Compares every Kalshi "Winner?" market against sportsbook consensus from 20+ bookmakers. Only trades when Kalshi is 5%+ cheaper.
This is the highest-priority strategy in the engine (score: 95). It works by pulling real-time odds from The Odds API across 20+ US bookmakers, computing the median implied probability for each game, and comparing it to the Kalshi YES price.
When the bookmaker consensus says a team has a 75% chance to win but Kalshi is pricing YES at 65¢, that's a 10% edge. The engine buys before Kalshi adjusts.
Pulls real-time odds from The Odds API. Median consensus across all bookmakers. Kalshi price vs bookie implied probability = your edge.
The Odds API aggregates h2h (head-to-head) odds from every major US sportsbook in real time. For each game, it returns odds from DraftKings, FanDuel, BetMGM, Caesars, PointsBet, and more.
The engine converts American odds to implied probability, then takes the median across all books (not the mean — this removes outlier books). The median is the truest representation of market consensus.
Both teams must match between the Kalshi title and bookmaker event. The engine uses fuzzy team name matching with city-word filtering to prevent "New York Yankees" from matching "New York Mets."
Dynamic bet sizing: f* = (edge × price) / (1 - price). Quarter-Kelly for safety. 10% edge at 70¢ = ~5.8% of bankroll.
The Kelly Criterion is the mathematically optimal way to size bets given your edge. It maximizes long-run growth while avoiding ruin. We use quarter-Kelly (25% of full Kelly) for conservative sizing.
Formula: kelly_f = (edge × price) / (1 - price) — then multiply by 0.25 for quarter-Kelly.
Full Kelly is mathemoptimal but produces gut-wrenching drawdowns. Quarter-Kelly gets you 75% of the growth rate with substantially less variance.
Only trades with 5%+ edge (covers fees + slippage). More bookmakers confirming = higher conviction.
Not every gap between Kalshi and sportsbooks is worth trading. Fees, slippage, and execution risk eat into your edge. The minimum edge gate ensures you only trade when the expected profit exceeds all costs.
The gate also rejects trades where sportsbooks disagree with our direction by more than 10%. If we're buying YES but the books think NO is more likely, the trade is blocked.
Monitors ESPN live scores every 30 seconds. When a team scores, Kalshi prices lag — we buy before the market adjusts.
This is the fastest edge in the engine (score: 85). While Cross-Market +EV has the strongest mathematical edge, ESPN Live Edge has the fastest execution. The edge exists for seconds, not hours.
The engine polls ESPN's live scoreboard API every 30 seconds. It tracks the previous score for every active game. When it detects a change (a team just scored), it immediately looks up the Kalshi market for that game.
Each scoring event triggers a Kalshi market lookup. If the market hasn't adjusted yet, buy the scoring team's YES.
When a goal is scored in hockey or a run is plated in baseball, Kalshi's market price doesn't update instantly. There's a 5-30 second lag where the old price is still available. This lag is the edge.
The engine doesn't just buy blindly — it calculates an expected win probability based on the scoring event:
Works across NHL, MLB, soccer (7 leagues), tennis (ATP), and cricket (IPL). Each sport has custom scoring models.
The ESPN Live Edge strategy covers 8+ leagues with sport-specific probability models:
Each sport's ESPN feed is mapped to the correct Kalshi series ticker. The engine uses a team abbreviation map to match ESPN teams to Kalshi market titles.
60-second cooldown per game prevents over-trading. Only fires on genuine score changes, not clock updates.
Without cooldown, a game with rapid scoring (e.g., 3 goals in 2 minutes) would trigger 3 separate trades, all chasing the same momentum. The cooldown prevents this.
Only trades when ESPN confirms a blowout in progress. Sweet spot: 71-89¢ entry price. Proven over 125+ trades.
Live Blowout is the proven workhorse (score: 80). It buys YES on the team that's winning big — but only when ESPN confirms the blowout is real.
The engine checks ESPN's live scoreboard for games where one team has a commanding lead:
Only when ESPN confirms the blowout does the engine look for a Kalshi "Winner?" market at 71-89¢. The price gate is critical — below 70¢ is a graveyard, above 89¢ has thin profit margins.
CS2 (+$6.09), ATP Tennis (+$6.00), LoL (+$3.87), Serie A, KBL, AFL, IPL, NHL — these sports have reliable blowout patterns.
After analyzing 125+ trades across all sports, certain sports consistently produce profits with the Live Blowout strategy:
NBA, WTA, NPB, UFL, ATP Challenger, KBO — these sports have unpredictable momentum swings that break the strategy.
Some sports are banned from the Live Blowout strategy because they consistently produce losses despite appearing to have blowout conditions:
Multi-layer filters: 70¢ hard floor, sport-specific minimums, spread limits, volume requirements, ESPN confirmation.
Every potential Live Blowout trade passes through 6 layers of filters before it can execute:
Monitors live prices for rapid drops (>10¢ in 5 minutes). When a team goes on a run, the market overreacts — we buy the underdog.
This is the most speculative strategy (score: 60). It's based on the behavioral finding that prediction markets overreact to scoring runs.
The engine tracks the YES price for every active "Winner?" market. It maintains a 5-minute rolling window of prices. When the price drops by more than 10¢ in 5 minutes, it signals an overreaction.
Take profit at 50¢+ (2x+ return). Stop-loss at 5¢. Max 2 contracts — speculative bets with asymmetric payoff.
Mean Reversion has a completely different exit profile than the other strategies:
Expected math: ~35% win rate × 3x average win = 1.05x expected value. You lose often but win big when you're right. The asymmetric payoff makes it profitable over volume.
The opposite of Live Blowout. Instead of buying the favorite on a run, buy the other team when the market overprices the lead.
Live Blowout and Mean Reversion are opposite strategies:
This isn't a contradiction — they operate in different price ranges. Live Blowout buys at 71-89¢ (favorites). Mean Reversion buys at 5-35¢ (underdogs). They're trading different contracts on the same game.
The research behind Mean Reversion comes from behavioral finance: humans over-weight recent events. A team that just scored 3 goals is treated as if they'll keep scoring at that rate. But the underlying win probability hasn't changed as much as the price suggests.