- Ranked needs 20 champions
- ~20 games to stop losing more
- Jungle punishes inexperience most
- Over 170 champions
Four champions. Twenty games each.
Over 170 champions, one ranked game at a time. The players who climb don't know more of them — they know a few of them far better.
- M · MainBlind-pickable, your first lock-in every game
- P · PartnerSame role, covers your main's bans and bad matchups
- S · SafetyOne simple pick for your secondary role and autofill
- Practice ringNew champions live here until they've earned ranked
Pool shape: BoostPirate's model. Curve: iTero, "Mastery: a statistical summary of 1M+ games" (24 Apr 2023, patch 13.7, ~Gold, NA/EUW/KR) — averages by mastery bucket under the pre-2024 mastery system.
Ask ten high-ranked players how many champions they play in ranked and most will give you a small number. Ask ten players stuck in the same division and you'll hear "I can play anything". Flexibility sounds like a strength, but in solo queue it's usually a cost: every champion you don't know well hands the enemy a small edge, and small edges decide close games. This guide shows what the data says about champion experience, gives you a simple pool shape that survives bans, autofill and patches, and explains how to pick, add and drop champions without throwing games while you learn.
- The first ~20 games are expensive: low-mastery picks lost far more often than the same players' practised champions.
- Then it flattens: 50,000 to a million mastery made no significant difference to win rate in the same data.
- Jungle punishes inexperience most, ADC the least — keep your jungle pool especially small.
- Switching during a losing streak hurts: top players who swapped champions and lanes on losing runs performed worse.
- Patches rarely kill a main: Riot's balance framework ignores mastery data and most changes are a few percent.
Why a small pool wins more games
Every champion has a learning cost: ability ranges, damage thresholds, power spikes, matchups, the combos you only find under pressure. That cost shows up in the numbers. iTero's Jack J. Williams looked at just over 1,000,000 ranked solo queue games on patch 13.7, split across NA, EUW and KR, mostly around Gold. About 20% of players were on a champion with under 10,000 mastery, and those games averaged about 44%. The other games averaged around 51.5%. The crossover to 50% came near 12,000 mastery, which at the time took about 20 games (roughly 600 mastery a game).
Past that point, the curve goes flat. Champions with 50,000 to a million mastery won 51.77% of games and champions above a million won 51.73% — a difference the author calls too small to be significant. A separate iTero study of pro play (2022) found the same shape: a veteran's 7th game on a champion was no worse than their 50th. In other words, you don't need hundreds of games on a champion — you need the first twenty out of the way.
The behaviour side points the same way. The 2024 CHI PLAY study of 597,680 top-player matches found players stayed on the same champion and lane during winning streaks and switched during losing streaks, and that consistency in champion and lane went with better performance. Switching feels like a fix when you're losing. The data says it's part of the problem. Our tilt guide covers the losing-streak side.
The 2 + 1 + 1 pool
This is the pool shape BoostPirate's boosters and coaches use for their own climbs. It's built around the three things that take a champion away from you in solo queue: bans, enemy picks and autofill.
The champion you'd lock in blind. Not the strongest on paper — the one with the fewest losing matchups that you play well.
Same role, different answer: covers the main's worst matchups and steps in when the main is banned or taken.
One simple, forgiving champion for your second role. It's your autofill plan, not a project.
Champions you're learning. They move into ranked once they've cleared their first twenty games.
Why two in the main role and not one? A one-trick is the purest version of the data above, and it works — until the champion is banned, picked by the enemy first, or countered so hard that the lane is lost at level 3. A partner who shares your role's job but plays differently covers all three without doubling your learning cost. Why not five? Because each extra champion you take into ranked spends games in that expensive under-twenty zone.
Choosing your main
Riot Support's advice on choosing a champion is worth taking seriously: look at the champion's playstyle, look and story, and not to buy a champion only for its current strength, because balance changes often. For a ranked main, add four practical tests:
- Blind-pickable. It can go first in champion select without being punished in most matchups.
- Does your role's job. A top laner who can't hold a side lane, or a support with no engage or peel, makes your team play around a hole.
- Survives patches. Champions whose power sits in one number (a single ratio, a single item) swing hardest when that number changes.
- You enjoy it after a loss. You'll play this champion on bad days. If it feels like work when you're tilted, it won't last.
- Readable for you. You know where its damage comes from and when it's strong — if you can't explain its power spikes, it's not a main yet.
If you're new to the game or a role, start with champions Riot rates as simpler — our easiest champions guide lists them by Riot's own difficulty ratings. Mastery takes care of the rest: Riot's champion mastery is uncapped since 2024, and the grades you earn (S+ down to D-) compare you with players in your region on the same champion and position, so they're a useful check on whether a champion has really clicked. Our mastery explainer has the details.
Choosing the partner
The partner's job is to cover the main's gaps. Riot sorts champions into six classes — assassin, fighter, mage, marksman, support and tank — and the easiest way to build a partner is to pick from a different class or range than your main. BoostPirate's pairing rules:
Matchup knowledge is the second half of the partner decision. The matchups guide shows how to learn which matchups your main loses and how to play them; the champions that keep showing up on that list tell you what the partner needs to beat.
Pools by role
Every role puts a different load on your pool. The mastery data and the 2026 role systems point to a few role-specific rules:
Long, isolated lanes punish unknown matchups. Keep two mains you can play into most of the class you meet, and a Teleport plan for each.
In iTero's data the jungle had the lowest win rate on low-mastery champions and the highest on high-mastery ones. Two champions, one clear each.
The most counter-picked lane. Your partner should be safe into the mid laners that beat your main.
The smallest mastery effect in the same data — marksmen share more patterns. You can afford a slightly wider pool here.
Have one engage support and one that peels. Support is a priority role that is often autofilled, so it's also the best secondary.
Role Quests make role consistency worth a little more in 2026: the quests follow the role you're assigned in champion select, so each role's reward — Top's Unleashed Teleport, the jungle's Primal Smite, Bot's extra item slot — only builds up in that role. Our Role Quests explainer covers them, and best role to climb helps if you're still choosing a main role.
Autofill and your safety pick
Riot's matchmaking asks for a primary and a secondary position and names Support and Jungle as priority roles that are often autofilled. The queue screen warns you: "Possible" means you may be autofilled, "Guaranteed" means you'll get one of your two choices, and "Carryover" means you dodged an autofilled lobby and will be autofilled next game. Your pool needs an answer for each case:
- 2ndSecondary role: your safety pick
One champion with a simple job that doesn't need matchup knowledge. Play it in normals until it's past twenty games.
- FillAutofilled to a third role
Pick the simplest champion you own for that role and play for your team's carries. This is not the game to learn something.
- DodgeDon't dodge to escape autofill
Since 26.1 a dodged autofill lobby carries the autofill to your next game — the role comes back, the LP doesn't.
The autofill guide explains autofill protection and the 2026 autofilled-versus-autofilled matching in full.
When a patch hits your main
Patches are the most common excuse for abandoning a pool. Most of the time they shouldn't be. Riot's published balance framework deliberately leaves mastery data out of buff and nerf decisions, so players aren't punished for investing in a champion, and most changes are a few percent. Patch 26.22 is a good example: Yasuo's Q went from 105% to 100% of total AD, a change a Yasuo main barely feels. Thresh took a bigger hit — 2 base AD and 50 damage off every rank of his ultimate — because Riot said he'd "remained at the very top of the support pecking order", and he's still the same champion with the same hook.
A few percent on one ability. Read it, adjust one habit, carry on — your experience is worth more than the nerf.
The thing you pick it for was removed or reworked. Play the partner for a patch while you relearn the main in normals.
Riot's framework doesn't hold new and reworked champions to its numbers for about a month — expect swings.
Our patch and meta guide shows how to read a patch in ten minutes and tell a real meta shift from noise.
How to add a champion without losing LP
New champions are fine — they just shouldn't learn in your ranked games. BoostPirate's routine for moving a champion from the practice ring into the pool:
- 01Practice Tool first
Combos, ranges and a full clear or a level-1 trade pattern. The Practice Tool unlocks at summoner level 9.
- 02Normals or Swiftplay
Play it into real opponents with nothing on the line. Notice the matchups that feel impossible.
- 03Twenty games
The iTero data's crossover sat around 20 games. Use it as your bar — and expect grades to rise as you get there.
- 04One ranked slot
Bring it in as the partner or safety, not the main. It earns the main slot with results.
- 05Drop someone
If the pool grows past four, the weakest champion goes back to normals. A pool you can't keep sharp isn't a pool.
One-trick or small pool?
- Maximum experience on one champion
- Every game teaches the same lessons
- Bans and first picks take it away
- A hard counter can lose you lane
- A big patch change hits everything you play
- Nearly all of the experience benefit
- An answer to bans and counters
- A plan for autofill
- Twice the matchups to learn in your role
- Needs discipline to stay small
Both climb. If you love one champion and it's rarely banned, one-tricking with a partner for emergencies is close to the same thing as our model. The version that doesn't climb is the "I can play anything" pool, where most ranked games are someone's first twenty.
What ranked asks of your collection
Riot Support lists three requirements for Ranked Solo/Duo and Flex: summoner level 30, 10 normal games and 20 champions owned. The free rotation of 14 champions a week (and a special rotation below level 11) is the cheap way to try champions before you buy them. Twenty champions is the entry fee, not the pool size — most of ours sit in the collection untouched. The ranked unlock guide and the Blue Essence guide cover getting there.
Common champion pool mistakes
- Keep two champions in your main role and one safety in your secondary.
- Learn new champions in normals until they're past twenty games.
- Pick a partner from a different class or range than your main.
- Keep your jungle pool especially small.
- Stay on your champions through a losing streak.
- Lock in a first-time champion in ranked because it's "strong this patch".
- Drop your main over a small nerf.
- Fill your pool with three champions that lose the same matchups.
- Dodge to escape autofill — it carries to the next game.
- Confuse owning 20 champions with playing 20.
- Riot Support — Choosing the right champion (19 Sep 2025: "over one hundred and seventy Champions", six classes, 14 free champions a week) · Unlocking game modes (29 May 2026: level 30, 20 champions, 10 normal games; Practice Tool at level 9) · Matchmaking and autofill (priority roles, queue indicators) · Champion mastery guide (grades)
- Riot Games — /dev: Champion Balance Framework (30 May 2019: mastery data excluded; new and reworked champions exempt for about a month) · /dev: Updating Champion Mastery (5 Jan 2024: uncapped levels) · Patch 26.22 notes (3 Nov 2026: Yasuo, Thresh) · Patch 26.1 notes (autofill carryover on dodge)
- Official wiki — Role Quests (read 8 Nov 2026)
- Data — iTero (Jack J. Williams): Mastery: a statistical summary of 1M+ games (24 Apr 2023, patch 13.7, pre-2024 mastery system) · Do Player Comforts Matter, Really? (30 Jun 2022, pro play 2020–22) · Deng et al., "Streaks and Coping", CHI PLAY Companion '24 (597,680 matches)
- The 2 + 1 + 1 pool, pairing rules and routines are BoostPirate's coaching advice. Checked 8 November 2026, live patch 26.22.
All free tools: calculators, percentiles & season countdowns →
Common questions
How many champions should I play in ranked?
Is one-tricking good for climbing in LoL?
How many champions do you need for ranked?
Should I change my main after a nerf?
Which role needs the most champion experience?
Ready to climb the Rift?
1-on-1 Coaching
A Master+ coach reviews your replays and your champion pool with you.
Solo/Duo Rank Boost
A verified booster on your roles takes the grind off your plate.
Ranked Net Wins
Pay per net win — climb steadily while you work on your game.
More League of Legends guides
All League of Legends guides →
League of Legends and all related assets are © Riot Games; splash art via Data Dragon. BoostPirate isn't endorsed by Riot Games.










