A two-rail framework for live sports markets that treats volatility as the product, not the picks. Polymarket is the liquid rail · Underdog/Betr/PP/DK is the locked architectural-insurance rail. You ride the bike. We hold the wheels. Built on real reps · operator-honest · 12+ doctrine layers deep.
$500 flat. A full operator-day on Polymarket with SideGuy holding the architectural insurance on every position you take. You bet the favorites · we absorb the variance · you keep the framework. Book a Day →
The platforms have opposite liquidity properties. That asymmetry isn't a bug to work around — it's the whole game. You use them differently because they ARE different.
The Lab compounded over six months of real reps. Each layer is a structural pattern with a name, not a clever trick. Naming the layer is what lets you reuse it without re-inventing it next slip.
Stake the high-probability core (~58% leg) bigger, hedge the low-probability tail with a smaller stake at plus odds. Reduces variance, accepts ~$1 of EV cost as cognition tax.
Bet over on one platform's line + under on another platform's higher line. If the final stat lands between the two lines, both cash. Pure cross-platform line-gap exploit.
Smaller version of The Middle — single platform, exploit the line-shift between when you place and when the line moves. Use Poly's tradeability to capture the shift.
Underdog's 2-of-3 (or 3-of-4) soft floor. Lose one leg, still cash partial. Matches UD's no-exit reality — the Flex floor is UD's only built-in "exit-equivalent."
Every position should produce two paydays — settle outcome + a second compounding-side outcome (mid-event exit, partial floor, doctrine validation). Refuse single-output positions as inefficient by default.
Parlay structures must leave enough time for each leg to settle independently. Don't stack 3 props that all settle in the same 15-minute window — the structure collapses when you can't react.
The framework, the platforms, the bankroll, the doctrine library, and the AI partner all operate as one stack. You don't switch between them — they're one operating system.
The first ~50 positions are data, not income. Size for survival, capture every result with doctrine tags, refuse heater mentality. Bankroll discipline ~20-25% session deployment.
The companion bet to the conviction bet. If the conviction bet loses, the hedge recovers enough capital that you're not mad. Anti-tilt structural protection.
The line moves into evening as more sharps weigh in. Late deployment of evening positions captures the line at its most informed price.
Overnight market repricing between yesterday's close and today's open. Position selling/buying across that window captures real volatility, not gameday volatility.
Don't pre-deploy outer hedges until inner legs settle. Wait for the data point, then bracket the next position based on what actually happened, not what you predicted.
The textbook Layer 05 TWOFOR position. One stake. Two paydays available depending on how the game develops.
Auto-generated from intelligence/bets/state.json · the architecture-under-the-game doctrine validated in real-money production · per operator-honest case study page.
The Lab isn't a pick-selling service. It's a framework you learn by doing. A productized day where PJ walks you through bonus extraction across the platforms, runs you through the bracketed portfolio framework using your actual money, and you leave with the framework + the day's net positive cash.
Text PJ. Tell him you want The SideGuy Day brief. He'll send you the doctrine page, the bonus-platform list, and the first bracket structure to study before the day runs.