Force

Submitted Feb 9, 2026

Application Details

Team & Background

Who are you?
Robotics @MIT media lab (MIT cheetah team) , SE@Microsoft, Built and launched 2 LSD protocols in 2020/2022 50-150mm TVL, 200mm-1.2b FDV), top 10 trader globally on GMX '24. Force's round was led by Reforge VC.
Previously Deployed
Yes
Previous Examples
unsheth and others

Market Proposal

Assets to Deploy
We plan to deploy NBA performance index perpetuals for players and teams (season-long, continuous perps). Each asset is backed by a deterministic, exogenous performance index constructed from real-time NBA game data. Index construction (high level): Underlying data source: live and official NBA game statistics (e.g., points, assists, rebounds, turnovers, efficiency, usage, pace, opponent context). Raw stats are normalized for pace, minutes, role, and matchup. Components are standardized against league baselines (z-score–style normalization). Each component is weighted and confidence-gated so early-game noise contributes weakly and ramps smoothly as sample size increases. Components are aggregated and lightly smoothed to produce a single performance score in the range [0, 100]. The index is fully deterministic, replayable, and does not ingest any market or price data. Reference rate / oracle mapping: Oracle price published to Hyperliquid is a simple linear mapping: Oracle Price = Performance Index × fixed dollar scaling Scaling is fixed per asset and versioned. Oracle updates at a steady cadence during games and via heartbeat between games. Asset behavior: Assets are season-long continuous perps (no per-game settlement). Individual games act as information inputs that move the oracle; positions may remain open across games. Hyperliquid derives the mark price from the oracle; traded price is discovered via the orderbook. These assets are designed to expose continuous, execution-driven volatility suitable for leveraged perpetual trading, while maintaining strict separation between oracle truth and market price discovery.
Market Type
Oracle priced perps on real-world performance indices
Exists on Other Venues
No
Differentiation
This doesnt exist on Hyperliquid or any other exchange yet.
Target Leverage
upto 10x
Market Makers
In discussion

Oracle & Distribution

Data Source
neither Pyth nor SEDA apply because this is not a price feed from another market; it’s a proprietary, exogenous index.
Price Construction
Underlying index construction (inputs and weighting) The oracle references a deterministic performance index constructed from real-time and official NBA game data. Inputs:
Raw inputs include player/team box score and play-by-play statistics such as scoring efficiency (points per attempt), assists, rebounds, turnovers, usage, on/off impact proxies, pace, and opponent-adjusted context. Inputs are sourced from a licensed sports data provider and reflect objective in-game actions only. Normalization:
All raw inputs are normalized for minutes played, possessions (pace), role/usage, and matchup context to ensure comparability across games and situations. Standardization:
Each normalized input is standardized against league-wide historical baselines (z-score–style normalization) so heterogeneous metrics can be combined on a common scale. Weighting:
Standardized components are combined using fixed, pre-defined weights that reflect their relative contribution to execution quality. Weights are set offline using historical analysis and are versioned and immutable during live games. No dynamic re-weighting or learning occurs intragame. Confidence gating: Each component’s contribution is confidence-gated based on sample size and relevance. Early-game observations contribute weakly and ramp smoothly to full weight as sufficient data accumulates, preventing early noise from dominating the index. Aggregation and output:
Weighted components are aggregated and lightly smoothed to produce a single performance index in the range [0, 100], representing execution quality at each moment. The index is fully deterministic, replayable from raw inputs, and does not ingest any market or price data. The oracle price published to Hyperliquid is a linear mapping of the index: Oracle Price = Performance Index × fixed dollar scaling The scaling factor is fixed and versioned per asset. Oracle updates are published at a steady cadence during games with heartbeat updates between games.
Volume Strategy
Volume is driven structurally through continuous, leveraged, event-driven perps. Each NBA game creates multiple intragame volatility events rather than a single terminal outcome, allowing capital to turn over repeatedly. We will launch with high-attention player and team indices, onboard sports-savvy perp traders and market makers, and ensure tight liquidity during predictable high-volatility windows.
Distribution Channel
No

Log Activity

Activity (0)

No activity yet