Can Sportsbooks Separate Bet Builder From the Trading Feed?
Same-game parlay and bet builder products have usually come bundled with a sportsbook's trading feed. Francesco Borgosano, CEO and co-founder of Huddle, explains what it takes to separate the two, using the company's SGPx engine as the example.
On 8 September 2026, Huddle launched SGPx, a same-game parlay engine that works from an operator's existing odds feed rather than replacing it. Huddle says SGPx prices combinations from another supplier's single-leg prices using its own simulation models, supports pre-match and in-play betting, and can be integrated in a matter of weeks.
Huddle is itself an odds feed supplier. The company in its current form dates from a 2022 merger of Huddle Gaming and DeckPrism Sports, after which Las Vegas Sands made a strategic investment. It has offered its own simulation-based same-game parlay product since 2023. Huddle describes SGPx as the same approach applied to odds it does not produce.
The vantage point is worth stating. A plug-in bet builder lets Huddle sell into sportsbooks whose trading feed comes from another supplier, and Huddle says its own additional content and pricing remain available separately as a premium service. The case for separating the two products comes from a supplier that benefits if they are separated.
Francesco Borgosano, CEO and co-founder of Huddle, explains how correlated prices are built from third-party singles, how SGPx separates impossible combinations from improbable ones, where acceptance, risk and settlement sit when two suppliers are involved, and what integration involves. He gave no figures for early deployments, so the performance claims below are Huddle's own.
Gaming Eminence: Bet builder has usually come bundled with the trading feed. Why are operators now looking to separate the two, and what has typically held them back?
Francesco Borgosano: "Historically, Same Game Parlays (SGPs) were developed as an extension of traditional parlays. The idea was to leverage the existing sportsbook infrastructure, allowing legacy platforms to handle SGPs through the familiar logic of a multi-leg betslip.
However, this architecture created an important dependency: since the individual legs had to be recognised and connected within the same betting ecosystem, the Bet Builder engine was tightly coupled with the underlying odds feed. As a result, operators typically had to source both products from the same supplier.
Today, operators are increasingly questioning this model for two main reasons.
Firstly, product differentiation. Bet Builder has evolved from a complementary feature into a significant driver of engagement, turnover, and margin. Operators want greater control over market coverage, pricing, and the ability to introduce new betting combinations without being constrained by their primary feed provider's roadmap.
Secondly, commercial flexibility. An operator might be perfectly satisfied with its existing odds feed but want to improve its Bet Builder offering. Historically, doing so could mean replacing the entire feed or accepting the limitations of the incumbent supplier.
The main obstacle to separating the two has been technical. A standalone Bet Builder must consume external odds, understand the relationships between selections, calculate correlated prices in real time, and manage the resulting bet throughout its lifecycle. Crucially, it must do so without requiring the sportsbook to change its existing trading infrastructure.
At Huddle, we've addressed this challenge with our standalone SGP solution, allowing operators to integrate a dedicated Bet Builder independently of their core feed provider.
The fundamental shift is from treating Bet Builder as an extension of the odds feed to treating it as an independent product capability. That opens the door to greater competition, faster innovation, and more flexibility for operators."
Gaming Eminence: Huddle's SGP was built on simulation from your own pricing models. When the single-leg prices come from another supplier, how does SGPx build reliable correlated prices from them?
Francesco Borgosano: "The key distinction is between pricing an individual outcome and understanding the relationship between multiple outcomes. Those are two different problems.
Our original SGP product was built around proprietary simulation models, which generate a coherent representation of the underlying sporting event. That allows us to evaluate both the individual probabilities and the joint probability of multiple outcomes, so the correlation.
With SGPx, we extend that same approach to third-party odds.
The external feed provides their prices for individual outcomes, which become the reference point for the individual selections. Our models then use that information to calibrate the underlying probability distributions while preserving the relationships between outcomes through our simulation framework.
For example, consider an NFL Bet Builder combining a quarterback's passing yards, a receiver's receiving yards, and the team's total points. The individual prices might come from another supplier, but our simulation models understand how those outcomes interact. We can therefore calculate a coherent combined probability that reflects both the external market prices and the underlying correlations.
There is also an important distinction between producing a mathematically valid price and producing a commercially reliable one. External odds can have different margins, update frequencies, and market conventions. Handling those differences, maintaining consistency as prices move, and applying the appropriate trading controls are all essential parts of the solution.
Ultimately, SGPx separates the source of the individual prices from the intelligence required to combine them. This allows our partners to leverage their existing odds and trading operations while benefiting from Huddle's specialised SGP pricing capabilities, without disrupting their core pricing."
Gaming Eminence: Not every combination can or should be priced. How does SGPx decide what to offer and what to reject, including what is genuinely impossible rather than just improbable?
Francesco Borgosano: "We approach this as three separate questions: Is the combination possible? Can we price it reliably? And are we comfortable offering it from a trading perspective?
The first distinction is particularly important. There is a fundamental difference between a combination that is unlikely and one that is mathematically impossible.
Take a simple NFL example: a customer wants to combine a quarterback throwing for over 300 yards with the same quarterback throwing for under 200 yards. Those outcomes are mutually exclusive, so the combination should never be offered.
But consider something less obvious: a quarterback throwing for over 350 yards while his team scores under 10 points. That might be extremely unlikely, but it is certainly possible. A well-designed Bet Builder needs to recognise that distinction rather than simply rejecting combinations because they exhibit strong negative correlation.
This is where our simulation-based architecture becomes particularly valuable. Our SGP engine runs significantly more simulations than our core odds feed, deliberately exploring a much broader range of potential outcomes through a Monte Carlo approach. This allows us to capture rare scenarios, improve the statistical reliability of joint probabilities, and distinguish unlikely combinations from those that are genuinely incompatible.
However, mathematical feasibility alone isn't sufficient. Some combinations may technically be possible but fall outside the boundaries where our models can produce reliable prices. In those cases, we'd rather reject the combination than offer a price we cannot confidently support.
Finally, there is the trading dimension. Even when a combination is both valid and accurately priced, operators may want to restrict it based on exposure, liability, or specific risk policies. The objective is to give customers the flexibility to build meaningful bets while ensuring operators retain control over pricing quality and risk."
Gaming Eminence: With one supplier pricing the singles and SGPx pricing the combination, how is risk managed across both systems, particularly in-play, and who has the final say on accepting and settling a bet?
Francesco Borgosano: "The key to making this architecture work is establishing clear responsibilities between the different components. Having two pricing suppliers should not mean having two independent systems making conflicting trading decisions.
As an odds feed provider, Huddle is not directly involved in the bet acceptance process. Our responsibility is to deliver accurate, reliable, and timely prices, while the ultimate decision to accept or reject a wager remains with the sportsbook operator and its risk management infrastructure.
This is particularly important in-play, where prices and market availability can change rapidly. SGPx must remain synchronised with the underlying odds feed, recalibrating its models as prices change and ensuring that combinations reflect the latest available market information.
Similarly, resulting and settlement are driven by the official event data and settlement rules powering the operator's core feed, rather than by Huddle. This ensures consistency between individual selections and the resulting Bet Builder, without introducing a separate source of truth.
Ultimately, SGPx is a specialised pricing engine, not a replacement for the operator's existing trading and risk management infrastructure. Our objective is to deliver the best possible correlated pricing while allowing operators to retain full control over bet acceptance, exposure management, and settlement."
Gaming Eminence: What does integration actually involve for an operator with an established trading supplier, and where do the real complications tend to sit?
Francesco Borgosano: "From a technical standpoint, integrating SGPx is relatively straightforward and involves two main integration points.
The first is inbound: the operator provides Huddle with its existing core odds feed, including events, markets, selections, and prices. This allows our models to consume the underlying market information and calibrate the simulations accordingly.
The second is outbound: the operator integrates with Huddle's asynchronous pricing API, allowing the sportsbook to submit Bet Builder combinations and receive correlated prices calculated by SGPx.
The real complexity tends to sit in the data layer, rather than the API integration itself.
Different odds suppliers have different ways of representing events, markets, selections, and betting lines. Before we can price a combination, we need to ensure that the information coming from the external feed is correctly mapped to the underlying outcomes recognised by our models.
For example, one supplier might represent a player's passing yards through a series of alternative lines, while another uses different market identifiers or conventions. These differences may seem trivial, but consistency is essential when calculating correlations between multiple outcomes.
This is where much of our integration effort is concentrated: normalising different data models into a consistent framework that our simulation engine can interpret.
Ultimately, the objective is to make SGPx an additional pricing capability rather than another infrastructure project."
Gaming Eminence: Where is SGPx live today, and what are early deployments showing? We'd be particularly interested in any running alongside a third-party trading feed. Anonymised or directional figures are fine.
Francesco Borgosano: "We deployed SGPx during the current sports season, so it's still too early to draw meaningful conclusions about its long-term commercial performance.
That said, the initial feedback has been very encouraging. Early adopters have particularly appreciated how seamlessly SGPx integrates into their existing sportsbook infrastructure, allowing them to introduce a new Bet Builder capability without disrupting their core trading operations or existing supplier relationships.
We're also seeing positive early momentum in terms of player engagement and adoption, which gives us confidence in the product's potential.
Ultimately, the most important validation at this stage is that we've demonstrated the ability to deliver a standalone SGP experience alongside an existing trading feed, without the operational complexity that has historically made this difficult."
Gaming Eminence: Looking ahead, which parts of the sportsbook stack do you expect operators to unbundle next, and which will stay with a single supplier?
Francesco Borgosano: "The B2B sportsbook model is giving way to a modular, best-in-class architecture. We're already seeing the unpacking of odds feeds, so that specialist suppliers can deliver the best performance on high-margin products such as player props and micro markets. With their recent boom in popularity, the pricing gaps have also become more apparent, with expectations of double-digit margins, rather than just next-gen content added.
I'd also expect cashout to continue as a priority retention tool, with sportsbooks ensuring their calculation remains separate to sluggish trading feeds, ensuring markets remain open during live play for longer. What will stay are those core markets that are tied to high data rights and league licensing. These tend to be the markets with little differentiation in their pricing or front-end positioning, unlike player props and SGPs."
