What is the best build for Lee Sin?

We do not publish a best build for Lee Sin, and the reason is worth thirty seconds. The most recorded Lee Sin jungle build on patch 16.14 is Death's Dance, Sundered Sky and Eclipse, in 618 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 Lee Sin?

The most recorded Lee Sin jungle rune page on patch 16.14 takes Conqueror out of Precision, with Triumph, Legend: Alacrity and Cut Down beside it and Magical Footwear and Cosmic Insight out of Inspiration. Players who ran it won 51.0% of 5,328 games, range 49.7% to 52.4%. 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.

Lee Sin build, patch 16.14

Q: What skill order should I use on Lee Sin?

Recorded Lee Sin jungle players max Q, then W, then E. 1,472 recorded games on patch 16.14. Win comparisons only count games that ran long enough to finish the max: 100% of 1,461 Q-first games and 20% of 10 E-first games. Short games are missing from both sides, and short games are usually decided ones.

6,002 games of Lee Sin recorded on patch 16.14, across jungle, top, support, mid and bot.

Win rate

50.5% of jungle games this patch

Pick rate

7.5% of jungle games

Games

5,782 jungle, patch 16.14

Rank in role

4 of 77 jungle champions

Win impact against the average jungle pick on patch 16.14: +0.4 wins per 100 games (range -0.7 to 1.5) 5,782 games Pooled

What most Lee Sin jungle players build, patch 16.14

Build Death's Dance, Sundered Sky and Eclipse, and run Conqueror

The rune page they run

Precision

Conqueror

Inspiration

Shards

5,328 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 1,472 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.

618 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 57.6% (range 53.7-61.4) · 618 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.

Lee Sin jungle is the role with the most recorded games on patch 16.14: 5,782 games, of which 2,921 were wins. That is the role this page leads with. Measured on this page: jungle, top, support and mid, the roles that cleared 30 recorded games.

wins 0.4 per 100 vs the jungle average · 5,782 games on 16.14 · first-time players measured (5,346 games) · no clear move vs 16.13 (+3.0, range -5.0 to +11.0)

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 Lee Sin 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.

Lee Sin, jungle

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

Death's Dance, Sundered Sky and Eclipse
57.6% 618 games
Black Cleaver, Sundered Sky and Eclipse
54.8% 363 games
Black Cleaver, Death's Dance and Sundered Sky
55.5% 227 games
Maw of Malmortius, Sundered Sky and Eclipse
52.7% 165 games
Guardian Angel, Death's Dance, Sundered Sky and Eclipse
68.1% 160 games
Guardian Angel, Sundered Sky and Eclipse
73.6% 159 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.

Lee Sin, top

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

No top build recorded yet

We have 92 games of Lee Sin top on patch 16.14, but no item set repeated often enough to record as a build.

Read the rune pages instead

Lee Sin, support

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

No support build recorded yet

We have 61 games of Lee Sin support on patch 16.14, but no item set repeated often enough to record as a build.

Read the rune pages instead

Lee Sin, mid

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

No mid build recorded yet

We have 56 games of Lee Sin 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 Lee Sin 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.

Lee Sin, jungle

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

Conqueror Precision with Inspiration

5,328 games

Precision

Conqueror

Inspiration

Shards

5,328 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 Domination

335 games

Electrocute Domination with Precision

61 games

Keystones picked, one slot on its own

Conqueror: 50.7% 5,690 games
Electrocute: 44.4% 81 games
Dark Harvest: 47.4% 57 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.

Lee Sin, top

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

Conqueror Precision with Resolve

42 games

Precision

Conqueror

Resolve

Shards

42 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

Conqueror: 49.3% 71 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.

Lee Sin, support

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

Electrocute Domination with Precision

not enough games yet. 29 games so far, this needs 30. 29 games

Domination

Electrocute

Precision

Shards

29 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 Inspiration

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

Conqueror Precision with Domination

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

Keystones picked, one slot on its own

Electrocute: 38.1% 42 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.

Lee Sin, mid

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

Conqueror Precision with Resolve

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

Precision

Conqueror

Resolve

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

Electrocute Domination with Precision

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

Electrocute Domination with Resolve

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

Keystones picked, one slot on its own

Conqueror: not enough games yet. 27 games so far, this needs 30. 27 games
Electrocute: not enough games yet. 26 games so far, this needs 30. 26 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.

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

1,472 recorded games on patch 16.14. Win comparisons only count games that ran long enough to finish the max: 100% of 1,461 Q-first games and 20% of 10 E-first games. Short games are missing from both sides, and short games are usually decided ones.

not enough games

Lee Sin jungle skill path Each filled cell is the skill most players took at that level, from 1,472 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 E first, so there is nothing to compare. +0.2 wins per 100 games (range -9.7 to 10.2) 1,460 games Measured 1,458 games maxed Q first, 2 games maxed E first.

Lee Sin, top

Not enough top level-ups yet

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

Lee Sin top has 92 games recorded, but only 17 games of them carry level-ups. A levelling order off that many games is luck.

How we count level-ups

Lee Sin, support

Not enough support level-ups yet

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

Lee Sin support has 61 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

Lee Sin, mid

Not enough mid level-ups yet

not enough games yet. 19 games so far, this needs 30.

Lee Sin mid has 56 games recorded, but only 19 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 Lee Sin jungle 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

Eclipse: 11:05 615 games
Sundered Sky: 15:30 1,165 games
Death's Dance: 25:04 1,079 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 Lee Sin?

The opponents Lee Sin jungle loses to most and the teammates Lee Sin 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

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

Best teammates

Every measured Lee Sin 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.

Lee Sin in the jungle

Camp times are counted across the whole store rather than one patch, so they do not move with the patch above.

First camp

raptors 30.6% of 1,539 games

First dragon

8:17 middle of the spread, across 1,530 games of every jungler

Jungle games

1,539 Lee Sin, the whole store

Routes, most recorded first

Camp times come from per-minute match frames and are estimates with about a minute of resolution. Clear speed varies with leash and skill order. A route needs 30 games before we put a win figure on it.

full clear not enough games yet. 5 games so far, this needs 30. 5 games
1. blue buff 1:30
2. wolves 1:53
3. gromp 2:15
4. red buff 2:30
5. krugs 2:45
6. raptors 3:00
3 camp gank not enough games yet. 9 games so far, this needs 30. 9 games
1. raptors 1:30
2. krugs 2:00
3. red buff 2:30

How does Lee Sin rank?

Every role we recorded Lee Sin 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.

Lee Sin by role

jungle B +0.4 wins per 100 games (range -0.7 to 1.5) 5,782 games Pooled: 62 5,782 games
top C 0.0 wins per 100 games (range -2.2 to 2.1) 92 games Pooled: 45 92 games
support C -0.2 wins per 100 games (range -2.4 to 1.9) 61 games Pooled: 44 61 games
mid C +0.1 wins per 100 games (range -2.1 to 2.3) 56 games Pooled: 45 56 games
bot not enough games yet. 11 games so far, this needs 30.: not scored 11 games

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

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.