Methodology & FAQ

1. The consensus tier list

We track every published League Classic ranking and normalise them onto one scale: each tier is mapped to the midpoint of its band within that ranking's own scale (a 6-tier S+…D scale and a 4-tier S…C scale don't mean the same thing by letter). A champion's consensus score is the average across all rankings that cover it for that role.

Split marks champions where rankings disagree by 30+ points — read those entries as "contested, form your own opinion". Champions covered by a single ranking are capped at A tier and labelled, because one opinion isn't a consensus. Full attribution is in the footer.

2. The damage simulator

The consensus above is other people's opinions, averaged. The damage numbers are not an opinion at all: they come from a Season 3 combat model that reads champion base stats, spell base damage and ratios, item stats and mastery effects out of the patch 3.13.1 data files and computes the result. Nothing in it is scraped from another site, and nothing in it is a win rate.

Two things that are easy to get wrong, and that this model handles explicitly. Season 3 stat growth is linear — stat = base + perlevel × (level − 1) — not the quadratic curve introduced in 2018, which would inflate every level-18 number. And armour reduction and armour penetration are different operations applied in a fixed order: flat reduction, then percent reduction, then percent penetration, then flat penetration. Penetration cannot push a resistance below zero; reduction can.

The reference measurement is a full spell rotation plus auto-attacks at level 18 against a target with 1900 HP, 75 armour and 50 magic resist. Every build on the site is measured the same way, against the same target, so the numbers are comparable to each other.

Where it stops. The model knows damage. It does not know range, crowd control, mobility, mana or cooldowns, and it cannot tell you that a 150-second ultimate is worth less than a 4-second Q. That is why a measured letter is labelled a damage rank, not a power rank everywhere it appears, and why every off-meta build also carries a written note saying when it is simply the wrong pick — the simulator cannot produce that line, so a human wrote it.

Where the model can't read a number out of the data unambiguously, it says so on the build instead of guessing: those spells are flagged rather than silently counted as zero. You can run any of it yourself on the damage calculator, which uses the same maths.

FAQ