Should I Dodge's win estimate is one input, not a full read on a lobby. These guides cover the reasoning around it: what the underlying win-rate data actually means, what to look for in champion select yourself, and how to think about dodging, drafting, and counterpicking when the numbers alone don't settle it.
- When to Dodge — what a ranked dodge costs under the 2026 penalty system, and when that cost is worth paying.
- How to Read Champion Select — role coverage, damage-type balance, frontline, engage, and the behavioural signals worth watching for.
- What a Champion Win Rate Actually Means — what an aggregate win rate measures, and what it hides.
- Aggregate, Matchup, and Personal Win Rates — three numbers people call "win rate," and which one to trust.
- Why the Same Win Rate Means Different Things at Different Tiers — why Challenger readings are noisier than Diamond ones, and what the model does about it.
- How to Read a Losing Draft — comp archetypes, win conditions, and telling "no win condition" from merely awkward.
- Counterpicking Without Griefing Yourself — why a counter you can't play is worse than a neutral pick you can.
- Draft Order, Blind Picks, and Flex Picks — what pick order costs and buys, and why last pick isn't automatically best.
- The Noise Floor Under Every League Tier List — this app's own per-tier calibration data, and how to tell a real win-rate gap from one that's just noise.
- What a 55% Lobby Actually Means at Your Rank — the full band-boundary table by tier, and why the same reading earns a different verdict at Challenger than at Diamond.
- Why the Highest-Elo Data Is the Least Trustworthy — the sample-size mechanism that makes Challenger's win rates the noisiest on the ladder, not the most authoritative.
- How Many Picks Before a Champ-Select Read Means Anything — why an early, partial reading needs to move further from 50% to mean the same thing a full lobby does.
- Why a 43% Champion Isn't Scored at 43% — the log-odds math behind the win estimate, and why a single champion's own number can never move it alone.
- How the App Guesses Enemy Roles When Champ Select Never Says — the minimum-cost matching behind role assignment, and the honest failure mode when the underlying data is empty.
- Why a Champ-Select Screenshot Gets Misread — the upscale, inversion, and double-threshold pipeline behind local OCR, and what actually helps if a read comes back wrong.
- Ban Rate Is Not a Power Ranking — why ban rate is identical across every role row a champion shows, and what it measures instead of strength.
- The Expected Value of a Dodge, With the Actual Numbers — a break-even win probability for dodging, and how often a lobby actually bad enough to cross it occurs at each tier.