Gaming & Esports Desk — Panel Ruling
Date: March 28, 2026
Panel: Claude Opus 4.6, Claude Sonnet 4.6, Gemini Pro 3.1, Grok 4.2
Judge: Boss
Grade: APPROVED — Build all 4 phases
Ruling Summary
The panel unanimously recommended building a full Gaming & Esports trading desk on Kalshi. All 4 panelists agreed on the core strategy and architecture. The boss approved all 4 implementation phases.
Key Decisions
Anchor: Sharp Composite Line (SCL)
- Unanimous. Pinnacle is the primary anchor but weaker for esports than traditional sports (wider vig, lower limits, slower updates).
- Build a weighted no-vig composite across Pinnacle (~50% weight) + bet365 + GG.bet.
- Pinnacle scraper is PRIMARY. The Odds API is BACKUP ONLY.
- For markets with no book anchor (Game Awards, Steam rankings, Twitch subs): we ARE the model.
Market Inefficiencies (all 4 panelists agreed)
- Roster change lag (12-24 hours before books reprice)
- Patch meta shifts (2-6 hour window post-patch)
- Map veto asymmetry (3-8% edge from map-specific modeling)
- Online vs LAN performance gap (documented but not priced)
- Casual bettor bias on Kalshi (favorites overpriced)
Highest-Value Metric: Map Veto Prediction Model
- All 4 panelists independently designed this as the #1 proprietary edge.
- Monte Carlo simulation of ban/pick sequences using historical veto patterns.
- Expected edge: 3-8% on map-dependent markets.
- Uses entirely public data (HLTV/VLR.gg ban/pick histories).
Custom Metrics Built (panel consensus)
- Map Veto Monte Carlo (CS2/Valorant) — 10K simulations, log5 probability
- Roster Change Impact Score (RCIS) — role-weighted with synergy penalty
- Patch Sensitivity Score (PSS) — usage × nerf magnitude × team dependency
- Online-LAN Adjustment Factor (OLAF) — LAN_WR / Online_WR + travel fatigue
- Custom Elo System — K-factor adjustments for LAN/tier/roster/patch + decay
- Tournament Bracket Simulator — 20K sims, 4 formats, fatigue/rest modifiers
- CS2 Pistol/Economy Model — pistol round impact + eco rating
- Early Game Composite (LoL/Dota 2) — weighted z-score of FB/FT/FD/GD@15
- Patch Meta Fluidity Index — team adaptability across patches
Novel Edge Vectors (unique contributions by panelist)
| Edge Vector |
Proposed By |
Status |
| Scrim leak monitoring (Reddit/Twitter/Discord) |
All 4 |
Built — sentiment-monitor.ts |
| Patch notes speed advantage (auto-parse within minutes) |
Opus, Sonnet, Gemini |
Built — patch-calendar.ts + patch-sensitivity.ts |
| SteamDB depot monitoring (pre-patch detection) |
Opus |
Built — steam-tracker.ts |
| Pro player practice tracking (Steam hours, FACEIT Elo) |
Opus |
Built — player-tracker.ts |
| Pro player account stalking (solo queue picks) |
Gemini |
Not built (future) |
| Tournament bracket simulation |
Opus, Sonnet |
Built — bracket-simulator.ts |
| Favorites bias calibration (6.5M data) |
Sonnet |
Not built (data analysis task) |
| Visa/travel disruption monitoring |
Sonnet |
Covered by sentiment-monitor.ts keywords |
Data Sources Approved
- 33 data sources across odds, stats, meta, sentiment, intelligence
- HLTV.org (CS2), VLR.gg (Valorant), Oracle's Elixir (LoL), OpenDota (Dota 2), Liquipedia (all)
- SteamDB, SullyGnome, Reddit sentiment, player activity tracking
Build Phases
| Phase |
Content |
Status |
| Phase 1: Foundation |
Pinnacle esports URLs, Odds API backup, Elo system, HLTV/VLR/Liquipedia scrapers |
BUILT + 2x SIMPLIFIED |
| Phase 2: Core Models |
Map veto, RCIS, patch calendar, Oracle's Elixir, OpenDota |
BUILT + 2x SIMPLIFIED |
| Phase 3: Proprietary |
PSS, OLAF, bracket sim, CS2 economy, early game composite |
BUILT + 2x SIMPLIFIED |
| Phase 4: Signals |
Steam/Twitch pipeline, Reddit sentiment, player tracking |
BUILT + 2x SIMPLIFIED |
Market Size
6.5M+ settled Kalshi volume across esports:
- FIFA: 7.4M | CoD: .2M | LoL: .2M+ | CS2: .1M | Valorant: .5M | R6: 79K
Kalshi Coverage
109 series tracked across: match markets (23), tournament winners (18), Esports World Cup (21), Game Awards/Rankings (13), Steam Awards (12), Twitch/Streamer (15), other (7).
What's Next
- Wire all scrapers into the scheduler (cron jobs)
- Run favorites bias calibration on settled volume data
- Backfill Elo ratings from historical match data
- Build pro player account stalking module (Riot API solo queue monitoring)
- Create formal desk spec document
Source: ~/edgeclaw/results/panel-results/esports-desk-ruling.md