Contract Reference

Core contracts deployed on Monad.

Contract Overview

Contract	Description
`PredictionMarket.sol`	Core SKC logic — create markets, predict, resolve, claim
`MarketFactory.sol`	Factory for deploying isolated PredictionMarket instances
`FixedPointMath.sol`	Library for on-chain ln() and cross-entropy scoring

Market Parameters

Parameter	Default	Description
Alpha (α)	20%	Stop probability per prediction
K	2	Last k predictors get flat reward
Flat Reward (R)	0.01 ETH	Reward per last-k predictor
Bond	0.1 ETH	Deposit per prediction
Liquidity (b)	1.0 ETH	LMSR scaling parameter
Initial Price	0.5	Starting market price

Write Functions

`createMarket`

function createMarket(
    string calldata question,
    uint256 alpha,
    uint256 k,
    uint256 flatReward,
    uint256 bondAmount,
    uint256 liquidityParam,
    uint256 initialPrice
) external payable returns (uint256 marketId)

`predict`

function predict(uint256 marketId, uint256 probability) external payable

Submit a prediction with bond attached. One prediction per wallet per market.

`claimPayout`

function claimPayout(uint256 marketId) external

`forceResolve`

function forceResolve(uint256 marketId) external

Force-resolve a stale market. Owner can call anytime, anyone can call after 2 days of inactivity.

Read Functions

Function	Returns	Description
`getMarketCount()`	uint256	Total markets created
`getMarketInfo(id)`	tuple	Core market data (question, price, status)
`getMarketParams(id)`	tuple	Market configuration (alpha, k, bond, etc.)
`getPrediction(id, idx)`	tuple	Specific prediction details
`getPayoutAmount(id, addr)`	uint256	Net payout amount (post-fee)
`isMarketActive(id)`	bool	Whether market is accepting predictions
`hasPredicted(id, addr)`	bool	Whether address has already predicted

Events

event MarketCreated(uint256 indexed marketId, string question, ...)
event PredictionMade(uint256 indexed marketId, address indexed predictor, ...)
event MarketResolved(uint256 indexed marketId, uint256 finalPrice, ...)
event PayoutClaimed(uint256 indexed marketId, address indexed predictor, ...)

Alpha Tuning Guide

Alpha	Avg Predictions	Best For
10%	~10	Deep analysis, many participants
20%	~5	Balanced (default)
33%	~3	Quick resolution
50%	~2	Very fast, binary questions