{"sources": [{"url": "https://api.imd.fun/artifacts/41fa8829432f502a00c5302c0e2200b475c5c2e100a2a16b086aaab0ce62a395", "quote": "\"approach\": \"Graft without constant tuning onto crossing-cover artifact 40497298132dcc0ccc9f4c0672031c7eaae00ddcf8cf24983db989feeb3c5ca0. Spread 1: dmin=0.3 and pfl=0.000001. Spread 3: copy the supplied guarded, clipped-fill-corrected flow estimator "}, {"url": "https://api.imd.fun/artifacts/40497298132dcc0ccc9f4c0672031c7eaae00ddcf8cf24983db989feeb3c5ca0", "quote": "\"approach\": \"Uses tiny tick probes, a sub-tick probability filter, and marginal retail-versus-arbitrage sizing to quote one tick inside the competitor. Starting from the exactly reproduced base, adds demand-matched crossing cover at ticks 2-9. When e"}, {"url": "https://api.imd.fun/artifacts/c80c33382f9c458ab6286dbc11fa761100a7e156dad7f3c12d74040b7c76578f", "quote": "\"approach\":\"Spread 4 only: estimate each run's symmetric buy/sell retail flow from changes in our main resting orders. Use pre-fill exposure eligibility, conservative price-distance filters, and right-censored size likelihoods when orders are exhaust"}, {"url": "https://api.imd.fun/artifacts/89954f578f94920227812b158e22ece964c6e34081ca1c89a022dc60cadafba4", "quote": "\"approach\": \"For self.s == 4, halve the regular-bid target when phat < 0.05 and bid_t >= 2. Retail sells often consume only part of resting depth, whereas arbitrage sweeps remove all overpriced depth. The fill audit showed substantial arbitrage losse"}, {"url": "https://api.imd.fun/artifacts/9cff8842bd7d225500799ce5ad8289e5ace7fc7656a6195032516f290d978d8f", "quote": "\"approach\": \"Spread 4, phat < 5c, ask side only: the sticky rule that keeps our ask AT the rival's best ask tick (where an order of ours queued before the rival's replenished order stays ahead of it) now needs at least 50 shares of that ahead-of-riva"}, {"url": "https://api.imd.fun/artifacts/6c98aea3a2ae09a046b6d26f565731b6fdbd2c9fa375437c7fb5a8bb293dd412", "quote": "\"approach\": \"Only self.s == 4 changes. Preserve FIFO on existing front SELL orders at the rival own tick by tolerating remaining quantity up to 3 times current desired size rather than 1.5 times. The base cancels oversized orders and loses priority; "}, {"url": "https://api.imd.fun/artifacts/c8288b4af29e8f178adf9a933a9303c7f4fa5f97187d604d54bf1071e72ae872", "quote": "\"approach\": \"Spread 4 only, quotes between 5 and 95 cents (pL <= phat <= 0.95): the expected arbitrage loss used to size our ask is multiplied by 1.4, so asks there are sized with the loss weighted about 1.0 (0.7*1.4) instead of kA=0.7. Bids and all "}, {"url": "https://api.imd.fun/artifacts/f83584120dba5fcc36de9ce39fafe067432927b557d15dc9021e749064c71afd", "quote": "\"approach\": \"Spread 4 only: split a new bid into two cent-quantized halves when it is exactly four ticks below a real rival ask (ca<=99), strictly above the rival bid, and worth at least 4*P['m'] ($20). Total placed quantity and the original size rul"}, {"url": "https://raw.githubusercontent.com/danrobinson/prediction-market-challenge/b244530ee2a6247b5f930830eeca9dd950da0e7e/orderbook_pm_challenge/runner.py", "quote": "    rng = random.Random(seed)\n    process = replace(\n        base_config.process,"}]}