Parlay Dutching: Impossible & Unlikely Worlds

Parlay dutching is a proactive strategy where you place every bet before any games start, covering every possible outcome combination across a set of correlated legs. Each bet is a complete parlay placed at whichever single sportsbook offers the best price for that specific combination, and stakes are sized so the payout is identical no matter which outcome hits. The edge comes from one thing: impossible and unlikely worlds — outcome states that cannot happen (or rarely happen) but that sportsbook pricing still treats as if they can.

Parlay Hedging vs. Parlay Dutching

These are fundamentally different strategies. Understanding the distinction is critical.

How Sportsbooks Price Parlays: Two Different Engines

This is the foundation of the entire strategy — and where most people get confused. Sportsbooks use two completely different pricing engines depending on whether your parlay legs come from the same game or different games.

Impossible Worlds: The Core Concept

An impossible world is an outcome combination that literally cannot happen. Not unlikely — impossible. A logical contradiction. When you build a set of parlays covering all possible outcomes but some outcomes can't happen, you're paying for coverage you'll never need — and that overpayment is pure edge.

The Definitive Example: 3 Legs, 8 States, 2 Impossible

Consider a single NFL game — Chiefs (-7.5) vs. Raiders. Three legs from the same game:

• Leg A: Chiefs Moneyline (win / lose)
• Leg B: Chiefs -7.5 Spread (cover / miss)
• Leg C: Game Total Over/Under 47.5

Three binary legs = 2³ = 8 theoretical outcome states. But two of those states are logically impossible:

The SGP Tax: The Book's Countermeasure

Here is the critical nuance that separates theory from practice. Sportsbooks are not stupid. They know ML + Spread are correlated. Their SGP builders apply correlation adjustments specifically to handle this.

The SGP pricing pipeline works like this:

1. Start with individual leg markets — each leg already has embedded bookmaker margin
2. Apply same-game repricing — the builder adjusts the joint quote for correlation and market rules
3. Layer in any live promo mechanics — profit boosts, insurance, bonus bets, or site credit only matter if the offer terms allow them on that exact wager

So the real question is not "do impossible worlds exist?" — they do. The real question is:

Where The Edge Actually Lives: Three Paths

Given the SGP tax, there are three realistic paths to profitable parlay dutching. Each has different mechanics and different practical requirements.

Path 1: Cross-Game Correlated Parlays (No SGP Tax)

If you can find legs from different games that are correlated, you get the best of both worlds: impossible or unlikely worlds exist, but the book prices the parlay using naive multiplication (no SGP engine, no SGP tax).

Path 2: SGP Dutching When The Tax Under-Corrects

This is the primary path. You build SGP parlays at books where the SGP engine's correlation model does not fully account for the impossible worlds. This happens because:

• Builder formulas are proprietary — you only see the final quoted price, not the operator's internal model or margin split
• Different books handle the same relationship differently — some reject more combinations, some reprice more aggressively, and some expose deeper builder menus
• Menu depth varies across books — alt lines, player props, and niche legs can create wider quote dispersion because the builders do not offer identical choices
• Promo compatibility differs — DraftKings notes some SGPx builds can support bonus bets and No Sweat bets, while other books restrict boosts or insurance to narrower bet types and expiry windows

Path 3: Cross-Book SGP Arbitrage (The Hybrid)

This is the most realistic execution path. You don't need any single book to fully underprice — you need the best price at each book across all real states to collectively sum below 100% implied probability. Different books will misprice different combinations in different directions.

What a Profitable Opportunity Looks Like

The example above intentionally shows a losing setup because that's what most games produce. But here's what success looks like — same structure, better cross-book pricing:

Stake Sizing: Equal Payout Across All States

When you find an opportunity (sum < 1.0), you size stakes so every real outcome pays the same total amount. The formula is simple:

Unlikely Worlds: The Softer Edge

Beyond the truly impossible, there are unlikely worlds — combinations that cantechnically happen but occur far less often than pricing implies because of strong correlations the book's model underweights.

Unlikely worlds don't eliminate states — they overweightthem in the pricing. If the book's SGP model assumes a state has 8% probability but the true probability is 3%, you're effectively paying for 8 cents of coverage that only costs 3 cents in reality. This compounds with impossible worlds to further reduce the sum-of-reciprocals below 1.0.

Push and Void Handling

Pushes are the silent killer of parlay dutching. If one leg pushes (e.g., Chiefs win by exactly 7 on a -7 spread), the 3-leg parlay becomes a 2-leg parlay. Your carefully sized stakes are now wrong because the payout structure changes.

Alternate Lines: Expanding Compression

Alternate lines create additional impossible worlds beyond the standard ML+Spread structure. The wider the spread, the more impossible states you generate — but the odds on the remaining states change accordingly.

Sport-by-Sport Breakdown

Not all sports are equally suited for parlay dutching. The strategy requires binary outcomes, half-point lines, liquid SGP markets, and structural correlations. Here's how the major sports stack up:

Why 3 Legs Is the Sweet Spot

Two legs give you 4 states, and removing 2 impossible gets you to 2 real states — but that's just a straight arbitrage, not really a parlay dutch. Four legs give you 16 states — harder to manage, harder to find cross-book pricing, and the compounding vig on 4-leg parlays is brutal.

• 8 theoretical states — manageable number of bets
• 2 impossible + 1–2 unlikely = 4–6 real states in the best cases
• Cross-book shopping is feasible — most books offer 3-leg SGPs
• SGP tax is lower on 3 legs than 4+ — 17-30% vs 23-35%+ on 4 legs
• Each bet looks like a normal recreational parlay — no red flags for the sportsbook

Finding the Right Legs

The key is selecting legs where correlations create impossible or highly unlikely states. Not all 3-leg combinations compress — you need legs that logically depend on each other.

Screening Process: Finding Opportunities

Most games will not produce a profitable parlay dutch. Here's how to screen efficiently:

1. Start with large favorites— the bigger the spread, the wider the gap between ML and spread probabilities, and the more "probability mass" sits in the impossible states
2. Check SGP prices at 4+ books — price all 6 real states (ML+Spread+Total) at each book. Large variance between books signals at least one model is wrong
3. Compute the sum— Σ(1/best_decimal_odds) for all 6 states. If it's above 1.05, move on. If it's between 0.95 and 1.05, check for boost/promo overlays (see below). Below 0.95 is a strong opportunity
4. Verify half-point lines — ensure no push risk on any leg
5. Check bet limits — SGPs often have lower max bets than straight wagers. Verify you can place the required stakes at each book

SGP Boosts: The Edge Amplifier

This is where theory meets the real world of promotions. Sportsbooks frequently run SGP-specific offers, but cadence and eligibility vary by book, sport, jurisdiction, and customer segment. The practical edge comes from checking the live promo inventory and pricing the offer correctly, not from assuming a fixed weekly supply of boosts.

Worked Example: Boost Turns Loser Into Winner

Remember the unprofitable Chiefs example from earlier with a sum of 1.134? Here's what happens when you apply a 30% SGP profit boost token to State #3 (Chiefs ML + Miss + Over):

Combining with Bonus Clearing

If you're currently clearing a deposit match, site-credit rollover, or product-specific playthrough, these wagers only help if the exact terms count sportsbook parlays or SGPs toward the remaining requirement. That is not universal, so bonus clearing is an overlay you verify offer by offer:

Realistic Expectations

Is This Strategy New?

No. The underlying idea — pricing correlated combinations more carefully than the market does — long predates the recent SGP era. What is newer is the scale of same-game builders and promo overlays now available at regulated operators, which creates more surface area to screen but also more book-side controls and offer-specific restrictions.

What is relatively newer is the "impossible worlds" framing, which borrows from modal logic in philosophy to make the concept clearer. Most sharps just call it "correlated parlay arb" or "SGP arb." The math is the same regardless of what you call it.

Putting It All Together

The edge comes from compounding layers, in order of importance:

1. Impossible worlds— Logical impossibilities that eliminate entire outcome states. Only ML + Spread (same game) creates provably impossible states. This is the primary structural edge, and it's not statistical — it's logical. The question is whether the book's final builder quote fully prices this in.
2. Unlikely worlds — Correlated outcomes that reduce the true probability of certain states below what pricing assumes. Unlike impossible worlds, these still have real probability — they just have less than priced.
3. Cross-book price shopping — Best single-book SGP for each state further reduces total implied probability. This is where the execution edge compounds the structural edge.
4. SGP boosts and promos — A well-placed profit boost or bonus bet can push a marginally unprofitable setup below the 1.0 threshold. But only the offer terms tell you whether the reward applies, how it settles, and how much stake it really supports.
5. Bonus clearing overlay — When the exact promotion counts those wagers toward playthrough, even break-even dutches can generate profit from the remaining bonus value.