What is the best build for Bard?

We do not publish a best build for Bard, and the reason is worth thirty seconds. The most recorded Bard support build on patch 16.14 is Bandlepipes, Locket of the Iron Solari and Dead Man's Plate, in 133 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 Bard?

The most recorded Bard support rune page on patch 16.14 takes Electrocute out of Domination, with Cheap Shot, Deep Ward and Relentless Hunter beside it and Magical Footwear and Cosmic Insight out of Inspiration. Players who ran it won 49.1% of 2,088 games, range 47.0% to 51.3%. 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.

Bard build, patch 16.14

Q: What skill order should I use on Bard?

Recorded Bard support players max Q, then W, then E. 953 recorded games on patch 16.14. Win comparisons only count games that ran long enough to finish the max: 93% of 951 Q-first games and 100% of 2 W-first games. Short games are missing from both sides, and short games are usually decided ones.

3,457 games of Bard recorded on patch 16.14, across support, top, mid and bot.

Win rate

49.4% of support games this patch

Pick rate

4.5% of support games

Games

3,444 support, patch 16.14

Rank in role

14 of 90 support champions

Win impact against the average support pick on patch 16.14: -0.4 wins per 100 games (range -1.7 to 1.0) 3,444 games Pooled

What most Bard support players build, patch 16.14

Build Bandlepipes, Locket of the Iron Solari and Dead Man's Plate, and run Electrocute

The rune page they run

Domination

Electrocute

Inspiration

Shards

2,088 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 953 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 W, 11 R, 12 W, 13 W, 14 E, 15 E, 16 R, 17 E, 18 E

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

133 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 53.4% (range 44.9-61.6) · 133 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.

Bard support is the role with the most recorded games on patch 16.14: 3,444 games, of which 1,702 were wins. That is the role this page leads with. Measured on this page: support, the role that cleared 30 recorded games.

wins 0.4 per 100 vs the support average · 3,444 games on 16.14 · first-time players measured (3,077 games) · -10.0 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 Bard 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.

Bard, support

8 item sets were recorded in this role and 7 of them cleared 30 games.

Bandlepipes, Locket of the Iron Solari and Dead Man's Plate
53.4% 133 games
Locket of the Iron Solari, Dead Man's Plate and Imperial Mandate
55.9% 93 games
Bandlepipes, Locket of the Iron Solari and Mikael's Blessing
64.4% 73 games
Locket of the Iron Solari, Dead Man's Plate and Liandry's Torment
62.3% 61 games
Locket of the Iron Solari, Mikael's Blessing and Dead Man's Plate
55.9% 59 games
Redemption, Locket of the Iron Solari and Dead Man's Plate
59.2% 49 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 Bard 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.

Bard, support

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

Electrocute Domination with Inspiration

2,088 games

Domination

Electrocute

Inspiration

Shards

2,088 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

801 games

Electrocute Domination with Resolve

351 games

Deathfire Touch Sorcery with Inspiration

86 games

Keystones picked, one slot on its own

Electrocute: 49.4% 3,252 games
Deathfire Touch: 47.4% 95 games
Unsealed Spellbook: 51.1% 45 games
Guardian: 48.8% 43 games

8 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 W, then E

953 recorded games on patch 16.14. Win comparisons only count games that ran long enough to finish the max: 93% of 951 Q-first games and 100% of 2 W-first games. Short games are missing from both sides, and short games are usually decided ones.

not enough games

Bard support skill path Each filled cell is the skill most players took at that level, from 953 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 W, 11 R, 12 W, 13 W, 14 E, 15 E, 16 R, 17 E, 18 E

Almost nobody maxed W first, so there is nothing to compare. +2 wins per 100 games (range -9 to 12) 882 games Measured 880 games maxed Q first, 2 games maxed W first.

When do players buy each piece?

The pieces of the most recorded Bard 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
Dead Man's Plate: 20:18 409 games
Bandlepipes: 22:46 305 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 Bard?

The opponents Bard support loses to most and the teammates Bard 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

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

Best teammates

Every measured Bard 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 Bard rank?

Every role we recorded Bard 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.

Bard by role

support B -0.4 wins per 100 games (range -1.7 to 1.0) 3,444 games Pooled: 53 3,444 games
top not enough games yet. 9 games so far, this needs 30.: not scored 9 games
mid not enough games yet. 3 games so far, this needs 30.: not scored 3 games
bot not enough games yet. 1 game so far, this needs 30.: not scored 1 game

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

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.