11-Level Smooth K-Factor Regulation & Mining Rules
While the decentralized "Trade-to-Earn" model can stimulate trading liquidity on the platform, it is susceptible to wash trading exploits. If a user buys 45 out of 49 numbers, their winning probability reaches 91.8%. They recover almost 100% of their betting principal while minting a massive amount of $DESIX tokens for free. This exploits honest Liquidity Providers (LPs) and triggers severe token inflation.
To preserve the protocol's integrity, DeSix implements an 11-level smooth K-factor regulation algorithm and an EOA identity detection mechanism within its mining engine.
1. 11-Level Smooth K-Factor Decay Formula
The hashrate generated by each bet is not simply equal to the USDC amount wagered; instead, it is inversely proportional to the risk exposure (number of selections) of the bet. Specifically, the contract enforces the following hashrate calculation on-chain via DeSixMiner.sol:
$$\text{Trade Hashrate} = \text{Bet Amount} \times K(n)$$
Where the regulation coefficient $K(n)$ is divided into 11 levels based on the number of selected numbers $n$ in a single bet:
| Level | Selections $n$ | Theoretical Win Rate | K-Value | Description |
|---|---|---|---|---|
| 1 | 1 number | 2.04% | 1.50x | Reward extreme risk prediction |
| 2 | 2 numbers | 4.08% | 1.40x | Reward ultra-high risk prediction |
| 3 | 3 numbers | 6.12% | 1.30x | Reward high risk prediction |
| 4 | 4-5 numbers | 8.16% ~ 10.20% | 1.20x | Medium-high risk |
| 5 | 6-8 numbers | 12.24% ~ 16.32% | 1.00x | Baseline hashrate |
| 6 | 9-12 numbers | 18.36% ~ 24.49% | 0.80x | Medium-low risk |
| 7 | 13-18 numbers | 26.53% ~ 36.73% | 0.60x | Low risk |
| 8 | 19-24 numbers | 38.78% ~ 48.98% | 0.35x | Extremely low risk |
| 9 | 25-30 numbers | 51.02% ~ 61.22% | 0.20x | Minor risk |
| 10 | 31-39 numbers | 63.27% ~ 79.59% | 0.10x | Virtually risk-free |
| 11 | 40+ numbers | 81.63% ~ 100.00% | 0.01x | Risk Control Block (Severe Penalty) |
2. Hashrate Allocation & Anti-Arbitrage Game
Through the 11-level decay curve, DeSix achieves a dynamic balance of game theory:
- Prohibitive Arbitrage Costs: If an arbitrageur attempts to wash tokens using 40 numbers, their mining hashrate is slashed by 99% ($0.01x$). Staking 10,000 USDC only yields 100 hashrate, whereas a single-number bettor taking real prediction risk generates the same hashrate with a mere 67 USDC bet. Stakers face a 0.3% protocol fee loss with virtually no token allocation, resulting in a negative expected return.
- High Risk, High Reward: Genuine players are incentivized to bet on fewer numbers (e.g., 1-3 numbers), which brings healthy variance to the platform and rewards them with a high mining multiplier of $1.3x \sim 1.5x$.
3. Anti-Sybil Identity Detection
Arbitrageurs might attempt another strategy: a Sybil Attack. By deploying scripts, they batch-create hundreds of smart contract wallets (AA accounts), each buying only 1 number within the same issue. Aggregated, they buy 40+ numbers, attempting to bypass the K-value single-address limit and claiming the high 1.5x hashrate.
To defend against this "phantom multi-wallet" attack, DeSix introduces the DeSixIdentity.sol resolution mechanism:
- Underlying EOA Extraction: When calculating hashrate,
DeSixMinerdoes not trust the smart contract wallet (AA account) address directly. Instead, via account abstraction tracking oftx.originor signed credentials, it trace back to determine the actual controlling EOA address behind the smart wallet. - Global Number Aggregation: Before calculating the K-value, the contract aggregates all numbers wagered across all different AA wallets controlled by the same EOA in the same issue, merging them into a single on-chain Bitmap.
- Cross-Wallet Resolution: If the aggregated numbers exceed the safety threshold, the system reapplies the K-value calculation using the total merged selection count. Even if split across 40 wallets, the chain identifies it as "40 numbers bought," dropping the K-value to $0.01x$ instantly and shattering Sybil feasibility.