What is the best build for Zed?

We do not publish a best build for Zed, and the reason is worth thirty seconds. The most recorded Zed mid build on patch 16.14 is Serylda's Grudge, Axiom Arc and Voltaic Cyclosword, in 91 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 Zed?

The most recorded Zed mid rune page on patch 16.14 takes First Strike out of Inspiration, with Cash Back, Triple Tonic and Cosmic Insight beside it and Cut Down and Legend: Haste out of Precision. Players who ran it won 48.8% of 1,238 games, range 46.0% to 51.6%. 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.

Zed build, patch 16.14

Q: What skill order should I use on Zed?

Recorded Zed mid players max Q, then E, then W. 588 recorded games on patch 16.14. Win comparisons only count games that ran long enough to finish the max: 99% of 587 Q-first games and 100% of 1 E-first games. Short games are missing from both sides, and short games are usually decided ones.

3,220 games of Zed recorded on patch 16.14, across mid, jungle, top, support and bot.

Win rate

47.5% of mid games this patch

Pick rate

3.1% of mid games

Games

2,409 mid, patch 16.14

Rank in role

100 of 109 mid champions

Win impact against the average mid pick on patch 16.14: -1.3 wins per 100 games (range -2.8 to 0.1) 2,409 games Pooled

What most Zed mid players build, patch 16.14

Build Serylda's Grudge, Axiom Arc and Voltaic Cyclosword, and run First Strike

The rune page they run

Inspiration

First Strike

Precision

Shards

1,238 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 588 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		Q
W		W												W	W	W		W
E			E					E		E	E		E
R						R						R					R

Levels 10 to 18

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

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

91 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 44.0% (range 34.2-54.2) · 91 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.

Zed mid is the role with the most recorded games on patch 16.14: 2,409 games, of which 1,145 were wins. That is the role this page leads with. Measured on this page: mid, jungle, top, support and bot, the roles that cleared 30 recorded games.

wins -1.3 per 100 vs the mid average · 2,409 games on 16.14 · first-time players measured (2,167 games) · -18.0 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 Zed 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.

Zed, mid

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

Serylda's Grudge, Axiom Arc and Voltaic Cyclosword
44.0% 91 games
Edge of Night, Serylda's Grudge, Axiom Arc and Voltaic Cyclosword
58.9% 90 games
Edge of Night, Axiom Arc and Voltaic Cyclosword
66.7% 54 games
Serylda's Grudge, Serpent's Fang, Axiom Arc and Voltaic Cyclosword
52.2% 46 games
Edge of Night, Serylda's Grudge, Serpent's Fang, Axiom Arc and Voltaic Cyclosword
55.6% 45 games
Bastionbreaker, Serylda's Grudge and Voltaic Cyclosword
42.1% 38 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.

Zed, jungle

8 item sets were recorded in this role and 0 of them cleared 30 games.

Edge of Night, Serylda's Grudge, Axiom Arc and Voltaic Cyclosword
not enough games yet. 25 games so far, this needs 30. 25 games
Serylda's Grudge, Axiom Arc and Voltaic Cyclosword
not enough games yet. 18 games so far, this needs 30. 18 games
Edge of Night, Axiom Arc and Voltaic Cyclosword
not enough games yet. 12 games so far, this needs 30. 12 games

Zed, top

0 item sets were recorded in this role and 0 of them cleared 30 games.

No top build recorded yet

We have 218 games of Zed top on patch 16.14, but no item set repeated often enough to record as a build.

Read the rune pages instead

Zed, support

0 item sets were recorded in this role and 0 of them cleared 30 games.

No support build recorded yet

We have 51 games of Zed support on patch 16.14, but no item set repeated often enough to record as a build.

Read the rune pages instead

Zed, bot

0 item sets were recorded in this role and 0 of them cleared 30 games.

No bot build recorded yet

We have 41 games of Zed bot on patch 16.14, but no item set repeated often enough to record as a build.

Read the rune pages instead

What runes do most Zed 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.

Zed, mid

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

First Strike Inspiration with Precision

1,238 games

Inspiration

First Strike

Precision

Shards

1,238 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

Electrocute Domination with Sorcery

414 games

Conqueror Precision with Sorcery

185 games

Electrocute Domination with Precision

151 games

Conqueror Precision with Resolve

125 games

First Strike Inspiration with Sorcery

110 games

Keystones picked, one slot on its own

First Strike: 48.9% 1,416 games
Electrocute: 46.3% 627 games
Conqueror: 44.8% 382 games

4 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.

Zed, jungle

10 complete pages were recorded in this role and 3 of them cleared 30 games.

First Strike Inspiration with Domination

335 games

Inspiration

First Strike

Domination

Shards

335 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

First Strike Inspiration with Precision

53 games

Conqueror Precision with Sorcery

39 games

Keystones picked, one slot on its own

First Strike: 47.1% 399 games
Conqueror: 35.4% 65 games
Electrocute: 43.8% 32 games

4 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.

Zed, top

10 complete pages were recorded in this role and 2 of them cleared 30 games.

First Strike Inspiration with Precision

127 games

Inspiration

First Strike

Precision

Shards

127 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

Conqueror Precision with Sorcery

30 games

Keystones picked, one slot on its own

First Strike: 53.9% 152 games
Conqueror: 58.5% 53 games

3 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.

Zed, support

4 complete pages were recorded in this role and 1 of them cleared 30 games.

Electrocute Domination with Sorcery

33 games

Domination

Electrocute

Sorcery

Shards

33 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.

Keystones picked, one slot on its own

Electrocute: 53.8% 39 games

2 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.

Zed, bot

5 complete pages were recorded in this role and 0 of them cleared 30 games.

Conqueror Precision with Inspiration

not enough games yet. 14 games so far, this needs 30. 14 games

Precision

Conqueror

Inspiration

Shards

14 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

First Strike Inspiration with Precision

not enough games yet. 12 games so far, this needs 30. 12 games

Electrocute Domination with Sorcery

not enough games yet. 9 games so far, this needs 30. 9 games

Keystones picked, one slot on its own

Conqueror: not enough games yet. 17 games so far, this needs 30. 17 games
First Strike: not enough games yet. 12 games so far, this needs 30. 12 games
Electrocute: not enough games yet. 11 games so far, this needs 30. 11 games

3 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 Q, then E, then W

588 recorded games on patch 16.14. Win comparisons only count games that ran long enough to finish the max: 99% of 587 Q-first games and 100% of 1 E-first games. Short games are missing from both sides, and short games are usually decided ones.

not enough games

Zed mid skill path Each filled cell is the skill most players took at that level, from 588 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		Q
W		W												W	W	W		W
E			E					E		E	E		E
R						R						R					R

Levels 10 to 18

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

Almost nobody maxed E first, so there is nothing to compare. 0 wins per 100 games (range -11 to 10) 583 games Measured 582 games maxed Q first, 1 game maxed E first.

Max Q, then E, then W

119 recorded games on patch 16.14.

Zed jungle skill path Each filled cell is the skill most players took at that level, from 119 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		Q
W			W											W	W	W		W
E		E						E		E	E		E
R						R						R					R

Levels 10 to 18

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

Max Q, then E, then W

Only 63 recorded games on patch 16.14. A soft read.

Zed top skill path Each filled cell is the skill most players took at that level, from 63 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		Q
W		W												W	W	W		W
E			E					E		E	E		E
R						R						R					R

Levels 10 to 18

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

Zed, support

Not enough support level-ups yet

not enough games yet. 11 games so far, this needs 30.

Zed support has 51 games recorded, but only 11 games of them carry level-ups. A levelling order off that many games is luck.

How we count level-ups

Zed, bot

Not enough bot level-ups yet

not enough games yet. 13 games so far, this needs 30.

Zed bot has 41 games recorded, but only 13 games of them carry level-ups. A levelling order off that many games is luck.

How we count level-ups

When do players buy each piece?

The pieces of the most recorded Zed mid 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

Voltaic Cyclosword: 14:21 1,140 games
Axiom Arc: 17:11 454 games
Serylda's Grudge: 25:00 482 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 Zed?

The opponents Zed mid loses to most and the teammates Zed 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

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

Best teammates

Every measured Zed 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 Zed rank?

Every role we recorded Zed 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.

Zed by role

mid C -1.3 wins per 100 games (range -2.8 to 0.1) 2,409 games Pooled: 43 2,409 games
jungle C -1.1 wins per 100 games (range -3.0 to 0.9) 501 games Pooled: 41 501 games
top B +0.5 wins per 100 games (range -1.5 to 2.6) 218 games Pooled: 48 218 games
support C +0.1 wins per 100 games (range -2.0 to 2.3) 51 games Pooled: 46 51 games
bot C +0.2 wins per 100 games (range -1.9 to 2.4) 41 games Pooled: 46 41 games

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

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.