Adding no contest outcome was wrong decision, because it increases risks both for traders and liquidity providers (LPs):
- LPs can lose all their liquidity due to huge IL in 49/49/2 pool when no contest price changes from 0.02 to 1, for example when MMA fight is suddenly canceled
- LPs can lose up to 5-7% due to IL in markets where real no contest probability is close to 0, for example in NFL markets. Changing initial price from 0.02 to 0.01 or lower can help to decrease IL losses but also proportionally reduces availble liquidity (bad for traders)
- Traders who opens positions without buying no contest shares for insurance also can lose all their position. That’s not expected behaviour for most bettors because other prediction markets usually return money in case of tie or no contest
Without no contest outcome in case of tie or event cancellation markets should resolve 50/50 - all shares are redeemed for equal amount of USDC collateral. It drastically decreases risks described above for all market participants.