What is the best build for Rakan?

We do not publish a best build for Rakan, and the reason is worth thirty seconds. The most recorded Rakan support build on patch 16.14 is Zeke's Convergence, Knight's Vow and Locket of the Iron Solari, in 129 games. A finished item set is read at the end of the game, so a build only appears once the game ran long enough to buy it, and a game that runs long is usually a game you were winning. Any share of wins taken from those games leans high for that reason on its own, which is why there is not one in this answer. The page lists every recorded build with the games behind it.

What are the best runes for Rakan?

The most recorded Rakan support rune page on patch 16.14 takes Guardian out of Resolve, with Font of Life, Conditioning and Overgrowth beside it and Ultimate Hunter and Deep Ward out of Domination. Players who ran it won 48.7% of 915 games, range 45.5% to 52.0%. A rune page is locked in at champion select, so that share is not inflated by long games the way a build is, but it counts who picked the page rather than what the page did.

Rakan build, patch 16.14

Q: What skill order should I use on Rakan?

Recorded Rakan support players max W, then E, then Q. 452 recorded games on patch 16.14. Win comparisons only count games that ran long enough to finish the max: 94% of 410 W-first games and 91% of 11 E-first games. Short games are missing from both sides, and short games are usually decided ones.

1,668 games of Rakan recorded on patch 16.14, across support, mid, top, jungle and bot.

Win rate

49.5% of support games this patch

Pick rate

2.1% of support games

Games

1,642 support, patch 16.14

Rank in role

27 of 90 support champions

Win impact against the average support pick on patch 16.14: -0.2 wins per 100 games (range -1.9 to 1.4) 1,642 games Pooled

What most Rakan support players build, patch 16.14

Build Zeke's Convergence, Knight's Vow and Locket of the Iron Solari, and run Guardian

The rune page they run

Resolve

Guardian

Domination

Shards

915 games ran this page exactly as it is drawn. We never stitch the most picked keystone onto the most picked second row and call it a page: nobody ran that one.

The skill path Each filled cell is the skill most players took at that level, from 452 games.
skill	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18
Q														Q	Q		Q	Q
W	W			W	W		W		W
E		E	E					E		E		E	E
R						R					R					R

Levels 10 to 18

10 E, 11 R, 12 E, 13 E, 14 Q, 15 Q, 16 R, 17 Q, 18 Q

Max W, then E, then Q. The level by level count.

129 games finished this exact set. It is the set recorded most often, which is a fact about players rather than about the items.

Share of those games that were wins, in percent 61.2% (range 52.6-69.2) · 129 games · Measured

Games that finish a build are games that lasted, and lasting usually means winning, so this figure leans high before the items do anything. It is how those games went, not what the build does to your chances. How far that can bend a number.

Rakan support is the role with the most recorded games on patch 16.14: 1,642 games, of which 812 were wins. That is the role this page leads with. Measured on this page: support, the role that cleared 30 recorded games.

wins 0.2 per 100 vs the support average · 1,642 games on 16.14 · first-time players measured (1,526 games) · -12.0 vs 16.13

The build above is the one recorded most often, nothing more. Every recorded build, with its own games and its own share, is in the build list, in the order players built them.

Why this answer

A champion and role pair needs 30 recorded games before we put a score or a win impact on it: under that, a handful of lucky games moves a win rate by ten points. The same floor applies to a single build and to a single rune page, and anything under it prints its games and no number.

The share on that figure says how the games went, not what the build does to you. Nothing on this page measures that; the method note at the end says why.

What do most Rakan players build?

Item sets recorded on patch 16.14, most built first, never sorted by how the games went. The share is how many of a build's own games were wins; a build under 30 games shows its games and no share.

Rakan, support

8 item sets were recorded in this role and 3 of them cleared 30 games.

Zeke's Convergence, Knight's Vow and Locket of the Iron Solari
61.2% 129 games
Shurelya's Battlesong, Zeke's Convergence and Locket of the Iron Solari
60.7% 56 games
Bandlepipes, Zeke's Convergence and Locket of the Iron Solari
60.0% 40 games

A finished set is read at the end of the game, so every share here leans high: games that last long enough to finish a build are usually games that were already going well. How far that can bend a number.

What runes do most Rakan players run?

Whole pages that players actually ran, most recorded first. A page is locked in at champion select, so long games do not inflate its share the way they inflate a build's; it still counts who picked the page rather than what the page did.

Rakan, support

10 complete pages were recorded in this role and 8 of them cleared 30 games.

Guardian Resolve with Domination

915 games

Resolve

Guardian

Domination

Shards

915 games ran this page exactly as it is drawn. We never stitch the most picked keystone onto the most picked second row and call it a page: nobody ran that one.

Other pages players ran

Guardian Resolve with Sorcery

252 games

Unsealed Spellbook Inspiration with Domination

201 games

Guardian Resolve with Inspiration

64 games

Glacial Augment Inspiration with Domination

44 games

Electrocute Domination with Sorcery

38 games

Keystones picked, one slot on its own

Guardian: 50.4% 1,235 games
Unsealed Spellbook: 46.7% 229 games
Electrocute: 46.2% 106 games
Glacial Augment: 45.3% 53 games

7 keystones were recorded in this role: one slot on its own, counted across every page that used it, so the share carries the rest of each page with it.

What skill order should I use?

Counted from recorded level-ups on patch 16.14, one player per game.

Max W, then E, then Q

452 recorded games on patch 16.14. Win comparisons only count games that ran long enough to finish the max: 94% of 410 W-first games and 91% of 11 E-first games. Short games are missing from both sides, and short games are usually decided ones.

not enough games

Rakan support skill path Each filled cell is the skill most players took at that level, from 452 games.
skill	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18
Q														Q	Q		Q	Q
W	W			W	W		W		W
E		E	E					E		E		E	E
R						R					R					R

Levels 10 to 18

10 E, 11 R, 12 E, 13 E, 14 Q, 15 Q, 16 R, 17 Q, 18 Q

Almost nobody maxed E first, so there is nothing to compare. -1 wins per 100 games (range -11 to 9) 394 games Measured 384 games maxed W first, 10 games maxed E first.

When do players buy each piece?

The pieces of the most recorded Rakan support build, in the order players buy them. The time is the middle of the spread across every recorded game on patch 16.14, measured at the buy rather than at the end of the game: half the recorded buys of the item happen before it, half after.

Buy times, earliest first

Locket of the Iron Solari: 14:41 1,116 games
Zeke's Convergence: 18:53 410 games
Knight's Vow: 23:21 422 games

The bar is scaled to the latest piece of the set. Each item links to its own page, where the whole spread sits beside the games behind it.

Who counters Rakan?

The opponents Rakan support loses to most and the teammates Rakan gains the most beside on patch 16.14. The chip is the edge in wins per 100 games, and the games behind it sit under the name.

Hardest opponents

74 more opponents sit under 50 games against Rakan and are counted, not ranked. Every measured Rakan opponent, with the full table

Best teammates

Every measured Rakan teammate, with the full table

These numbers come from 108,616 ranked games we recorded on EUW, Korea and NA, found by following Diamond players and the people they play with. That is not every rank and it is not every region. Everything on this page is patch 16.14. Counted up to 24 July 2026.

How does Rakan rank?

Every role we recorded Rakan in on patch 16.14, scored or not. Win impact is measured against the average pick in that role on the same patch, and a role needs 30 recorded games before it is scored.

Rakan by role

support B -0.2 wins per 100 games (range -1.9 to 1.4) 1,642 games Pooled: 49 1,642 games
mid not enough games yet. 17 games so far, this needs 30.: not scored 17 games
top not enough games yet. 5 games so far, this needs 30.: not scored 5 games
jungle not enough games yet. 2 games so far, this needs 30.: not scored 2 games
bot not enough games yet. 2 games so far, this needs 30.: not scored 2 games

Why no share of wins on this page says a build is good for Rakan

A finished item set is read at the end of the game. A player who bought the first two items and lost at twenty minutes never finished the set, so they are not in its row. Everybody in the row lasted long enough to buy it, and lasting is most of what winning looks like. The share those games won is pushed up by that on its own, before the items do anything at all, and the push is largest on the most expensive builds and backwards on the ones you buy when a game is going badly.

We know the size of it. In simulation, with every item's real effect set to exactly zero, an item read at the end of the game shows about 92% as a sixth completed item. A defensive item with a real 5 point benefit shows about 20 points the wrong way, because it gets bought when a player is behind. The measurement is here.

So this page counts and labels rather than ranks. The estimate that answers the real question is taken at the moment of the buy, from the state the purchase happened in, and it runs in the desktop app because it needs the game in front of it.