Sports markets have traditionally asked straightforward questions.
Who wins?
How many points will be scored?
Will a team cover the spread?
A new generation of prediction-market contracts is beginning to ask something more complicated: what path will a team take to victory?
On September 21, 2026, ARST certified two binary event contracts with the Commodity Futures Trading Commission tied to the Green Bay Packers’ September 24 game against the Atlanta Falcons.
One asks whether Green Bay will defeat Atlanta without its win probability falling below 20%.
The other asks whether the Packers will win without their point differential ever being above 6.5 points.
Those conditions fundamentally change what determines whether the contract settles Yes or No.
Winning the football game is necessary.
Winning alone is no longer sufficient.
The entire sequence of events during the game can determine settlement.
That makes these contracts an important development in the continuing convergence of prediction markets, live sports data and conventional sports wagering.
Green Bay-Atlanta Introduces Two Unusual NFL Contracts
The CFTC’s designated contract market product filings show ARST certifying the Green Bay-Atlanta contracts on September 21.
Both are listed as binary event contracts.
The first asks whether the Packers will defeat the Falcons without Green Bay’s win probability falling below 20%.
The second asks whether Green Bay will win without its point differential being above 6.5.
The NFL lists Atlanta at Green Bay for Thursday, September 24 at 8:15 p.m. ET.
Green Bay enters the matchup at 1-1 after losing 39-22 to Minnesota in Week 1 and defeating the New York Jets 20-17 in Week 2.
Atlanta enters after a difficult start that included a 34-3 loss to Carolina on September 20.
Those football results provide the setting, but the larger story is the structure of the contracts themselves.
Winning Is Only the First Requirement
Consider the win-probability contract.
A conventional game-winner market effectively asks one question:
Did Green Bay win?
The new structure requires at least two conditions.
Green Bay must win the game, and its designated win probability must never fall below the specified threshold.
Imagine Green Bay falling behind significantly in the second quarter.
Its live win probability drops to 18%.
The Packers then recover and eventually win.
Under an ordinary winner contract, Green Bay’s victory determines the outcome.
Under the path-dependent contract, the earlier probability decline could determine settlement even though Green Bay ultimately wins.
That transforms the market from a prediction about a final result into a prediction about the trajectory of the game.
The Point-Differential Contract Works in the Opposite Direction
The second contract is equally unusual.
It asks whether Green Bay wins without its point differential ever being above 6.5.
That means the Packers cannot simply win.
Their path to victory must also remain within the contract’s specified scoring condition.
For example, imagine Green Bay leads by 10 points during the third quarter but Atlanta rallies and the Packers ultimately win by three.
A normal moneyline-style contract would care only that Green Bay finished ahead.
A final-margin market would care about the closing score.
This contract potentially cares about what happened earlier.
The temporary 10-point advantage could matter even though it disappears before the final whistle.
That is an important conceptual shift.
Prediction Markets Are Moving From Outcomes to Paths
Sports markets have always been able to create increasingly specific propositions.
But these contracts introduce a different dimension.
Instead of asking only what occurs, they can ask how it occurs.
That creates what can be understood as a path-dependent sports contract.
The settlement outcome depends on the sequence followed by the underlying event.
A football game can therefore generate two identical final scores but different contract outcomes.
Suppose Green Bay wins 24-21 in two hypothetical games.
In Game A, Green Bay leads narrowly throughout.
In Game B, Green Bay falls behind 17-0, its live win probability collapses below 20%, and it produces a dramatic fourth-quarter comeback.
The final score could be identical.
The path-dependent contract could settle differently.
That Requires Much More Data Than a Moneyline
A conventional winner market needs a reliable final result.
These new products require considerably more information.
A win-probability threshold contract potentially needs a continuous or sufficiently granular probability feed throughout the game.
A point-differential path contract requires an accurate record of every scoring state.
That increases the importance of data infrastructure.
It also raises a basic settlement question:
What is the authoritative source?
A final NFL score is objective.
A live win probability is a calculated metric.
Different statistical models can produce different probabilities at the same moment.
One model might estimate Green Bay at 21%.
Another might calculate 19%.
That difference is irrelevant to the football result but potentially decisive for a contract with a 20% threshold.
Win Probability Is a Model, Not a Score
This distinction deserves particular attention.
A team’s score is directly observable.
Win probability is estimated.
A model can incorporate variables such as score differential, possession, field position, time remaining, timeouts and historical game situations.
Different models can weight those variables differently.
Suppose Green Bay trails by seven points with six minutes remaining.
One model could estimate a 24% chance of winning.
Another might calculate 19%.
Neither number is the official score.
They are statistical estimates derived from the game state.
For a contract tied to whether probability ever falls below 20%, the methodology behind the probability calculation becomes part of the product itself.
That makes settlement specifications unusually important.
The Data Feed Becomes Part of the Contract
For this type of market, data is no longer merely something traders use to make decisions.
It can become part of what determines settlement.
That is a major distinction.
In conventional sports markets, official league results usually establish whether a wager wins or loses.
For path-dependent prediction contracts, a specified third-party calculation or data methodology may also become essential.
Participants therefore need to understand not only the sporting condition but the contract’s formal rules and designated settlement source.
A visually simple question such as “Did the team’s probability stay above 20%?” can hide substantial technical complexity underneath.
That complexity becomes particularly important around the threshold.
A probability moving from 20.1% to 19.9% may look insignificant statistically.
For a binary contract, it can be decisive.
This Is Different From Ordinary Live Betting
Live sportsbooks already change prices continuously during games.
If Atlanta scores a touchdown, Green Bay’s live moneyline can move.
If Green Bay responds, the market can move back.
But those price changes usually describe the market at that moment.
They do not normally become permanent historical conditions attached to a later winner market.
Path-dependent contracts change that.
A moment occurring early in the game can permanently determine whether a contract remains capable of settling Yes.
That creates a kind of market memory.
Once the specified threshold has been crossed, what happens later may not reverse its effect.
The Contract Remembers What Happened Earlier
This is perhaps the clearest way to understand the difference.
Traditional winner markets largely forget the path.
A team can play terribly for three quarters and still win.
The final result is what matters.
A path-dependent contract remembers.
It can remember that the team once fell below a probability threshold.
It can remember that the scoring margin once exceeded a specified level.
That means two contracts centered on the same team can behave very differently during the same game.
A standard Packers winner contract could recover dramatically during a comeback.
A contract requiring Green Bay never to fall below 20% could already be determined by an earlier game state.
Similar Products Are Appearing Outside the NFL
The Packers-Falcons filings are not isolated.
ARST also certified comparable contracts tied to the Los Angeles Dodgers-San Diego Padres MLB game scheduled for September 23.
One asks whether Los Angeles wins without its win probability falling below 20%.
Another asks whether the Dodgers win without their point differential being above 2.5 runs.
The CFTC database also lists comparable NBA contracts involving the New York Knicks and Philadelphia 76ers for their October 20 matchup.
One asks whether New York wins without its point differential exceeding 9.5 points.
Another applies a 20% win-probability floor.
The same underlying concept is therefore being adapted across football, baseball and basketball.
Different Sports Create Different Path Dynamics
The concept may be portable, but each sport behaves differently.
NFL games have relatively few scoring events.
One touchdown can dramatically change the game state.
Basketball contains constant scoring and much more frequent changes in point differential.
Baseball can remain scoreless for several innings before a single sequence changes the game.
That means the same path-dependent contract design can behave very differently depending on the sport.
A 20% probability threshold in football may be crossed only after a significant event.
Basketball probabilities can fluctuate repeatedly during runs.
Baseball probabilities can shift sharply after a home run, pitching change or bases-loaded situation.
Contract designers therefore need to account for the statistical characteristics of the underlying sport.
These Markets Create a New Kind of Pricing Problem
Pricing a simple game-winner contract primarily requires estimating the probability that one team defeats another.
A path-dependent contract requires more.
The market needs to estimate both the probability of victory and the probability of satisfying the additional condition along the way.
Those probabilities are related but not identical.
A strong favorite may have a high probability of winning.
But that does not automatically mean it has an equally high probability of winning while never experiencing a specified adverse game state.
Similarly, a team expected to win comfortably may have a very different probability of winning while never leading by more than 6.5 points.
The additional condition changes the event being priced.
Traditional Odds Provide Useful Context
Understanding this distinction starts with ordinary implied probability.
Sportsbook prices translate expectations about outcomes into odds while incorporating operator margin.
Our guide to how sportsbooks price favorites and underdogs explains the basic relationship between market expectations, probability and price.
Path-dependent contracts add another variable.
Instead of estimating:
Probability that Green Bay wins
the market might need to estimate:
Probability that Green Bay wins AND never crosses condition X.
That intersection can produce a meaningfully different price.
The 20% Threshold Raises an Important Settlement Question
Of the two Green Bay contracts, the probability-based product is technically more interesting.
A scoring margin can be reconstructed directly from the official play-by-play.
Win probability requires a model.
That makes transparency critical.
Participants need to know which probability source governs settlement, how frequently the calculation updates, what happens if data are corrected and how edge cases are resolved.
Consider a play that is initially recorded as a touchdown but overturned after review.
A live probability model could change sharply when the touchdown is recorded and then reverse after the ruling.
Would the temporary probability reading count?
The answer depends on the contract rules and settlement methodology.
Those details become increasingly important as contracts depend on rapidly changing live data.
Latency Becomes More Important Too
Path-dependent markets also increase the importance of timing.
Sports data move through several stages.
An event happens on the field.
The event is recorded.
A data provider distributes it.
Models process the new information.
Trading platforms update.
Participants react.
Even small differences in that sequence can matter for live markets.
The problem is not unique to prediction markets.
Sportsbooks have dealt with live-data latency for years.
But contracts tied directly to an evolving probability metric make data timing even more central because the metric itself may determine settlement.
Liquidity Could Determine Whether These Products Work
A sophisticated contract is not automatically a useful market.
It still needs buyers and sellers.
Traditional NFL markets concentrate enormous attention around a relatively small number of familiar products such as moneylines, spreads and totals.
Every additional contract divides potential trading activity.
That can create liquidity fragmentation.
Imagine one game supporting separate contracts for:
winner, point differential, total points, player outcomes, scoring events, win-probability thresholds and path-dependent margins.
If participation is distributed across too many products, some contracts may develop wide spreads or shallow order books.
The innovation therefore creates a trade-off.
More precise markets can provide more expressive ways to price an event.
Too many markets can make individual contracts less efficient.
Complex Contracts Can Be Harder to Understand
Consumer comprehension is another issue.
“Will Green Bay win?” is immediately understandable.
“Will Green Bay win without its win probability falling below 20%?” requires several additional concepts.
Which win probability?
Measured when?
Updated how often?
What happens after a correction?
Does touching exactly 20% qualify?
What source settles the market?
Those questions are not minor technicalities.
They define the product.
As prediction markets become more complex, clear contract specifications become increasingly important.
A user should be able to understand exactly what must happen before entering a position.
Certification Does Not Mean the CFTC Endorses an Outcome
The regulatory terminology also deserves clarification.
The CFTC database lists these ARST products as certified binary event contracts.
That should not be interpreted as a government prediction that the contract is accurately priced or that a particular side represents good value.
Nor is certification equivalent to the CFTC recommending participation.
It refers to the regulatory filing status of the product.
The distinction matters because prediction-market products can carry the visual appearance of financial instruments while still involving uncertain sporting outcomes.
The Concept Already Predates the Atlanta Game
The September 21 filings are a fresh implementation, but ARST had already certified the generic NFL structure earlier.
CFTC records show an NFL contract template certified on August 14, 2026 asking whether a team would win a game without its point differential being above a specified number.
The database also contains a Green Bay-specific version tied to the Packers’ September 13 game against the Minnesota Vikings.
That indicates this is not simply a one-game experiment.
It appears to be part of a broader effort to develop reusable path-dependent sports contracts.
Sports Prediction Markets Are Becoming More Granular
The development fits a much larger trend.
Prediction exchanges have expanded from basic championship and game-winner contracts into increasingly detailed sports outcomes.
Markets now cover player statistics, individual periods of games, tournament qualification, awards and other specific events.
Path-dependent contracts push that evolution further.
They allow markets to encode not only an outcome but a condition attached to the journey toward that outcome.
Mathematically, that produces a much larger universe of possible contracts.
Commercially, however, the challenge will be deciding which of those markets attract enough participation to justify their complexity.
This Could Eventually Extend Far Beyond Football
The same structure can theoretically apply to soccer.
A market could depend on a club winning without ever falling behind.
Another could depend on winning while never exceeding a specified goal differential.
Basketball offers even more possible paths because leads change frequently.
Baseball can create conditions around run differentials and changing win probabilities.
Tennis could produce similar structures around match probabilities or set trajectories.
That does not mean every possible contract will become commercially viable.
But the CFTC filings demonstrate that exchanges are already experimenting with the underlying architecture across multiple American sports.
The Final Score Is No Longer Enough
For decades, the fundamental unit of sports wagering has been the result.
Prediction markets are beginning to break that unit into something more complicated.
The Green Bay-Atlanta contracts certified on September 21 illustrate the shift clearly.
A Packers victory may not be enough.
The market can ask whether Green Bay won without ever becoming too unlikely to win.
It can separately ask whether Green Bay won without ever building too large a lead.
Those conditions turn the timeline of the game into part of the financial product.
That creates interesting possibilities for market design, but it also increases the importance of transparent settlement rules, authoritative data, model methodology, liquidity and consumer understanding.
The next phase of sports prediction markets may therefore be less about creating another way to ask who wins.
It may be about creating increasingly precise ways to trade how the winner gets there.