Anchored Horizon is Iridae’s probabilistic commercial forecasting system. It estimates how revenue outcomes change across different game concepts, configurations, launch assumptions, and strategic paths.
Most revenue forecasting in games is still built around static comp decks: pick a few titles, cite a rough range, and call it a plan. That is not enough for decision systems that need to compare multiple futures and reason about uncertainty in a structured way. Anchored Horizon is built for that harder job.
What Anchored Horizon Is
Anchored Horizon is not just a point forecast for a single game. It is the commercial outcome model that helps Iridae evaluate how different choices reshape the space of plausible revenue futures.
It starts by retrieving comparable games algorithmically using Iridae’s latent market representation, then turns those comps into structured prior evidence. From there, it adjusts based on the actual features, positioning, and uncertainty profile of the concept or scenario being evaluated.
That allows the system to answer questions like:
- How commercial expectations change if the concept evolves in one direction versus another
- How pricing, launch timing, or scope decisions affect the outcome distribution
- How much upside and downside attach to one strategic branch compared with another
In other words, Anchored Horizon does not just estimate revenue. It helps the platform reason about commercial consequences under different plausible futures.
How It Works in Practice
Anchored Horizon combines market-neighbor retrieval, comp-informed priors, and feature-sensitive probabilistic modeling into one forecasting layer.
Relevant comps are identified algorithmically rather than hand-selected. Their evidence is weighted by similarity, freshness, and confidence in the underlying revenue estimate. That gives the model a grounded commercial baseline without pretending that every comp is equally informative.
From there, the forecast moves beyond the comp set. The specific structure of the title or scenario can pull the prediction away from its neighbors when there is real signal to do so. Rare breakout outcomes remain possible, but they are handled explicitly rather than allowed to distort the center of the forecast.
The result is a posterior predictive distribution that can be queried by other systems: not just “expected revenue,” but the shape of the range, the downside, the upside, and the probability of hitting meaningful thresholds.
What It Enables
Because Anchored Horizon is scenario-native, it supports more than standalone revenue estimation.
It can be used to compare commercial implications across:
- Alternative product directions
- Pricing and packaging decisions
- Launch timing assumptions
- Concept refinements
- Broader strategic pathways being evaluated elsewhere in the stack
That makes it especially valuable when revenue is not the only question, but one of several constraints shaping a larger decision.
How It Powers the Iridae Stack
Anchored Horizon is one of the core systems that makes Iridae’s planning layer commercially grounded.
It uses Latent Spark to understand what a game or concept is, how it sits in the market, and which comps actually matter. Its probabilistic forecasts feed directly into Patient Cartographer, where projected commercial outcomes become part of larger strategic pathway evaluation. It can also absorb information from experimentation and concept refinement elsewhere in the stack as uncertainty narrows.
That means commercial forecasting is not isolated from representation or planning. It becomes part of the same uncertainty-aware reasoning layer that helps studios compare paths, not just endpoints.
Why It Matters
Studios rarely struggle because they have no revenue estimate at all. They struggle because commercial assumptions are often too static, too hand-waved, or too detached from the decisions that actually change outcomes.
Anchored Horizon closes that gap. It gives the Iridae stack a way to reason about revenue as something shaped by strategy, design, timing, and uncertainty, not just something to summarize after the fact.
That is what makes it more than a forecasting tool. It is the commercial consequence model that helps the rest of the platform compare futures in a way studios can actually plan around.