Block path-dependent settlement barriers

This commit is contained in:
Theodore Song
2026-08-18 12:41:22 -04:00
parent 25622d2ee5
commit c78a13bbb7
3 changed files with 34 additions and 24 deletions
+17 -12
View File
@@ -15,7 +15,7 @@ https://polymarket-site-eta.vercel.app/personal.html
The site fetches live Polymarket markets, generates agent suggestions, lets you
run frequent paper cycles, and syncs the shared arena state through Neon or
Vercel Blob. Build 48 also installs an offline app shell and caches timestamped
Vercel Blob. Build 49 also installs an offline app shell and caches timestamped
market snapshots. During an outage, cycles continue locally; cached entries are
allowed for 90 minutes, older snapshots become mark-only, and all cached data
expires after 24 hours.
@@ -27,7 +27,7 @@ The learner shrinks small samples toward neutral, caps sizing changes to
15% of candidates for deterministic exploration so a stale regime cannot become
permanent.
Strategy 45 treats each binary stake as capable of falling to zero even when the
Strategy 46 treats each binary stake as capable of falling to zero even when the
18% stop cannot fill. New core positions are capped at 2.5%-4% of equity and
aggressive positions at 3%-5%, with lower limits for near-term, extreme-price,
reversal, and fast-moving setups. Oversized positions inherited from older
@@ -65,7 +65,7 @@ negative in all three chronological segments. Reversals averaged -3.69%, with a
market-clustered 90% interval entirely below zero. Crypto and Sports were also
negative but covered only three and five markets. The 24-hour cohort improved to
-0.82% row mean and +1.31% market mean, with no rule robustly negative across all
segments. Strategy 45 therefore keeps reversal entries observation-only until
segments. Strategy 46 therefore keeps reversal entries observation-only until
their recent signal and quality cohorts independently earn promotion, retains
their signals for paper grading, and evaluates adaptation at 24 hours.
@@ -74,7 +74,7 @@ markets with no failures and produced 3,597 net-of-cost 24-hour outcomes across
142 markets. No tested follow or fade rule was robustly positive. Crypto trends
averaged -3.83% per observation and -3.99% per market; Sports trends averaged
-5.44% and -6.61%. Both stayed negative in every chronological segment and their
market-clustered 90% intervals were entirely below zero. Strategy 45 therefore
market-clustered 90% intervals were entirely below zero. Strategy 46 therefore
keeps Crypto and Sports trends observation-only while continuing to grade them.
A corrected 200-market audit paged through 197 markets with usable history and
@@ -83,23 +83,23 @@ segment and averaged -4.13%. Sports trends were negative in train and test and
averaged -3.53% at 72 hours. Politics trends were the sole cohort with positive
row-level returns in all three 72-hour segments, but its market-cluster interval
still crossed zero; that supports a longer hold test, not a larger entry bet.
Strategy 45 gives Politics trend positions that 72-hour observation window before
Strategy 46 gives Politics trend positions that 72-hour observation window before
ordinary signal exits. Stops, profit locks, settlement handling, and risk-budget
reductions remain immediate.
Strategy 45 also subtracts a half-cent round-trip cost when grading each live
Strategy 46 also subtracts a half-cent round-trip cost when grading each live
walk-forward signal. Confidence uses the largest independent matching bucket,
not the sum of five overlapping feature buckets, and evidence from older engine
versions is down-weighted. This prevents a handful of duplicated observations
from authorizing larger positions or hiding a modest negative regime.
Strategy 45 adds uncertainty-aware promotion and demotion. A matching setup must
Strategy 46 adds uncertainty-aware promotion and demotion. A matching setup must
accumulate at least eight effective observations and agree across at least two
feature views before repeatable positive evidence can increase size or repeatable
negative evidence can block a new entry. Mixed evidence stays close to neutral
instead of being mistaken for an edge.
Build 48 enforces the documented offline boundary end to end. Cached snapshots
Build 49 enforces the documented offline boundary end to end. Cached snapshots
under 90 minutes old may continue paper execution. Older snapshots remain usable
for valuation and chart snapshots for up to 24 hours, but cannot trigger entries,
stop-losses, gain-stops, risk rebalances, settlements, or policy exits. Network
@@ -112,12 +112,12 @@ adaptive baselines, pending signal grades, and trade evidence remain in one stra
lineage until the actual entry, sizing, or exit logic changes. Legacy build 40 and 41
records are migrated into the same strategy lineage without losing evidence.
Build 48 independently refreshes markets for matured pending signals that have
Build 49 independently refreshes markets for matured pending signals that have
left the current top-500 activity scan. Unavailable markets remain queued for a
bounded retry window. This prevents activity-rank survivorship from deciding
which wins and losses reach the adaptive calibration ledger.
Strategy 45 coordinates high-risk exploration globally. Near-term, extreme-price,
Strategy 46 coordinates high-risk exploration globally. Near-term, extreme-price,
and other gap-prone positions may be held materially by only one agent, while
ordinary independently confirmed markets retain the two-agent cap. The robustly
negative Sports- and Crypto-trend cohorts cannot enter through exploration.
@@ -139,7 +139,7 @@ whole-event YES or NO bundles using executable best asks/bids, per-leg costs, an
a minimum-liquidity requirement. The 500-event audit found 33 complete liquid
negative-risk events and zero positive worst-case bundle returns after costs.
Midpoint price sums sometimes looked attractive, but executable spreads removed
the apparent edge, so Strategy 45 does not pretend those snapshots are arbitrage.
the apparent edge, so Strategy 46 does not pretend those snapshots are arbitrage.
The expanded event-clustered run loaded history for 498 of the 500 highest-volume
resolved markets with no fetch failures. No side, price band, category, trend,
@@ -150,13 +150,18 @@ static settlement-direction boost from this audit.
The earlier 200-resolved-market audit found short-dated NO entries strongly
negative, but the 500-market rerun did not reproduce that loss in its newer test
segment. Strategy 45 therefore treats the result as a provisional prior instead
segment. Strategy 46 therefore treats the result as a provisional prior instead
of a permanent ban: NO entries with 21 days or less remain observation-only until
the recent walk-forward calibration promotes their matching side and duration
cohorts. Exact numeric-range contracts are excluded from new entries because a
settlement jump can pass directly through an 18% stop; the live audit found that
this failure mode caused the largest latest-day loss.
Strategy 46 also excludes path-dependent barriers such as "reach $66,000," "hit
$90," and "dip to $62,000." These contracts can resolve abruptly as soon as the
barrier is touched, so a later hourly stop cannot reliably cap the loss. Fixed-date
level questions such as "above $66,000 on August 23" remain eligible.
Paper accounts created with a password are also saved through the backend, so a
user can log in from another device and see the same paper portfolio, activity,
and value history. Passwordless paper accounts remain local-only.
+16 -11
View File
@@ -341,7 +341,7 @@ footer{margin-top:34px;padding-top:22px;border-top:1px solid var(--border);color
<nav class="topnav">
<div class="brand">
<div class="logo">🏆</div>
<div><div class="brand-name">Polymarket Arena</div><div class="brand-sub">10 agents · 5 core + 5 aggressive</div><div class="build-badge">Adaptive strategy 45 · build 48</div></div>
<div><div class="brand-name">Polymarket Arena</div><div class="brand-sub">10 agents · 5 core + 5 aggressive</div><div class="build-badge">Adaptive strategy 46 · build 49</div></div>
</div>
<div class="tabs" id="tabs">
<button class="tab" data-tab="overview">Overview</button>
@@ -363,7 +363,7 @@ footer{margin-top:34px;padding-top:22px;border-top:1px solid var(--border);color
<div class="personal-banner" id="personalBanner">
<b>Personal research mode.</b> This copy is for your own analysis, paper tracking, and manual trade research only. It does not pool money, onboard investors, custody funds, bypass eligibility rules, or place orders without your manual approval.
</div>
<div class="live-build-banner"><b>Build 48 active:</b> Crypto and Sports trends are observation-only after a 500-market walk-forward audit found repeatable 24-hour losses. Reversal and short-dated NO signals can return only after their own recent cohorts earn promotion. Exact-range contracts remain excluded because settlement jumps can bypass stops. This remains paper trading; profits are not guaranteed.</div>
<div class="live-build-banner"><b>Build 49 active:</b> Crypto and Sports trends are observation-only after a 500-market walk-forward audit found repeatable 24-hour losses. Reversal and short-dated NO signals can return only after their own recent cohorts earn promotion. Exact ranges and path-dependent "reach / hit / dip" barriers are excluded because abrupt resolution can bypass stops. This remains paper trading; profits are not guaranteed.</div>
<!-- ============ OVERVIEW ============ -->
<section class="tabpanel" data-tab="overview">
@@ -745,7 +745,7 @@ footer{margin-top:34px;padding-top:22px;border-top:1px solid var(--border);color
</section>
<footer>
Build 48 · Adaptive strategy 45 · Paper trading only · Live prices from Polymarket's public Gamma API · Not financial advice ·
Build 49 · Adaptive strategy 46 · Paper trading only · Live prices from Polymarket's public Gamma API · Not financial advice ·
<a class="market-link" href="https://github.com/theodore-song/polymarket-analyst" target="_blank" rel="noopener">Source on GitHub</a>
</footer>
</div>
@@ -774,9 +774,9 @@ const POLITICS_TREND_MIN_HOLD_HOURS = 72;
const EXIT_CONFIRM_HOURS = 6;
const AGENTS_KEY = "pma_agents_v2";
const SUG_KEY = "pma_suggestions_v5";
const BUILD_VERSION = 48;
const SUGGESTION_ENGINE_VERSION = 45;
const PREVIOUS_STRATEGY_VERSION = 44;
const BUILD_VERSION = 49;
const SUGGESTION_ENGINE_VERSION = 46;
const PREVIOUS_STRATEGY_VERSION = 45;
const LEGACY_BUILD_STRATEGY_LINEAGE = Object.freeze({40:40,41:40});
function normalizedStrategyVersion(value){
const version=Number(value||0);
@@ -1175,7 +1175,8 @@ function fastSettlementRisk(m){
const numericRange=/\b\d+(?:\.\d+)?\s*(?:%|percent)?\s*(?:-||—|to)\s*\d+(?:\.\d+)?\s*(?:%|percent|votes?|points?|seats?|bps|basis points?|tweets?|posts?|goals?)(?![a-z])/;
const currencyRange=/[$€£]\d+(?:\.\d+)?\s*(?:-||—|to)\s*[$€£]?\d+(?:\.\d+)?/;
const betweenRange=/\bbetween\s+[$€£]?\d+(?:\.\d+)?\s*(?:%|percent)?\s+(?:and|to)\s+[$€£]?\d+(?:\.\d+)?/;
if(numericRange.test(text)||currencyRange.test(text)||betweenRange.test(text))return true;
const pathDependentBarrier=/\b(?:reach|hit|touch|dip(?:\s+(?:to|below))?|rise\s+(?:to|above)|fall\s+(?:to|below))\b[^?]{0,45}(?:[$€£]\s*)?\d[\d,]*(?:\.\d+)?\s*(?:%|percent|k|m|b|points?|bps|basis points?)?/;
if(numericRange.test(text)||currencyRange.test(text)||betweenRange.test(text)||pathDependentBarrier.test(text))return true;
if(/\bexact score\b|\bscore:\s*\d|\bposts? \d+-\d+|\bnumber of (tweets|posts)\b/.test(text))return true;
if(/\bvs\.?\b/.test(text))return true;
if((m&&m.category)==="Sports"){
@@ -4321,7 +4322,7 @@ window.PMA_ENGINE_DIAGNOSTICS=Object.freeze({
staleCacheIsMarkOnly:true,strategyEvidenceSurvivesBuilds:true,signalEvaluationHours:SIGNAL_EVAL_HOURS,signalRetryHours:SIGNAL_LEDGER_RETRY_HOURS,
signalDueFetchLimit:SIGNAL_LEDGER_DUE_FETCH_LIMIT,survivorshipSafeSignalGrading:true,
signalRoundTripCostCents:SIGNAL_ROUND_TRIP_COST*100,independentConfidence:true,uncertaintyGatedCalibration:true,
historicalPrior:"Sports and Crypto trends observation-only; reversal and short-dated NO require recent cohort promotion; exact ranges excluded; longshots and YES sized down"}),
historicalPrior:"Sports and Crypto trends observation-only; reversal and short-dated NO require recent cohort promotion; exact ranges and path-dependent barriers excluded; longshots and YES sized down"}),
});
function runEngineSelfTest(){
const market=(overrides={})=>Object.assign({
@@ -4337,6 +4338,8 @@ function runEngineSelfTest(){
const noSignal=analyzeMarket(market({price_change_1h:0,price_change_1d:0.002,price_change_1w:-0.002}),news);
const reversal=analyzeMarket(market({price_change_1h:-0.01,price_change_1d:0.07,price_change_1w:0.02}),news);
const intervalContract=analyzeMarket(market({id:"range-test",question:"Will Count Binface win 2030% of votes?"}),news);
const pathBarrierContract=analyzeMarket(market({id:"barrier-test",question:"Will Bitcoin reach $66,000 August 17-23?",category:"Crypto"}),news);
const fixedDateLevelContract=analyzeMarket(market({id:"fixed-level-test",question:"Will Bitcoin be above $66,000 on 2026-08-23?",category:"Crypto"}),news);
const ordinaryDatedContract=analyzeMarket(market({id:"dated-test",question:"Will the policy pass on 2026-08-22?"}),news);
const targets=window.PMA_ENGINE_DIAGNOSTICS.gainStopTargets(0.82);
const hoursAgo=h=>new Date(Date.now()-h*3600000).toISOString();
@@ -4462,11 +4465,11 @@ function runEngineSelfTest(){
delete executableBook.positions[0].price_status;
markToMarket(executableBook,{"offline-mark":market({id:"offline-mark",yes_price:0.8,no_price:0.2})},AGENTS[0],{policyExits:true,executeTrades:true});
const buildMigrationState=defaultState();
buildMigrationState.engine_version=47;buildMigrationState.strategy_version=SUGGESTION_ENGINE_VERSION;
buildMigrationState.engine_version=48;buildMigrationState.strategy_version=SUGGESTION_ENGINE_VERSION;
buildMigrationState.agents.value.engine_baseline={version:SUGGESTION_ENGINE_VERSION,started_at:hoursAgo(2),equity:9876.54};
reconcileStateVersions(buildMigrationState);
const strategyMigrationState=defaultState();
strategyMigrationState.engine_version=47;strategyMigrationState.strategy_version=PREVIOUS_STRATEGY_VERSION;
strategyMigrationState.engine_version=48;strategyMigrationState.strategy_version=PREVIOUS_STRATEGY_VERSION;
strategyMigrationState.agents.value.cash=9876.54;
strategyMigrationState.agents.value.engine_baseline={version:PREVIOUS_STRATEGY_VERSION,started_at:hoursAgo(2),equity:10000};
reconcileStateVersions(strategyMigrationState);
@@ -4512,7 +4515,9 @@ function runEngineSelfTest(){
noisyEvidenceStaysNeutral:noisyCalibration.state==="observing"&&noisyCalibration.multiplier>=0.99&&noisyCalibration.multiplier<=1.01,
blocksConfirmedLosingRegime:!lossOpportunity.allowed,containmentMode:lossDecision.mode,containmentMaxNew:lossDecision.maxNew,
legacyLossDoesNotFreezeCurrentEngine:legacyLossDecision.mode!=="Loss Regime Containment"},
riskBudget:{core:coreRiskBudget,aggressiveGap:aggressiveGapBudget,blocksExactRange:intervalContract.jump_risk&&!intervalContract.trade_ready,preservesOrdinaryDatedContract:!ordinaryDatedContract.jump_risk,boundedStake:boundedStakeForRisk(10000,1000,riskCfg,{entry_price:0.82,days_to_resolution:5,price_change_1d:0.06}),
riskBudget:{core:coreRiskBudget,aggressiveGap:aggressiveGapBudget,blocksExactRange:intervalContract.jump_risk&&!intervalContract.trade_ready,
blocksPathDependentBarrier:pathBarrierContract.jump_risk&&!pathBarrierContract.trade_ready,preservesFixedDateLevel:!fixedDateLevelContract.jump_risk,
preservesOrdinaryDatedContract:!ordinaryDatedContract.jump_risk,boundedStake:boundedStakeForRisk(10000,1000,riskCfg,{entry_price:0.82,days_to_resolution:5,price_change_1d:0.06}),
legacyNormalizedValue:riskPos.value,equityPreserved:+equity(riskBook).toFixed(2),capsBinaryGap:aggressiveGapBudget===0.03&&riskPos.value<=300.01},
offline:{fresh:offlineCachePolicy(30*60000),staleEntry:offlineCachePolicy(3*3600000),expired:offlineCachePolicy(25*3600000),
staleMarkUpdatesValue:markOnlyBook.positions.length===1&&markOnlyBook.positions[0].value===800,
+1 -1
View File
@@ -1,4 +1,4 @@
const CACHE_NAME = "polymarket-arena-build-48";
const CACHE_NAME = "polymarket-arena-build-49";
const APP_SHELL = ["/", "/index.html", "/personal.html", "/cycle-worker.js"];
self.addEventListener("install", event => {