{"approach":"Starting from the exact seat-845 crossing-cover code (artifact 40497298132dcc0ccc9f4c0672031c7eaae00ddcf8cf24983db989feeb3c5ca0), keep its low-price funding and spread-dependent risk sizing. At estimated probability >=0.20, refill both bid and ask orders inside the competitor whenever retained depth is below 95% of the current target, instead of 70%. Thus a competitive order remaining at 80% of target now receives additional depth. Quote targets still depend on the price gap to the inferred probability and its arbitrage-loss model, so sizing remains tied to spread. At estimated probability >=0.95 only, multiply the ordinary bid target by 1.5. Add the symmetric upper-tail duration estimate when the observed probability lower bound stays at least 99 cents, with depth coefficient 1.5 and probability ceiling 0.9995. Preserve collateral budgeting and FIFO retention at competitor prices. Reorganize order scans, reuse identical probe action values, and accelerate the unchanged eight-point Gaussian convolution. The optimized file matched the selected parent action-for-action on 40000 steps from seeds 4000-4019. Simulator commit b244530ee2a6247b5f930830eeca9dd950da0e7e; Python standard library plus the supplied challenge API; default 2000 steps; no sandbox and no seed-dependent strategy logic. Scorer reproduced 108.63974289615767 exactly on seeds 1000-1599. Development and diagnostic fitting used only seeds 4000-4999; separate validation used 6000-7999. Seeds 5000-5999 were first scored for the selected file after freezing. Selected version frozen at 2026-11-06T07:29:06.636271+00:00 before any 50000-53999 results were accessed. The full 4000-7999 score and the unseen 50000-53999 score use that exact file. Unseen fills were attributed with a read-only market subclass; all 200 instrumented base results were exactly equal to the ordinary simulator results in the equivalence check. Eligibility (no failures in either 4000-run batch, no retail decrease on unseen, and no slower 200-run elapsed time): True. Price-zone audit below uses execution price, not estimated probability; values are mean raw fill edge per simulation over all 4000 unseen runs. The competitor column measures realized retail edge left to the competitor; base plus competitor is the realized whole-book maker edge pool. Price and FIFO constraints prevent capturing that entire pool. The frontier capacity column is an optimistic reference on the baseline path: unlimited depth one tick inside the competitor, or at an existing participant queue position ahead of it, excluding collateral limits and arbitrage losses; it is not achievable net profit. Unseen change at execution prices >=20 cents: {\"retail_buy\":0.3420015210557006,\"retail_sell\":0.14562813455592494,\"arb\":-0.2975848318374279}. Price-zone audit: [{\"fill_price_zone\":\"1-4 cents\",\"base_retail_buy\":4.167817531646527,\"base_retail_sell\":69.05648686319574,\"base_arb\":-17.102616188876492,\"selected_retail_buy\":4.168367903964556,\"selected_retail_sell\":69.06342042131477,\"selected_arb\":-17.095758651419068,\"base_competitor_retail\":11.938124853306675,\"base_frontier_gross_retail_capacity\":79.69219007954432,\"base_realized_total_book_retail_edge\":85.16242924814895},{\"fill_price_zone\":\"5-19 cents\",\"base_retail_buy\":5.923316926472388,\"base_retail_sell\":23.180728155190717,\"base_arb\":-6.347185420903609,\"selected_retail_buy\":5.931716763766476,\"selected_retail_sell\":23.1682265107291,\"selected_arb\":-6.321764669141292,\"base_competitor_retail\":20.25016743390375,\"base_frontier_gross_retail_capacity\":37.33396307135261,\"base_realized_total_book_retail_edge\":49.35421251556686},{\"fill_price_zone\":\"20-49 cents\",\"base_retail_buy\":5.881582974087,\"base_retail_sell\":7.258028579232872,\"base_arb\":-3.3764834240274744,\"selected_retail_buy\":5.94856056809674,\"selected_retail_sell\":7.340897315130777,\"selected_arb\":-3.501216655954097,\"base_competitor_retail\":19.97122742681657,\"base_frontier_gross_retail_capacity\":21.906587214238805,\"base_realized_total_book_retail_edge\":33.11083898013644},{\"fill_price_zone\":\"50-79 cents\",\"base_retail_buy\":3.403860314426901,\"base_retail_sell\":3.286397722850749,\"base_arb\":-1.8304061468725363,\"selected_retail_buy\":3.4505205302363966,\"selected_retail_sell\":3.330993890417672,\"selected_arb\":-1.9084089570242577,\"base_competitor_retail\":10.328276741853927,\"base_frontier_gross_retail_capacity\":11.286701222461293,\"base_realized_total_book_retail_edge\":17.018534779131578},{\"fill_price_zone\":\"80-99 cents\",\"base_retail_buy\":3.612762798205608,\"base_retail_sell\":1.157921015818951,\"base_arb\":-0.7251766746034796,\"selected_retail_buy\":3.841126509442073,\"selected_retail_sell\":1.1760842469100483,\"selected_arb\":-0.8200254643625633,\"base_competitor_retail\":3.6239901305666047,\"base_frontier_gross_retail_capacity\":6.201206078850069,\"base_realized_total_book_retail_edge\":8.394673944591164}]. Timing used one process and eight alternating 25-run blocks on matching seeds while the full scoring batches were also active; the seconds fields are measured elapsed time, and CPU times are reported in tried to disclose the load effect. These are local simulation measurements, not a new Paradigm leaderboard score.","blocked":null,"python_version":"3.13.5","base_mean_edge":108.63974289615767,"base_unseen":{"mean_edge":97.54703502577897,"retail_edge":126.92890288111944,"arb_edge":-29.381867855334075},"score":{"mean_edge":98.24925168071465,"retail_edge":125.6600463017868,"arb_edge":-27.410794621081074,"simulations":4000,"seed_start":4000,"failures":0},"holdout":{"mean_edge":97.77274026204337,"retail_edge":127.41991465999835,"arb_edge":-29.6471743979495,"simulations":4000,"seed_start":50000,"failures":0},"diff":{"mean":0.22570523626440003,"standard_error":0.03437928452040865,"retail":0.4910117788789,"arb":-0.265306542615425},"seconds":{"base":206.40725212194957,"yours":189.05823257332668},"code":"\"\"\"Retail refill and selective high-price depth on the crossing-cover base.\n\nKeep the base's spread-dependent risk sizing and low-price inventory funding.\nAt estimated YES probabilities of at least 20 cents, replenish a competitive\nquote whenever its remaining depth falls below 95 percent of the current target.\nAbove 95 cents, increase the bid target by 50 percent. Track time beyond the\nupper tick to refine the upper-tail probability estimate. Decisions use only\npublic StepState observations. The common-grid convolution and reusable probe\nactions reduce computation without changing the filter or order values.\n\"\"\"\nimport math\nfrom statistics import NormalDist\n_INVERSE_CDF = NormalDist().inv_cdf\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)\nRHO = P['lam_r'] * 0.5\nMU = math.log(P['m']) - 0.5 * P['sig_n'] ** 2\nVAR = P['sig'] ** 2 + P['lam_j'] * P['sig_j'] ** 2\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    return _INVERSE_CDF(1e-9 if p < 1e-9 else (1 - 1e-9 if p > 1 - 1e-9 else p))\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)]\nJUMP_OFFSETS = [(n * P['sig_j'], w) for n, w in NODES]\nPROBE = 0.01\nPROBE_BUYS = tuple(PlaceOrder(side=Side.BUY, price_ticks=t, quantity=PROBE) for t in range(1, 100))\nPROBE_SELLS = tuple(PlaceOrder(side=Side.SELL, price_ticks=t, quantity=PROBE) for t in range(1, 100))\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.ta = 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        sb = set()\n        sa = set()\n        bids = []\n        asks = []\n        probes = []\n        for o in state.own_orders:\n            if o.side is Side.BUY:\n                if o.price_ticks > lo:\n                    lo = o.price_ticks\n                if o.remaining_quantity <= PROBE + 1e-9:\n                    sb.add(o.price_ticks)\n                    probes.append(o)\n                else:\n                    bids.append(o)\n            else:\n                if o.price_ticks < hi:\n                    hi = o.price_ticks\n                if o.remaining_quantity <= PROBE + 1e-9:\n                    sa.add(o.price_ticks)\n                    probes.append(o)\n                else:\n                    asks.append(o)\n        if P['pinf'] > 0:\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            # Same Gaussian convolution on an unchanged, evenly spaced grid.\n            # Expand the small matrix to reuse its eight distinct exponentials.\n            if nb == 8 and wid == 1.0 and self.bc == cs and all(cs[i] - cs[i-1] == 0.125 for i in range(1, 8)):\n                ev = []\n                for i in range(8):\n                    d = cs[i] - cs[0]\n                    d2 = d * d * k\n                    ev.append(math.exp(d2) if d2 > -12.5 else 0.0)\n                e0, e1, e2, e3, e4, e5, e6, e7 = ev\n                w0, w1, w2, w3, w4, w5, w6, w7 = self.bw\n                ws = [\n                    uni + w0 * e0 + w1 * e1 + w2 * e2 + w3 * e3 + w4 * e4 + w5 * e5 + w6 * e6 + w7 * e7,\n                    uni + w0 * e1 + w1 * e0 + w2 * e1 + w3 * e2 + w4 * e3 + w5 * e4 + w6 * e5 + w7 * e6,\n                    uni + w0 * e2 + w1 * e1 + w2 * e0 + w3 * e1 + w4 * e2 + w5 * e3 + w6 * e4 + w7 * e5,\n                    uni + w0 * e3 + w1 * e2 + w2 * e1 + w3 * e0 + w4 * e1 + w5 * e2 + w6 * e3 + w7 * e4,\n                    uni + w0 * e4 + w1 * e3 + w2 * e2 + w3 * e1 + w4 * e0 + w5 * e1 + w6 * e2 + w7 * e3,\n                    uni + w0 * e5 + w1 * e4 + w2 * e3 + w3 * e2 + w4 * e1 + w5 * e0 + w6 * e1 + w7 * e2,\n                    uni + w0 * e6 + w1 * e5 + w2 * e4 + w3 * e3 + w4 * e2 + w5 * e1 + w6 * e0 + w7 * e1,\n                    uni + w0 * e7 + w1 * e6 + w2 * e5 + w3 * e4 + w4 * e3 + w5 * e2 + w6 * e1 + w7 * e0,\n                ]\n            else:\n                prior = tuple(zip(self.bc, self.bw))\n                for c in cs:\n                    t = uni\n                    for ci, wi in prior:\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 = VAR\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        self.ta = self.ta + 1 if lo >= 99 else 0\n        if self.ta > 0:\n            phat = max(phat, min(0.9995, cdf(inv(0.99) + 1.5 * 0.02 * math.sqrt(self.ta) / 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 + offset) / sd1), w) for offset, w in JUMP_OFFSETS]\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 * (x - p) for p, w in jp if x > p)\n            jb = sum(w * (p - x) for p, w in jp if p > x)\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 = RHO\n        mu = MU\n        kA = P['kL'] if phat < P['pL'] else P['kA']\n\n        def qsize(loss, gain, px):\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'], 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):\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)\n\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        if phat >= 0.95:\n            qb *= 1.5\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 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 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\n        # Bridge expected retail demand until another tick crossing is available.\n        if 0.02 < phat < 0.1:\n            ct = max(2, min(99, int(math.floor(ptick))))\n            wealth = state.cash + Y * phat\n            goal = min(1 * qa, 0.3 * max(wealth, 0.0) / phat)\n            if Y < goal and ct < ask_t:\n                amt, qam = ct, min(3000.0, goal - 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 probes:\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        ka = manage(asks, ask_t, qa, ca)\n        add_b = (qb - kb) if (kb < qb * (0.95 if phat >= 0.2 else P['keep_lo']) and bid_t != cb) else 0.0\n        add_a = (qa - ka) if (ka < qa * (0.95 if phat >= 0.2 else P['keep_lo']) and ask_t != ca) 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            q = fl(q1)\n            out.append(PlaceOrder(side=Side.BUY, price_ticks=1, quantity=q))\n            free -= q / 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                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        # 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(PROBE_BUYS[t - 1])\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(PROBE_SELLS[t - 1])\n                    self.pa.add(t)\n        return actions + out\n","code_sha256":"4d34c482983b88e18053535d96c3fc256d0fb8baf8c5ec607e3097098276b297","tried":["Widen mid/high quote prices by 2 ticks; seeds 4000-4199; mean_edge=81.16686769140699 retail_edge=114.838583563231 arb_edge=-33.671715871821 failures=0 paired_mean=-32.292689468119995 SE=1.7334530037415872","Join competitor when inside quote is too risky: join0x4; seeds 4000-4199; mean_edge=87.58501755032799 retail_edge=138.9467234337425 arb_edge=-51.361705883376 failures=0 paired_mean=-25.874539609199 SE=1.6937021155453467","Widen mid/high quote prices by 1 ticks; seeds 4000-4199; mean_edge=93.762121806608 retail_edge=113.715606620672 arb_edge=-19.9534848140585 failures=0 paired_mean=-19.697435352918998 SE=1.5127424373493976","Above 20c, use 0.15 times modeled arbitrage loss in target-size calculation; seeds 4000-4199; mean_edge=100.120559741578 retail_edge=146.2243522318085 arb_edge=-46.103792490169496 failures=0 paired_mean=-13.338997417949 SE=0.9603017331826027","Above 20c, multiply desired bid and ask quantities by 4.0; seeds 4000-4199; mean_edge=100.8724306178095 retail_edge=141.87720873761052 arb_edge=-41.004778119761006 failures=0 paired_mean=-12.587126541717499 SE=1.8288759850613918","Change mid/high ask sizes: ascale3; seeds 4000-4199; mean_edge=103.49819985870099 retail_edge=135.68441282584251 arb_edge=-32.1862129671315 failures=0 paired_mean=-9.961357300826 SE=2.030980504664581","Add a second price level: layerinner10; seeds 4000-4199; mean_edge=106.60987294162 retail_edge=134.792526454777 arb_edge=-28.1826535131285 failures=0 paired_mean=-6.849684217906998 SE=0.5855011120816811","Join competitor when inside quote is too risky: join5x1; seeds 4000-4199; mean_edge=107.265677098939 retail_edge=138.777782751469 arb_edge=-31.512105652509 failures=0 paired_mean=-6.193880060588 SE=0.5587687734744022","Separate diffusion and jump loss weights: risk0p3j0p5; seeds 4000-4199; mean_edge=107.702758626829 retail_edge=143.915257835847 arb_edge=-36.21249920896 failures=0 paired_mean=-5.7567985326979985 SE=0.6553039244354996","Join competitor when inside quote is too risky: join1x1; seeds 4000-4199; mean_edge=107.706238124506 retail_edge=138.92062070324948 arb_edge=-31.214382578722002 failures=0 paired_mean=-5.753319035021 SE=0.5411778890785697","Join competitor when inside quote is too risky: join0x1; seeds 4000-4199; mean_edge=107.8460586967745 retail_edge=138.919218857734 arb_edge=-31.073160160933497 failures=0 paired_mean=-5.613498462752499 SE=0.5388118701114996","Add a second price level: layerinner05; seeds 4000-4199; mean_edge=108.629958523739 retail_edge=135.239825846766 arb_edge=-26.6098673230215 failures=0 paired_mean=-4.829598635788 SE=0.3918607199238031","Change mid/high ask sizes: ascale2; seeds 4000-4199; mean_edge=110.22285816050949 retail_edge=138.74304351566 arb_edge=-28.520185355148502 failures=0 paired_mean=-3.236698999017499 SE=0.8760073809209001","Add a second price level: layerinner025; seeds 4000-4199; mean_edge=110.23322524883349 retail_edge=136.05430451369 arb_edge=-25.821079264828498 failures=0 paired_mean=-3.226331910693499 SE=0.2539568519194399","Separate diffusion and jump loss weights: risk0p3j1p5; seeds 4000-4199; mean_edge=110.2429823487095 retail_edge=141.0000422203955 arb_edge=-30.757059871642 failures=0 paired_mean=-3.216574810817499 SE=0.5476491774322778","Above 20c, use 0.3 times modeled arbitrage loss in target-size calculation; seeds 4000-4199; mean_edge=110.48886886934001 retail_edge=142.7906059910745 arb_edge=-32.301737121726 failures=0 paired_mean=-2.9706882901869993 SE=0.4941941231834099","Separate diffusion and jump loss weights: risk0p3j2p5; seeds 4000-4199; mean_edge=110.9190307719205 retail_edge=138.659206952756 arb_edge=-27.740176180765 failures=0 paired_mean=-2.5405263876064996 SE=0.5150241268803891","Above 20c, use 2.0 times modeled arbitrage loss in target-size calculation; seeds 4000-4199; mean_edge=111.51112517963249 retail_edge=133.524777056889 arb_edge=-22.013651877254503 failures=0 paired_mean=-1.9484319798945 SE=0.3302217718404139","Increase mid/high bid sizes: bscale0p5; seeds 4000-4199; mean_edge=111.527644318963 retail_edge=135.819530333327 arb_edge=-24.291886014364 failures=0 paired_mean=-1.931912840563999 SE=0.4044601446220207","Separate diffusion and jump loss weights: risk0p5j0p5; seeds 4000-4199; mean_edge=111.659106104468 retail_edge=141.908661991483 arb_edge=-30.249555886940996 failures=0 paired_mean=-1.8004510550589992 SE=0.3821021527324994","Above 20c, multiply desired bid and ask quantities by 2.0; seeds 4000-4199; mean_edge=111.74792620944 retail_edge=141.69001102491052 arb_edge=-29.942084815457 failures=0 paired_mean=-1.7116309500870008 SE=0.4847367783984947","Increase depth only ahead of predicted opposite-side replenishment: queuevalue1p6both1; seeds 4000-4199; mean_edge=111.7930633413085 retail_edge=139.6222563147865 arb_edge=-27.829192973436 failures=0 paired_mean=-1.6664938182184983 SE=0.323384710220081","Join competitor when inside quote is too risky: join0x0p25; seeds 4000-4199; mean_edge=112.1423558559445 retail_edge=138.71883055158452 arb_edge=-26.576474695625002 failures=0 paired_mean=-1.3172013035824992 SE=0.38620197504790593","Size up only when a modeled opposite-boundary crossing may give queue priority: crossqueue1p6w2; seeds 4000-4199; mean_edge=112.250373744749 retail_edge=139.34930999617652 arb_edge=-27.0989362513715 failures=0 paired_mean=-1.209183414778 SE=0.300195805111371","Increase mid/high bid sizes: bscale3; seeds 4000-4199; mean_edge=112.255343616562 retail_edge=140.9865569968195 arb_edge=-28.7312133802435 failures=0 paired_mean=-1.204213542965 SE=0.47863307529035903","Separate diffusion and jump loss weights: risk0p5j2p5; seeds 4000-4199; mean_edge=112.27802861727851 retail_edge=137.33546158554452 arb_edge=-25.057432968239 failures=0 paired_mean=-1.1815285422484998 SE=0.28061048042191905","Expand asks only against surplus YES stock: acover1; seeds 4000-4199; mean_edge=112.47183421596799 retail_edge=139.0970785573845 arb_edge=-26.6252443414005 failures=0 paired_mean=-0.9877229435589986 SE=0.3145912300749696","Separate diffusion and jump loss weights: risk0p5j1p5; seeds 4000-4199; mean_edge=112.50847433446751 retail_edge=139.3474672084175 arb_edge=-26.838992873921 failures=0 paired_mean=-0.9510828250595 SE=0.2794824987821617","Separate diffusion and jump loss weights: risk1p2j2p5; seeds 4000-4199; mean_edge=112.579905585867 retail_edge=135.654049976836 arb_edge=-23.0741443909485 failures=0 paired_mean=-0.8796515736599985 SE=0.23070576819831556","Size up only when a modeled opposite-boundary crossing may give queue priority: crossqueue1p6w1; seeds 4000-4199; mean_edge=112.59201134798501 retail_edge=138.9419947239695 arb_edge=-26.3499833759425 failures=0 paired_mean=-0.867545811541999 SE=0.2132934659304358","Above 20c, use 1.4 times modeled arbitrage loss in target-size calculation; seeds 4000-4199; mean_edge=112.65448918803449 retail_edge=135.3775600836 arb_edge=-22.723070895523 failures=0 paired_mean=-0.8050679714924999 SE=0.23583495779760352","Separate diffusion and jump loss weights: risk0p8j2p5; seeds 4000-4199; mean_edge=112.6778628354015 retail_edge=136.33386660362 arb_edge=-23.656003768168997 failures=0 paired_mean=-0.7816943241254996 SE=0.22875721925416997","Change mid/high ask sizes: ascale1p5; seeds 4000-4199; mean_edge=112.754671661314 retail_edge=139.3353603175625 arb_edge=-26.580688656226997 failures=0 paired_mean=-0.7048854982129985 SE=0.21523408202723743","Mid/high target retention and refill thresholds: refresh1p5l0p3; seeds 4000-4199; mean_edge=112.777987523062 retail_edge=137.33116808993452 arb_edge=-24.553180566842997 failures=0 paired_mean=-0.6815696364649986 SE=0.13576140363018335","Change mid/high quote diffusion width: vol0p8; seeds 4000-4199; mean_edge=112.810575183447 retail_edge=140.257910967069 arb_edge=-27.447335783564498 failures=0 paired_mean=-0.6489819760799993 SE=0.16434781636845566","Above 20c, use 0.45 times modeled arbitrage loss in target-size calculation; seeds 4000-4199; mean_edge=112.84939299461799 retail_edge=140.599221689186 arb_edge=-27.7498286945465 failures=0 paired_mean=-0.6101641649089998 SE=0.18783193754667415","Boost both-side size in probability band: zone0p2x1p8; seeds 4000-4199; mean_edge=112.931597542926 retail_edge=139.647270593066 arb_edge=-26.7156730501235 failures=0 paired_mean=-0.5279596166009987 SE=0.32417604452370335","Expand asks only against surplus YES stock: acover0p5; seeds 4000-4199; mean_edge=112.93612932863451 retail_edge=138.98344838783999 arb_edge=-26.047319059189498 failures=0 paired_mean=-0.5234278308924994 SE=0.16793111592045853","Boost both-side size in probability band: zone0p4x1p8; seeds 4000-4199; mean_edge=112.98298086023249 retail_edge=139.1554657087645 arb_edge=-26.172484848527002 failures=0 paired_mean=-0.47657629929449924 SE=0.15225174483108053","Condition probe information on aggregate fill size: probefill0p03; seeds 4000-4199; mean_edge=113.01569584051451 retail_edge=138.17917418559648 arb_edge=-25.1634783450875 failures=0 paired_mean=-0.44386131901249953 SE=0.08600241874615502","Above 20c, multiply desired bid and ask quantities by 1.5; seeds 4000-4199; mean_edge=113.0392278534605 retail_edge=140.499319558315 arb_edge=-27.4600917048415 failures=0 paired_mean=-0.4203293060664995 SE=0.24870717918003105","Change mid/high quote diffusion width: vol1p2; seeds 4000-4199; mean_edge=113.070397629247 retail_edge=136.7045752569895 arb_edge=-23.634177627742503 failures=0 paired_mean=-0.38915953028000017 SE=0.12039463317710813","Above 20c, keep existing quote depth up to 10.0 times current target; seeds 4000-4199; mean_edge=113.0875640108225 retail_edge=139.3784118221455 arb_edge=-26.290847811315498 failures=0 paired_mean=-0.37199314870449884 SE=0.15572534543070016","Extra crossing-funded cover above 20c: cover1off0; seeds 4000-4199; mean_edge=113.13026330814901 retail_edge=138.543969138063 arb_edge=-25.4137058299045 failures=0 paired_mean=-0.3292938513780003 SE=0.22971719791616643","Increase depth only ahead of predicted opposite-side replenishment: queuevalue1p6both0; seeds 4000-4199; mean_edge=113.1529148829165 retail_edge=139.1464376672895 arb_edge=-25.993522784351498 failures=0 paired_mean=-0.30664227661049936 SE=0.1418894925916623","Separate diffusion and jump loss weights: risk0p8j0p5; seeds 4000-4199; mean_edge=113.155261722233 retail_edge=140.2753890725985 arb_edge=-27.120127350364502 failures=0 paired_mean=-0.30429543729400005 SE=0.1909859188263157","Increase depth only ahead of predicted opposite-side replenishment: queuevalue1p3both1; seeds 4000-4199; mean_edge=113.1568043675445 retail_edge=139.5126643774105 arb_edge=-26.355860009816503 failures=0 paired_mean=-0.30275279198249877 SE=0.16579377791513067","Above 20c, use 1.0 times modeled arbitrage loss in target-size calculation; seeds 4000-4199; mean_edge=113.202516414773 retail_edge=136.9246360135975 arb_edge=-23.7221195987885 failures=0 paired_mean=-0.25704074475399935 SE=0.12096745669789208","Separate diffusion and jump loss weights: risk0p8j1p5; seeds 4000-4199; mean_edge=113.203626164174 retail_edge=138.063136082785 arb_edge=-24.8595099185335 failures=0 paired_mean=-0.2559309953529993 SE=0.10833621543191567","Change mid/high filter prediction width: predict0p8; seeds 4000-4199; mean_edge=113.21488380293849 retail_edge=138.4443487938095 arb_edge=-25.2294649908465 failures=0 paired_mean=-0.24467335658849929 SE=0.06780016017576625","Soften information from removed mid/high probes: softprobe1; seeds 4000-4199; mean_edge=113.21851276637149 retail_edge=138.2597065104685 arb_edge=-25.0411937441075 failures=0 paired_mean=-0.2410443931554991 SE=0.05974033038288065","Extra crossing-funded cover above 20c: cover1off1; seeds 4000-4199; mean_edge=113.2443717273545 retail_edge=138.5636739171285 arb_edge=-25.31930218977 failures=0 paired_mean=-0.2151854321725003 SE=0.14918960893664504","Condition probe information on aggregate fill size: probefill0p015; seeds 4000-4199; mean_edge=113.2566784581985 retail_edge=138.365474870093 arb_edge=-25.1087964118975 failures=0 paired_mean=-0.2028787013284993 SE=0.06631651959643456","Change mid/high ask sizes: ascale0p5; seeds 4000-4199; mean_edge=113.26223098052549 retail_edge=136.51208477467802 arb_edge=-23.249853794133 failures=0 paired_mean=-0.197326179001499 SE=0.2513762191173476","Extra crossing-funded cover above 20c: cover0p5off0; seeds 4000-4199; mean_edge=113.262328234976 retail_edge=138.523058510352 arb_edge=-25.260730275365 failures=0 paired_mean=-0.19722892455099802 SE=0.16043048366376153","Separate diffusion and jump loss weights: risk1p2j1p5; seeds 4000-4199; mean_edge=113.26341150051199 retail_edge=137.1932089428295 arb_edge=-23.9297974423 failures=0 paired_mean=-0.19614565901499964 SE=0.114110495662659","Size up only when a modeled opposite-boundary crossing may give queue priority: crossqueue1p3w2; seeds 4000-4199; mean_edge=113.3027956127785 retail_edge=139.375461372758 arb_edge=-26.0726657599465 failures=0 paired_mean=-0.15676154674849838 SE=0.15345816097378123","Increase mid/high bid sizes: bscale2; seeds 4000-4199; mean_edge=113.30327161833048 retail_edge=140.2327893624425 arb_edge=-26.9295177440965 failures=0 paired_mean=-0.1562855411964993 SE=0.23282961264982385","Size up only when a modeled opposite-boundary crossing may give queue priority: crossqueue1p3w1; seeds 4000-4199; mean_edge=113.3220973614435 retail_edge=139.0202042823455 arb_edge=-25.6981069208705 failures=0 paired_mean=-0.13745979808349995 SE=0.1309684507154638","Soften information from removed mid/high probes: softprobe0p5; seeds 4000-4199; mean_edge=113.3254148815865 retail_edge=138.3657316955995 arb_edge=-25.0403168140035 failures=0 paired_mean=-0.1341422779404998 SE=0.04129416952785342","Boost both-side size in probability band: zone0p4x1p3; seeds 4000-4199; mean_edge=113.32562435956001 retail_edge=138.8065249354225 arb_edge=-25.4809005758495 failures=0 paired_mean=-0.13393279996699936 SE=0.06584812511596784","Boost both-side size in probability band: zone0p6x1p8; seeds 4000-4199; mean_edge=113.33352261378249 retail_edge=138.997204942331 arb_edge=-25.663682328529 failures=0 paired_mean=-0.126034545744499 SE=0.08058624568181978","Above 20c, keep existing quote depth up to 4.0 times current target; seeds 4000-4199; mean_edge=113.34244930597 retail_edge=139.1461200903335 arb_edge=-25.803670784349002 failures=0 paired_mean=-0.11710785355699815 SE=0.09721127248296016","Change mid/high filter regularization: uniform0p05; seeds 4000-4199; mean_edge=113.35148891850349 retail_edge=138.360673361105 arb_edge=-25.009184442565502 failures=0 paired_mean=-0.10806824102349914 SE=0.03798869538206027","Boost both-side size in probability band: zone0p2x1p3; seeds 4000-4199; mean_edge=113.35911369895501 retail_edge=139.0511596126385 arb_edge=-25.692045913670498 failures=0 paired_mean=-0.10044346057199903 SE=0.1423279654732468","Increase mid/high bid sizes: bscale1p5; seeds 4000-4999; mean_edge=101.84229720665681 retail_edge=129.2884908783347 arb_edge=-27.4461936716854 failures=0 paired_mean=-0.09614891981719947 SE=0.06717899835757105","Mid/high target retention and refill thresholds: refresh1p2l0p8; seeds 4000-4199; mean_edge=113.365007772765 retail_edge=138.4062851785275 arb_edge=-25.0412774057405 failures=0 paired_mean=-0.09454938676199957 SE=0.05671109417275102","Soften information from removed mid/high probes: softprobe0p25; seeds 4000-4199; mean_edge=113.37011856160599 retail_edge=138.39841917847198 arb_edge=-25.028300616856498 failures=0 paired_mean=-0.08943859792099959 SE=0.031379749215874164","Mid/high target retention and refill thresholds: refresh1p1l0p9; seeds 4000-4199; mean_edge=113.3841964438845 retail_edge=138.424041484923 arb_edge=-25.039845041007002 failures=0 paired_mean=-0.07536071564249934 SE=0.07347009164727955","Integrate sub-tick belief mixture for mid/high quote risk: mix1above0p2; seeds 4000-4199; mean_edge=113.419644647907 retail_edge=138.4757327266725 arb_edge=-25.056088078753 failures=0 paired_mean=-0.039912511619998745 SE=0.03131790426027066","Faster refill with spread-dependent size: spreadcombo1p01p01p41p8; seeds 4000-4199; mean_edge=113.4200277166285 retail_edge=139.828514813574 arb_edge=-26.408487096913 failures=0 paired_mean=-0.039529442898499474 SE=0.20894515255022095","Separate diffusion and jump loss weights: risk1p2j0p5; seeds 4000-4199; mean_edge=113.421613526567 retail_edge=139.1308040895655 arb_edge=-25.709190562955 failures=0 paired_mean=-0.037943632960000126 SE=0.164179484066602","Mid/high target retention and refill thresholds: refresh2l0p5; seeds 4000-4199; mean_edge=113.430879913561 retail_edge=138.476279294055 arb_edge=-25.0453993804775 failures=0 paired_mean=-0.028677245966000773 SE=0.10317982819906844","Above 20c, retain competitor-tick priority once ahead quantity exceeds 0.01; seeds 4000-4199; mean_edge=113.43196705329099 retail_edge=138.4209604087475 arb_edge=-24.988993355452 failures=0 paired_mean=-0.027590106235998904 SE=0.04765132627511101","Above 20c, retain competitor-tick priority once ahead quantity exceeds 0.25; seeds 4000-4199; mean_edge=113.438306393169 retail_edge=138.4280249972185 arb_edge=-24.9897186040435 failures=0 paired_mean=-0.021250766357998753 SE=0.047504461563157394","Faster refill plus mid/high size adjustment: combo_b1p4a1; seeds 4000-4999; mean_edge=101.92683890973841 retail_edge=129.41607096822452 arb_edge=-27.4892320584874 failures=0 paired_mean=-0.011607216735599602 SE=0.07959601178392642","Above 20c, keep existing quote depth up to 2.0 times current target; seeds 4000-4999; mean_edge=101.9380389611591 retail_edge=128.5152474675953 arb_edge=-26.577208506439202 failures=0 paired_mean=-0.00040716531489963433 SE=0.032923754107384566","Speed-only refactor fast0; seeds 4000-4199; mean_edge=113.459557159527 retail_edge=138.4995598434705 arb_edge=-25.040002683935498 failures=0 paired_mean=0.0 SE=0.0","Speed-only refactor fast1; seeds 4000-4199; mean_edge=113.459557159527 retail_edge=138.4995598434705 arb_edge=-25.040002683935498 failures=0 paired_mean=0.0 SE=0.0","Expand bids only when YES cover is below desired asks: bcover1; seeds 4000-4199; mean_edge=113.4645832085405 retail_edge=138.600166661447 arb_edge=-25.1355834528995 failures=0 paired_mean=0.0050260490135015665 SE=0.07061520048205272","Integrate sub-tick belief mixture for mid/high quote risk: mix1above0p8; seeds 4000-4199; mean_edge=113.4662551678975 retail_edge=138.5014161185335 arb_edge=-25.035160950623 failures=0 paired_mean=0.006698008370500056 SE=0.015363574520515195","Integrate sub-tick belief mixture for mid/high quote risk: mix0p5above0p2; seeds 4000-4199; mean_edge=113.46629082352351 retail_edge=138.5006673776575 arb_edge=-25.0343765541075 failures=0 paired_mean=0.006733663996501418 SE=0.014792242949728156","Integrate sub-tick belief mixture for mid/high quote risk: mix0p5above0p8; seeds 4000-4199; mean_edge=113.467915923599 retail_edge=138.50432317795298 arb_edge=-25.036407254333998 failures=0 paired_mean=0.008358764072000062 SE=0.006648088998156246","Symmetric belief beyond 99c: top0p5cap0p9995; seeds 4000-4199; mean_edge=113.47300191112599 retail_edge=138.51291914792 arb_edge=-25.039917236786 failures=0 paired_mean=0.013444751599000106 SE=0.003615198336563407","Expand bids only when YES cover is below desired asks: bcover2; seeds 4000-4199; mean_edge=113.47352530291201 retail_edge=138.6308187707845 arb_edge=-25.1572934678705 failures=0 paired_mean=0.013968143384999713 SE=0.0717001958928515","Above 20c, retain competitor-tick priority once ahead quantity exceeds 1.0; seeds 4000-4199; mean_edge=113.476624792431 retail_edge=138.4797792297745 arb_edge=-25.003154437337 failures=0 paired_mean=0.017067632904000363 SE=0.042218760433133766","Spread-aware size, faster refill, and upper-tail belief: synth_queue; seeds 4000-4999; mean_edge=101.9617859088614 retail_edge=129.180945476252 arb_edge=-27.2191595673853 failures=0 paired_mean=0.023339782387400306 SE=0.0668770534416001","Symmetric belief beyond 99c: top1cap0p9995; seeds 4000-4199; mean_edge=113.485636897547 retail_edge=138.52561970466002 arb_edge=-25.0399828071045 failures=0 paired_mean=0.026079738020000093 SE=0.006963765974846748","Change mid/high filter regularization: uniform0p001; seeds 4000-4199; mean_edge=113.48726959133 retail_edge=138.554598026811 arb_edge=-25.0673284354725 failures=0 paired_mean=0.027712431803001526 SE=0.027585397922957538","Symmetric belief beyond 99c: top1cap0p99995; seeds 4000-4199; mean_edge=113.49008327134399 retail_edge=138.5300655784585 arb_edge=-25.039982307104502 failures=0 paired_mean=0.030526111817000108 SE=0.008181814109324708","Boost both-side size in probability band: zone0p8x1p8; seeds 4000-4199; mean_edge=113.49424567686951 retail_edge=138.9557286849925 arb_edge=-25.4614830080975 failures=0 paired_mean=0.03468851734249991 SE=0.0753923510176751","Symmetric belief beyond 99c: top1p5cap0p9995; seeds 4000-4199; mean_edge=113.495808957517 retail_edge=138.5370625358745 arb_edge=-25.041253578348503 failures=0 paired_mean=0.03625179799000017 SE=0.009745481581237891","Boost both-side size in probability band: zone0p9x1p8; seeds 4000-4199; mean_edge=113.50055990548401 retail_edge=138.7522466332085 arb_edge=-25.2516867276925 failures=0 paired_mean=0.04100274595700009 SE=0.04830285636997189","Symmetric belief beyond 99c: top2cap0p9995; seeds 4000-4199; mean_edge=113.50090062380899 retail_edge=138.5415372195375 arb_edge=-25.0406365957175 failures=0 paired_mean=0.04134346428200011 SE=0.010326217723405416","Symmetric belief beyond 99c: top1p5cap0p99995; seeds 4000-4199; mean_edge=113.501295706613 retail_edge=138.5425492849715 arb_edge=-25.041253578348503 failures=0 paired_mean=0.04173854708600015 SE=0.011307036495771248","Expand bids only when YES cover is below desired asks: bcover0p5; seeds 4000-4199; mean_edge=113.502360191728 retail_edge=138.57558335531598 arb_edge=-25.073223163583503 failures=0 paired_mean=0.04280303220100189 SE=0.026194724006878525","Change mid/high filter prediction width: predict1p2; seeds 4000-4199; mean_edge=113.51009020515299 retail_edge=138.464597439318 arb_edge=-24.954507234119 failures=0 paired_mean=0.05053304562599981 SE=0.040968429881653584","Boost both-side size in probability band: zone0p9x1p3; seeds 4000-4199; mean_edge=113.518972202925 retail_edge=138.6261499157855 arb_edge=-25.1071777128315 failures=0 paired_mean=0.05941504339800014 SE=0.025030161000377023","Boost both-side size in probability band: zone0p6x1p3; seeds 4000-4199; mean_edge=113.52163811136101 retail_edge=138.769689929286 arb_edge=-25.2480518179105 failures=0 paired_mean=0.062080951834001155 SE=0.04948726849089599","Above 20c, keep existing quote depth up to 2.0 times current target; seeds 4000-4199; mean_edge=113.52359473631901 retail_edge=138.77420738193752 arb_edge=-25.250612645609 failures=0 paired_mean=0.06403757679200175 SE=0.06917320091069798","Boost both-side size in probability band: zone0p8x1p3; seeds 4000-4199; mean_edge=113.53453678703751 retail_edge=138.71983334964298 arb_edge=-25.185296562574 failures=0 paired_mean=0.07497962751049997 SE=0.035602061427952844","Boost both-side size in probability band: zone0p95x1p3; seeds 4000-4199; mean_edge=113.5425586697405 retail_edge=138.59634487844951 arb_edge=-25.0537862086835 failures=0 paired_mean=0.08300151021350008 SE=0.021150423683221786","Faster refill plus mid/high size adjustment: combo_b1p6a1p1; seeds 4000-4199; mean_edge=113.547726427864 retail_edge=140.2916509893745 arb_edge=-26.7439245614695 failures=0 paired_mean=0.08816926833700123 SE=0.20162483388804225","Refill every positive mid/high shortfall: fullrefillh1p5; seeds 4000-4199; mean_edge=113.55574694348799 retail_edge=138.868095931844 arb_edge=-25.3123489883185 failures=0 paired_mean=0.09618978396100093 SE=0.08674468315710973","Mid/high target retention and refill thresholds: refresh1p5l0p95; seeds 4000-4999; mean_edge=102.04519208571729 retail_edge=128.5973567051985 arb_edge=-26.5521646194757 failures=0 paired_mean=0.10674595924330038 SE=0.029849376365289778","Boost only high-probability bids: topbid0p8x1p5; seeds 4000-4199; mean_edge=113.57254285457749 retail_edge=138.7622583516485 arb_edge=-25.1897154970575 failures=0 paired_mean=0.11298569505050013 SE=0.04890718773427378","Boost only high-probability bids: topbid0p9x2; seeds 4000-4199; mean_edge=113.578704038587 retail_edge=138.75385870827148 arb_edge=-25.175154669659 failures=0 paired_mean=0.11914687906000028 SE=0.05643232334651901","Spread-aware size, faster refill, and upper-tail belief: synth_balanced; seeds 4000-4999; mean_edge=102.0608330865 retail_edge=129.2801794384146 arb_edge=-27.2193463519086 failures=0 paired_mean=0.12238696002600073 SE=0.066297792750218","Spread-aware size, faster refill, and upper-tail belief: synth_mild; seeds 4000-4999; mean_edge=102.0691276592029 retail_edge=129.0219466404799 arb_edge=-26.952818981283702 failures=0 paired_mean=0.13068153272890048 SE=0.043661399532025386","Faster refill plus mid/high size adjustment: combo_b1p2a1p1; seeds 4000-4199; mean_edge=113.59311961480348 retail_edge=139.55874681406948 arb_edge=-25.9656271992265 failures=0 paired_mean=0.13356245527650035 SE=0.1273457033526174","Boost both-side size in probability band: zone0p95x1p8; seeds 4000-4199; mean_edge=113.5956848477575 retail_edge=138.69742842331098 arb_edge=-25.10174357552 failures=0 paired_mean=0.13612768823050003 SE=0.03435475547738225","Refill every positive mid/high shortfall: fullrefillh2; seeds 4000-4199; mean_edge=113.60145903552849 retail_edge=139.1590973350155 arb_edge=-25.557638299455004 failures=0 paired_mean=0.14190187600150098 SE=0.10371081317155541","Increase depth only ahead of predicted opposite-side replenishment: queuevalue1p3both0; seeds 4000-4199; mean_edge=113.6033355075705 retail_edge=139.0606059517885 arb_edge=-25.4572704441835 failures=0 paired_mean=0.14377834804350148 SE=0.09443866144594676","Mid/high target retention and refill thresholds: refresh1p5l0p95; seeds 4000-4199; mean_edge=113.612595069389 retail_edge=138.864747168443 arb_edge=-25.2521520990195 failures=0 paired_mean=0.1530379098620007 SE=0.07844866253666234","Faster refill plus mid/high size adjustment: combo_b1p2a1; seeds 4000-4999; mean_edge=102.0928435755259 retail_edge=129.0748696707763 arb_edge=-26.982026095254803 failures=0 paired_mean=0.15439744905190053 SE=0.05868557981558526","Increase mid/high bid sizes: bscale1p5; seeds 4000-4199; mean_edge=113.61861908842201 retail_edge=139.5610425354855 arb_edge=-25.9424234470585 failures=0 paired_mean=0.1590619288950008 SE=0.13371722865631921","Boost only high-probability bids: topbid0p95x3; seeds 4000-4199; mean_edge=113.6384521490475 retail_edge=138.782954761541 arb_edge=-25.1445026124825 failures=0 paired_mean=0.17889498952050012 SE=0.04884429545734554","Above 20c refill at 95% of target; multiply bids by 1.5 only above 0.8; symmetric upper-tail belief; seeds 4000-4999; mean_edge=102.1226240020148 retail_edge=128.8873126051738 arb_edge=-26.764688603161 failures=0 paired_mean=0.18417787554080026 SE=0.04563436935799051","Above 20c refill at 95% of target; multiply bids by 1.5 only above 0.95; symmetric upper-tail belief; seeds 4000-4999; mean_edge=102.1375188465659 retail_edge=128.771217378264 arb_edge=-26.6336985316951 failures=0 paired_mean=0.19907272009190036 SE=0.0393006708476538","Faster refill plus mid/high size adjustment: combo_b1p6a1; seeds 4000-4199; mean_edge=113.658889216041 retail_edge=140.041168469369 arb_edge=-26.3822792533115 failures=0 paired_mean=0.19933205651400057 SE=0.19529579579177406","Faster refill plus mid/high size adjustment: combo_b1p4a1p1; seeds 4000-4199; mean_edge=113.665884343721 retail_edge=140.01111306539 arb_edge=-26.345228721626 failures=0 paired_mean=0.20632718419400092 SE=0.17530523016181584","Faster refill with spread-dependent size: spreadcombo1p01p31p41p5; seeds 4000-4199; mean_edge=113.7001089520015 retail_edge=139.821390819867 arb_edge=-26.121281867834 failures=0 paired_mean=0.24055179247450093 SE=0.17594374678793714","Faster refill plus mid/high size adjustment: combo_b1p2a1; seeds 4000-4199; mean_edge=113.75486749691399 retail_edge=139.35326309533448 arb_edge=-25.598395598391498 failures=0 paired_mean=0.29531033738700235 SE=0.12384168610356786","Faster refill plus mid/high size adjustment: combo_b1p4a1; seeds 4000-4199; mean_edge=113.8046030775865 retail_edge=139.777611024454 arb_edge=-25.973007946836002 failures=0 paired_mean=0.3450459180595002 SE=0.169379681689866","Scorer check: unchanged base, seeds 1000-1599; {\"mean_edge\":108.63974289615767,\"retail_edge\":136.2713209771645,\"arb_edge\":-27.63157808101933,\"simulations\":600,\"seed_start\":1000,\"failures\":0}","Initial development baseline; {\"mean_edge\":113.459557159527,\"retail_edge\":138.4995598434705,\"arb_edge\":-25.040002683935498,\"simulations\":200,\"seed_start\":4000,\"failures\":0}","Larger development baseline; {\"mean_edge\":101.93844612647399,\"retail_edge\":128.2601835529157,\"arb_edge\":-26.3217374264462,\"simulations\":1000,\"seed_start\":4000,\"failures\":0}","Full tuning baseline; {\"mean_edge\":98.0996956804193,\"retail_edge\":125.17961744069841,\"arb_edge\":-27.079921760293374,\"simulations\":4000,\"seed_start\":4000,\"failures\":0}","Full unseen baseline; {\"mean_edge\":97.54703502577897,\"retail_edge\":126.92890288111944,\"arb_edge\":-29.381867855334075,\"simulations\":4000,\"seed_start\":50000,\"failures\":0}","Separate validation refill_fast; {\"mean_edge\":94.48844794023701,\"retail_edge\":122.51381010238745,\"arb_edge\":-28.02536216217,\"simulations\":2000,\"seed_start\":6000,\"failures\":0} paired_delta={\"mean\":0.05245481169095018,\"standard_error\":0.025907111776544943,\"retail\":0.2977540156721008,\"arb\":-0.2452992039759501}","Separate validation synth_balanced; {\"mean_edge\":94.48172460144895,\"retail_edge\":123.26654957771544,\"arb_edge\":-28.784824976287698,\"simulations\":2000,\"seed_start\":6000,\"failures\":0} paired_delta={\"mean\":0.04573147290289985,\"standard_error\":0.05452870168022264,\"retail\":1.0504934910000998,\"arb\":-1.0047620180936498}","Separate validation high_refill; {\"mean_edge\":94.53784388204176,\"retail_edge\":122.7867213345975,\"arb_edge\":-28.2488774525727,\"simulations\":2000,\"seed_start\":6000,\"failures\":0} paired_delta={\"mean\":0.1018507534957001,\"standard_error\":0.035459470559968655,\"retail\":0.5706652478821509,\"arb\":-0.4688144943786501}","Separate validation tail_refill; {\"mean_edge\":94.5767981642628,\"retail_edge\":122.68539875973431,\"arb_edge\":-28.108600595489598,\"simulations\":2000,\"seed_start\":6000,\"failures\":0} paired_delta={\"mean\":0.14080503571675015,\"standard_error\":0.03214374868593006,\"retail\":0.46934267301895083,\"arb\":-0.3285376372955501}","Frozen optimized file versus selected parent: 3000 of 3000 overlapping complete SimulationResult records are exactly equal; optimized validation paired_delta={\"mean\":0.14080503571675015,\"standard_error\":0.03214374868593006,\"retail\":0.46934267301895083,\"arb\":-0.3285376372955501}.","Speed/equivalence: final optimized selected parent matched 40000 actions on 20 development seeds, with zero failed simulations; instrumented base matched every SimulationResult field on 200 seeds. Timing: same process, same 200 seeds 4000-4199, eight alternating 25-run blocks; elapsed_seconds={\"base\":206.40725212194957,\"yours\":189.05823257332668} cpu_seconds={\"base\":122.20658505500003,\"yours\":117.00238145700001}.","Single-step diagnostic fit on 1000 development runs, 1520978 mid/high-price states: simply scaling both-side targets upward worsened immediate expected realized edge. Extra depth and earlier refill must be assessed through complete simulator runs because they change future inventory, quote retention, and competitor fills. This diagnostic was not used as the reported strategy score.","Frozen unseen outcome attribution: {\"NO\":{\"simulations\":2000,\"paired_delta\":{\"mean\":0.030160645249899993,\"standard_error\":0.033770273558087434,\"retail\":0.28699782485440006,\"arb\":-0.25683717960350017}},\"YES\":{\"simulations\":2000,\"paired_delta\":{\"mean\":0.4212498272789001,\"standard_error\":0.059583849752574546,\"retail\":0.6950257329033999,\"arb\":-0.2737759056273499}}}","Selected frozen strategy final.py; full tuning paired_delta={\"mean\":0.1495560002953501,\"standard_error\":0.021143127933978118,\"retail\":0.48042886108840077,\"arb\":-0.3308728607877}; unseen paired_delta={\"mean\":0.22570523626440003,\"standard_error\":0.03437928452040865,\"retail\":0.4910117788789,\"arb\":-0.265306542615425}; code_sha256=4d34c482983b88e18053535d96c3fc256d0fb8baf8c5ec607e3097098276b297"],"next":"Graft the above-20-cent refill rule onto independently validated gains, then test on fresh seeds and Paradigm. The broad depth increases, extra inner price levels, and joining behind the competitor mostly bought retail at excessive arbitrage cost. Investigate selective queue-continuation value or observable demand for the remaining 20-80-cent opportunity, while keeping the existing low-price inventory funding. The 50000-53999 seeds in this result are now used; do not use them as a fresh holdout."}