How 9to5bets finds edge
A fully automated pipeline that compares Kalshi sports totals against sharp bookmaker consensus, filters ruthlessly, and alerts you only when the math checks out.
The problem: structural inefficiency
Kalshi sports totals markets are structurally less efficient than traditional sharp sportsbooks. Why? Kalshi is a prediction exchange — prices are set by retail order flow, not by professional market-makers with sophisticated models. When a sharp bookmaker consensus (e.g. Pinnacle, Circa) prices a total at a certain level and Kalshi's price lags behind, a quantifiable mispricing exists.
These gaps are small (often 3-10 cents), short-lived, and require systematic detection to capture. That's where the pipeline comes in.
Our pipeline: 5 layers of filtering
1.Data ingestion
Every 10 minutes, the pipeline fetches live Kalshi market data and current sharp bookmaker lines from the Odds API. It normalizes teams, matches events, and identifies totals markets with two-sided pricing.
2.Signal-0 lag gate
Before any math runs, we check if the Kalshi price is stale enough to signal. The lag gate measures the raw gap between Kalshi's mid and the sharp consensus, rejecting noise and only passing markets with meaningful dislocation.
3.Probability model
A multi-sigma model estimates the true probability of each outcome using the sharp consensus as the expected value. Edge is calculated at multiple sigma levels (1.2, 1.3, 1.4) to ensure robustness. All three scenarios must show positive edge after fees.
4.Suppression + anti-filters
Context-aware signals reduce strength when conditions are risky: fatigue (back-to-backs, rest days), tempo mismatches, player availability changes, and drift convergence. Anti-filters catch edge cases like close-game scenarios, OT risk, and foul trouble.
5.Tier scoring + liquidity gate
Each signal receives a composite strength score (0-14). Only signals above tier thresholds with sufficient Kalshi liquidity pass through. A-tier signals (strongest) trigger instant alerts; B-tier signals are logged for monitoring.
What you get
A-Tier Alerts
Instant Discord + email notifications with the exact market, side, target price, and edge percentage.
Signal Detail
Full analysis breakdown: strength score, key factors (fatigue, tempo, availability), and expandable advanced model data.
CLV Tracking
Settled A-tier signals get a Kalshi NO-side CLV read after the game so you can see whether we beat the exchange line — not hand-waved picks.
How to act on a signal: When you receive an A-tier alert, place a maker limit order on Kalshi at or below the target price. The recommended stake percentage accounts for your edge and the market's liquidity.
CLV: proving the edge
Closing Line Value (CLV) on 9to5bets is measured on the Kalshi contract for our totals strategy: we quote signals on the NO side (unders). After the game, once the Kalshi market has a reliable settlement print (several hours post-start), we compare that NO price to the NO target price stored at detection. CLV = NO price at close snapshot minus NO target at signal (in cents on the $1 contract). Positive CLV means the NO side finished higher on Kalshi than our reference entry — we beat the Kalshi closing line on that side.
CLV here is not realized profit: we do not subtract maker fees or slippage from this number (fees are surfaced separately at signal time). It is still the standard way to ask whether you are getting better prices than the market you actually trade. Win/loss is noisy in small samples; CLV is a cleaner quality signal — but no single metric guarantees profit after costs.
By sport — same A-tier, deduped-per-game rules
| Sport | Avg CLV | Median | Hit rate | Games |
|---|---|---|---|---|
| NBA | +2.3¢ | +2.0¢ | 67% | 6 |
FAQ
Is this gambling advice?
No. 9to5bets is a data analysis tool that identifies structural mispricings. We surface opportunities with quantified edge; you decide whether and how to act. This is not financial or gambling advice.
How often do signals fire?
It depends on market conditions. During active NBA/MLB seasons, A-tier signals typically fire a few times per week. We intentionally maintain a high bar — quality over quantity.
What sports are covered?
Currently NBA and MLB totals (unders). We focus on markets where our pipeline has demonstrated consistent CLV. New sports are added only after rigorous backtesting.
What's the catch during beta?
None. We're building a public track record and iterating on the product. Beta access is free. When we launch paid tiers, early users will be grandfathered in with the best terms.
Ready to see the signals?
Join during beta — free access to A-tier alerts, full signal analysis, and transparent CLV tracking.
Get started free