1_mean=53.91360066209165, paired_delta=0.4458695581270832, paired_SE=0.3149904735920444, retail=85.81813591735134, arb=-31.904535255237484, failures=0.", "finance600F1p0: spread1_mean=53.92276326242013, paired_delta=0.45503215845555545, paired_SE=0.31829576253348596, retail=85.826764901843, arb=-31.904001639400676, failures=0.", "finance1000F0p25: spread1_mean=53.930228815555544, paired_delta=0.4624977115909721, paired_SE=0.31554953910563877, retail=85.83476407081524, arb=-31.904535255237484, failures=0.", "comboLate100Safe: spread1_mean=53.93749446087985, paired_delta=0.46976335691527726, paired_SE=0.32551385165944696, retail=85.65203756281244, arb=-31.714543101907623, failures=0.", "comboHystDeep: spread1_mean=53.967564538089576, paired_delta=0.4998334341249994, paired_SE=0.3161054754363401, retail=85.87209419646524, arb=-31.904529658353454, failures=0.", "financeSafe1000: spread1_mean=53.96858973934652, paired_delta=0.5008586353819443, paired_SE=0.31189533941865516, retail=86.05387682412356, arb=-32.085287084750675, failures=0.", "comboFinance: spread1_mean=53.9877867246986, paired_delta=0.5200556207340272, paired_SE=0.31664241550973177, retail=85.89231920218053, arb=-31.904532477459707, failures=0.", "finance1000F0p5: spread1_mean=53.98985669663054, paired_delta=0.522125592665972, paired_SE=0.3246268554514899, retail=85.85274028992148, arb=-31.862883593268734, failures=0.", "comboFinanceHalf: spread1_mean=54.0474146057736, paired_delta=0.5796835018090272, paired_SE=0.3256155742050835, retail=85.91029542128678, arb=-31.862880815490954, failures=0.", "earnedPriority: spread1_mean=54.25864350303332, paired_delta=0.7909123990687493, paired_SE=0.5176100542000285, retail=85.14533503457842, arb=-30.886691531524985, failures=0.", "earnedNoRefill: spread1_mean=54.27074485089999, paired_delta=0.8030137469354158, paired_SE=0.5426646586588244, retail=85.38150586437078, arb=-31.110761013447203, failures=0.", "observedDemand: spread1_mean=54.430267101614575, paired_delta=0.9625359976499994, paired_SE=0.6211122831245962, retail=85.37315268382218, arb=-30.94288558218331, failures=0.", "Full tuning paired diagnostics: mean gain=0.23100863943599984, paired_SE=0.1496181830009063, changed simulations=7. These are selected training estimates, not evidence of an independently replicated improvement.", "Frozen candidate observedDemand at 2026-11-05 23:55:30 UTC before any evaluation of seeds 30000-30599. No tuning or code changes followed the holdout. Direct CLI training and holdout each completed 600/600 simulations with zero failures. Final training paired gain=0.2310086394360269.", "Sandbox: unmodified --sandbox on seeds 1000-1019 passed 20/20 for examples/starter_strategy.py and 20/20 for the final frozen strategy on Python 3.10.12. No sandbox patch was used. --workers 4 worked.", "Frozen holdout attribution: final retail=124.39478094730094, arb=-28.35695927689983; conditional spread-1 n=159, final mean=38.3182020907321, base mean=38.18072532361135. Every non-spread-1 simulation result matched the base exactly, in training and holdout.", "Post-freeze comparison only, no selection or retuning: base holdout mean=96.00139032710622, retail=124.33139605111212, arb=-28.33000572399766; final paired holdout gain=0.036431343287000005, paired_SE=0.01989893793084752.", "Final frozen observedDemand: direct full tuning mean=107.47676082579723, retail=135.99677430634952, arb=-28.520013480566334; untouched holdout mean=96.03782167039321. All final scores belong to SHA256 3d11f8011ce248338b7474891e8da9b952235089d169927796cdf4d81f42774c."], "next": "Correct the retail-demand estimator for censoring: it learns only when the 1c ask survives, omitting the largest stock-exhausting orders. Test a censored-flow estimate for the stock/remaining-demand switch while holding total jump exposure fixed. Prioritize independent replication over further threshold sweeps; the selected training gain comes from few paths."}