Beat a stronger rival
1200vs1400
With a 32 K-factor, a 1200-rated player is expected to score about 0.24 against a 1400-rated rival. A win would add about 24 points; a loss would subtract about 8.
COMPETITIVE RULEBOOK
Every point should
have a reason.
A transparent explanation of how TradeDuel intends to calculate match performance, adjust ELO ratings, and make duels fairer across different wallet sizes. Ratings change only after a match's settlement evidence has been verified and finalized.
A rating change is applied only when the match result and required portfolio evidence pass server-side checks. Incomplete evidence stays pending; it must not produce an unverified ELO change.
01 / RATING
ELO, explained simply.
ELO estimates relative skill from results against other rated players. Expected score is calculated from the rating gap; beating a higher-rated opponent earns more points, while losing to a lower-rated opponent costs more.
STANDARD ELO MODEL · ILLUSTRATIVE
New rating = Old rating + K × (Actual − Expected)
Actual result
Win = 1 · Draw = 0.5 · Loss = 0
Expected result
Estimated from both players' ratings before the match
K-factor
Controls how quickly ratings move; final value not set
02 / EXAMPLE
How a rating can move.
Illustrative example only, using a K-factor of 32. The final K-factor and rating parameters remain subject to validation.
Lose to a lower-rated rival
1400vs1200
Against a 1200-rated rival, a 1400-rated player is expected to score about 0.76. A win would add about 8 points; a loss would subtract about 24.
ELO is updated only after a match is complete, the required evidence is verified, and integrity checks pass. Matches with incomplete or conflicting evidence remain pending.
03 / PERFORMANCE
Win on performance, not wallet size.
Raw dollar profit would favour larger starting balances. A fairer comparison should use normalized return and controlled exposure, with both players measured under identical rules.
Normalize return
Compare percentage return on eligible starting equity or a fixed standardized allocation, not just raw SOL or USD profit.
Cap exposure
Apply the same maximum position size and concentration limits to both sides. The exact cap needs testing and publication.
Count the real costs
Account for trading fees, priority fees, slippage, transfers, and price-source reliability under one shared method.
Prevent rule gaming
Define eligible trades, start/end snapshots, open positions at timeout, token liquidity rules, and disqualification cases in advance.
04 / MATCH LENGTHS
Choose the clock.
Available duration options. Longer matches need robust price history and consistent start/end valuation.
05 / GAME MODES
Four modes.
One standard of fairness.
Mode rules must be enforced by the trade engine—not just shown in a dropdown. The four modes below can be selected in matchmaking, but their trading rules are not currently enforced.
01RULESET IN DESIGN
Standard
Baseline format
Both players use the same match duration and shared scoring rules. This is the default head-to-head mode.
02TRADE ENGINE REQUIRED
One Position
Execution discipline
Each player may hold only one open eligible position at a time. A new position starts only after the prior position is fully closed and indexed.
03CAP VALUE TBD
Exposure Cap
Limit concentration
Each position is limited to the same maximum share of eligible starting equity for both players, preventing a single oversized position from dominating.
04SCORING DESIGN
Equal Start
Reduce wallet-size advantage
Both players are evaluated from the same standardized starting allocation and compared on normalized percentage return, not raw dollars earned.
THE STANDARD WE WANT TO MEET
No black-box rankings.
Before ELO goes live, the final K-factor, return formula, tie handling, rating floor, provisional-player rules, and dispute process will be published alongside a test suite.
Open matchmaking ↗