What is the best build for Varus?

We do not publish a best build for Varus, and the reason is worth thirty seconds. The most recorded Varus bot build on patch 16.14 is Statikk Shiv, Guinsoo's Rageblade and Terminus, in 87 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 Varus?

The most recorded Varus bot rune page on patch 16.14 takes Lethal Tempo out of Precision, with Presence of Mind, Legend: Alacrity and Cut Down beside it and Triple Tonic and Cash Back out of Inspiration. Players who ran it won 49.5% of 392 games, range 44.6% to 54.4%. 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.

Varus build, patch 16.14

Q: What skill order should I use on Varus?

Recorded Varus bot players max Q, then W, then E. 214 recorded games on patch 16.14. Win comparisons only count games that ran long enough to finish the max: 99% of 160 Q-first games and 98% of 52 W-first games. Short games are missing from both sides, and short games are usually decided ones.

1,122 games of Varus recorded on patch 16.14, across bot, top, mid, support and jungle.

Win rate

47.2% of bot games this patch

Pick rate

0.9% of bot games

Games

725 bot, patch 16.14

Rank in role

73 of 77 bot champions

Win impact against the average bot pick on patch 16.14: -0.8 wins per 100 games (range -2.6 to 1.1) 725 games Pooled

What most Varus bot players build, patch 16.14

Build Statikk Shiv, Guinsoo's Rageblade and Terminus, and run Lethal Tempo

The rune page they run

Precision

Lethal Tempo

Inspiration

Shards

392 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 214 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
E	E		E											E	E			E
R						R					R	R				R	R

Levels 10 to 18

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

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

87 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 57% (range 47-67) · 87 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.

Varus bot is the role with the most recorded games on patch 16.14: 725 games, of which 342 were wins. That is the role this page leads with. Measured on this page: bot, top and mid, the roles that cleared 30 recorded games.

wins 0.8 per 100 vs the bot average · 725 games on 16.14 · first-time split not measured yet, neutral 50 · no clear move vs 16.13 (+3.0, range -10.0 to +16.0)

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 Varus 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.

Varus, bot

6 item sets were recorded in this role and 2 of them cleared 30 games.

Statikk Shiv, Guinsoo's Rageblade and Terminus
57.5% 87 games
Statikk Shiv, Guinsoo's Rageblade, Terminus and Jak'Sho, The Protean
46.9% 32 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.

Varus, top

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

Dusk and Dawn, Experimental Hexplate and Riftmaker
not enough games yet. 27 games so far, this needs 30. 27 games
Dusk and Dawn, Experimental Hexplate and Zhonya's Hourglass
not enough games yet. 13 games so far, this needs 30. 13 games

Varus, mid

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

No mid build recorded yet

We have 53 games of Varus mid 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 Varus 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.

Varus, bot

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

Lethal Tempo Precision with Inspiration

392 games

Precision

Lethal Tempo

Inspiration

Shards

392 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

Arcane Comet Sorcery with Inspiration

178 games

Press the Attack Precision with Inspiration

50 games

Lethal Tempo Precision with Resolve

33 games

Keystones picked, one slot on its own

Lethal Tempo: 49.2% 441 games
Arcane Comet: 38.0% 192 games
Press the Attack: 54.0% 87 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.

Varus, top

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

Press the Attack Precision with Resolve

187 games

Precision

Press the Attack

Resolve

Shards

187 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

Lethal Tempo Precision with Resolve

54 games

Lethal Tempo Precision with Inspiration

46 games

Keystones picked, one slot on its own

Press the Attack: 54.4% 228 games
Lethal Tempo: 52.4% 105 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.

Varus, mid

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

Press the Attack Precision with Inspiration

not enough games yet. 17 games so far, this needs 30. 17 games

Precision

Press the Attack

Inspiration

Shards

17 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

Lethal Tempo Precision with Inspiration

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

Press the Attack Precision with Resolve

not enough games yet. 8 games so far, this needs 30. 8 games

Keystones picked, one slot on its own

Press the Attack: 40.0% 30 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 W, then E

214 recorded games on patch 16.14. Win comparisons only count games that ran long enough to finish the max: 99% of 160 Q-first games and 98% of 52 W-first games. Short games are missing from both sides, and short games are usually decided ones.

too close

Varus bot skill path Each filled cell is the skill most players took at that level, from 214 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
E	E		E											E	E			E
R						R					R	R				R	R

Levels 10 to 18

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

Maxing Q and maxing W are too close to separate. -3.7 wins per 100 games (range -13.5 to 6.2) 209 games Measured 158 games maxed Q first, 51 games maxed W first.

No measured difference between maxing Q and W first, 209 games.

Max W, then Q, then E

Only 98 recorded games on patch 16.14. A soft read. Win comparisons only count games that ran long enough to finish the max: 100% of 88 W-first games and 100% of 10 Q-first games. Short games are missing from both sides, and short games are usually decided ones.

not enough games

Varus top skill path Each filled cell is the skill most players took at that level, from 98 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 Q, 11 R, 12 Q, 13 Q, 14 E, 15 E, 16 R, 17 E, 18 E

Almost nobody maxed Q first, so there is nothing to compare. +2 wins per 100 games (range -10 to 13) 98 games Measured 88 games maxed W first, 10 games maxed Q first.

Varus, mid

Not enough mid level-ups yet

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

Varus mid has 53 games recorded, but only 14 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 Varus bot 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

Statikk Shiv: 12:22 355 games
Guinsoo's Rageblade: 17:25 362 games
Terminus: 23:24 126 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 Varus?

The opponents Varus bot loses to most and the teammates Varus 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

Not enough games yet to rank Varus's opponents. Varusbot has 769 games recorded on patch 16.14, and ranking opponents needs 1,000 before the baseline is steady; every opponent so far is counted, not ranked.

Best teammates

Not enough games yet to rank the champions Varus wins beside.

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 Varus rank?

Every role we recorded Varus 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.

Varus by role

bot C -0.8 wins per 100 games (range -2.6 to 1.1) 725 games Pooled: 43 725 games
top B +0.5 wins per 100 games (range -1.5 to 2.6) 337 games Pooled: 49 337 games
mid C -0.2 wins per 100 games (range -2.4 to 1.9) 53 games Pooled: 44 53 games
support not enough games yet. 6 games so far, this needs 30.: not scored 6 games
jungle 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 Varus

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.