What the world's best prioritized: MSI & EWC 2026 draft trends
Updated 18 July 2026. Sample: 110 international games (71 MSI + 39 EWC so far), a single patch (Leaguepedia 26.13 = Riot 16.13). Source: Leaguepedia competitive drafts.
Pro drafts are a small, strange, high-signal dataset. There are only a hundred-odd games here, but every champion in them was chosen by the best players in the world under cutthroat bans, after coaching staffs spent the week deciding what is strictly good. You don't need thousands of games to read that. You just have to read the right number, and know what it can't tell you.
How to read a 110-game sample without fooling yourself. The load-bearing number here is presence — how often a champion is picked or banned. Presence is a decision, not a rate: when a team bans Poppy, that is five analysts saying "not this one," and it's stable even at this sample size. Win rates are the opposite: a champion played 20 times going 11–9 tells you almost nothing, so per-champion win rates below carry their game count and should be read as soft. And the whole board is the pro, coordinated-play meta — it is not solo-queue advice. Some of the most-contested pro champions need a five-person setup that your soloq game will never have.
The ban board: who pros refuse to let you have
The clearest signal in pro draft isn't what gets picked — it's what never gets the chance. Across the two events, one champion sits alone at the top of the ban board while barely being played at all.
| Champion | Ban rate | Pick rate | Read |
|---|---|---|---|
| Poppy | 61% | 7% | Banned first, rarely allowed through |
| Vi | 53% | 20% | Contest-or-remove engage |
| Jayce | 51% | 19% | Lane tyrant, denied on sight |
| Nocturne | 46% | 13% | Rising ban (see below) |
| Orianna | 45% | 22% | The mid pros both want and fear |
| Akali | 41% | 9% | Ban-to-deny flex threat |
Poppy is the story of the meta. She is banned in 61% of games and picked in 7%. That gap — the widest on the board — is what a "too strong to allow" champion looks like: teams would rather spend a ban than solve her, so almost no one gets to play her. When she does slip through, the games she's in are won 62% of the time (n=8, so read it soft, but it points the same direction as the ban rate).
The MSI → EWC shift
Two internationals a few weeks apart, same patch, let us watch the meta move. The bans tightened around two junglers:
| Champion | MSI | EWC | Shift |
|---|---|---|---|
| Poppy | 57% | 69% | ↑ more feared |
| Nocturne | 39% | 58% | ↑ sharply up |
| Vi | 53% | 53% | flat, still core |
| Akali | 45% | 35% | ↓ cooling |
| Orianna | 47% | 41% | ↓ slightly |
By EWC, teams were spending bans to remove Nocturne and Poppy from the jungle before the draft even started — EWC's presence board is topped by Poppy (82%) and Vi (79%). The meta didn't broaden between the two events; it narrowed onto a tighter set of must-answer picks.
Priority picks: what teams actually first-pick
When a champion isn't banned, presence and pick-order show what pros reach for. Jarvan IV was the single most-picked champion of the two events; Camille quietly became a priority support flex.
| Champion | Role | Presence | Pick | Win rate |
|---|---|---|---|---|
| Vi | Jungle | 74% | 20% | 43% (n=23) |
| Jayce | Top | 70% | 19% | 61% (n=21) |
| Orianna | Mid | 68% | 22% | 56% (n=25) |
| Camille | Support | 54% | 28% | 70% (n=31) |
| Lee Sin | Jungle | 50% | 26% | 51% (n=29) |
| Gnar | Top | 48% | 25% | 50% (n=28) |
| Jarvan IV | Jungle | 47% | 31% | 45% (n=35) |
Camille support is the read worth flagging: 28% pick rate as a support and a 70% win rate across 31 games. On a soloq tier list that would be a rounding error; in the pro meta it's a coordinated engage-and-lockdown pattern teams drafted on purpose. It's also a clean example of why pro tier ≠ soloq tier — a roaming engage support like this lives or dies on five-person timing you won't get in your ranked game.
The blue-side tax
The single most decisive draft fact in the sample isn't a champion at all — it's which side you're on. Blue side, which drafts first and controls the first-pick sweep, won 62% of all 110 games. Unlike per-champion win rates, this one is meaningful at this sample size, because every game contributes to it.
| Event | Games | Blue-side win rate |
|---|---|---|
| MSI 2026 | 71 | 66% |
| EWC 2026 (so far) | 39 | 56% |
| Combined | 110 | 62% |
At the highest level of coordinated play, first draft rotation converts into wins. It cooled from MSI to EWC — teams on red side answering the blue priority a little better — but the advantage never disappeared. If you take one thing from this page into your own games: side selection and first-pick priority are worth more than any single pocket pick.
Pairing trends: the duos that got drafted together
Which champions the best teams put on the same side, and how those pairs did. At this sample size treat these as what teams chose to pair, not proof the pairing is strong — the co-pick frequency is the signal; the win rate is directional.
| Pairing | Games together | Win rate |
|---|---|---|
| Camille + Ziggs | 12 | 66% |
| Ambessa + Lee Sin | 9 | 55% |
| Ezreal + Renekton | 8 | 62% |
| Bard + Ezreal | 8 | 50% |
| Lucian + Milio | 8 | 50% |
The Camille + Ziggs pairing — a diving engage front paired with siege-and-zone poke — was the most-repeated composition seed of the two events, and the one that most often won. That's the kind of intentional draft shape you can read off pro data even when the sample is thin.
Counter trends, and their limit
We can see which lane matchups the best teams walked into on purpose — same role, opposite sides. But this is where the sample runs out: the most-contested single matchup happened seven times. So we'll show you the matchups pros repeatedly drafted, and we will not pretend a 7-game win rate is a matchup verdict.
| Lane matchup | Games | Blue won |
|---|---|---|
| Camille vs Leona | 7 | 57% |
| Jarvan IV vs Xin Zhao | 6 | 66% |
| Cassiopeia vs Syndra | 6 | 50% |
| Orianna vs Viktor | 5 | 60% |
The honest read is the left column, not the right: these are the mirrors and answer-picks pro teams kept drafting into. The win rates move with a single game and should be read as nothing more than a hint.
The honest caveats
- Pro is not soloq. Everything here is the coordinated, banned-to-the-teeth pro meta on one patch. It tells you what the best teams think is strong right now; it does not tell you what climbs in your ranked games, and for some champions the two disagree hard.
- Presence over win rate. With 110 games, pick and ban rates are stable and meaningful; per-champion, per-pairing and per-matchup win rates are small samples and are labelled with their game counts on purpose. Read the counts.
- EWC is live. The Esports World Cup is still being played as this publishes, so its numbers are 39 games and counting. This page re-runs from the same competitive database as new games are played — the trends will sharpen.
This is the pro-meta lens. For the soloq version applied to your own lobby, the Champion Scout reads pro pick-priority by role, and the Scrypick overlay ranks your live champ-select options by matchup, synergy and your own mastery, with a calibrated win chance on each.