Dark Willow
How Dark Willow is doing in ranked games, and what works with it and against it.
Patch 7.41fRanked games from
Dark Willow wins 49.2% of ranked All Pick games, over 246,584 games. About the same as the week before.
Win rate by rank
Ranked All Pick, last full week, smoothed so small samples sit closer to 50%.
Herald46.7%4,898
Guardian47.4%18,508
Crusader48.0%32,072
Archon49.0%47,289
Legend49.4%59,055
Ancient49.8%44,913
Divine50.8%26,131
Immortal50.8%13,718
Where it's played
Share of its games in each position, and its win rate there.
- Carry1% of games38.4%603
- Mid5% of games46.1%4,422
- Offlane1% of games48.9%688
- Soft support48% of games49.2%46,760
- Hard support47% of games49.4%45,701
Matchups
Measured against what both heroes' overall strength predicts, so a strong hero doesn't top every list.
Good against
Dark Willow beats these heroes more often than expected.
Largo3.4 pts better than expected54.3%1,066
Medusa2.9 pts better than expected53.3%1,574
Drow Ranger2.5 pts better than expected55.9%6,174
Razor2.4 pts better than expected51.2%4,062
Sven2.2 pts better than expected50.0%6,864
Batrider2.2 pts better than expected57.3%492
Ember Spirit1.8 pts better than expected51.7%4,596
Lich1.7 pts better than expected48.2%7,842
Countered by
These heroes beat Dark Willow more often than expected.
Techies4.4 pts better than expected56.3%2,526
Naga Siren3.5 pts better than expected53.0%728
Elder Titan2.6 pts better than expected54.4%485
Bristleback2.4 pts better than expected51.9%3,618
Broodmother2.3 pts better than expected54.1%1,058
Juggernaut2.0 pts better than expected55.4%12,412
Puck2.0 pts better than expected50.1%2,068
Oracle2.0 pts better than expected52.7%2,613
Teammates
Teams with both heroes, against what each one's strength predicts.
Best with
Dark Willow wins more than expected next to these heroes.
Lycan6.6 pts better than expected54.6%806
Batrider5.6 pts better than expected48.8%409
Alchemist4.6 pts better than expected50.6%1,377
Enchantress3.9 pts better than expected50.2%1,043
Omniknight3.7 pts better than expected54.1%932
Morphling3.7 pts better than expected50.6%2,334
Visage3.5 pts better than expected56.4%844
Largo2.9 pts better than expected50.4%979
Worst with
Dark Willow wins less than expected next to these heroes.
Elder Titan4.0 pts worse than expected46.3%366
Faceless Void2.5 pts worse than expected45.2%4,500
Zeus2.5 pts worse than expected46.8%5,470
Ancient Apparition2.4 pts worse than expected48.3%2,800
Lich2.3 pts worse than expected49.7%4,041
Slardar1.5 pts worse than expected46.8%4,120
Disruptor1.5 pts worse than expected48.2%3,105
Warlock1.5 pts worse than expected47.8%1,990
Items
What Dark Willow players buy, how often, and how those games go.
Starting items
Share of games starting with each item.
Tango94%
Blood Grenade70%
Iron Branch69%
Magic Stick54%
Sentry Ward41%
Circlet38%
Observer and Sentry Wards35%
Faerie Fire34%
Boots
Share of games buying each, the average time, and the win rate.
Tranquil Boots51% of games · on average at 8:5149.8%64,567
Arcane Boots38% of games · on average at 11:4149.6%48,225
Boots of Speed29% of games · on average at 6:0049.4%36,149
Power Treads7% of games · on average at 14:2244.6%8,813
Core items
Items of 1,400 gold or more. The time is the median first purchase; the two win rates split those games at that minute.
Eul's Scepter of Divinity58% of games · usually at 23:00By 23:00: 53.7%Later: 51.1%52.5%73,630
Blink Dagger46% of games · usually at 26:00By 26:00: 59.1%Later: 57.1%58.2%58,071
Aghanim's Scepter26% of games · usually at 30:00By 30:00: 52.8%Later: 55.2%54.0%33,496
Glimmer Cape17% of games · usually at 26:00By 26:00: 52.0%Later: 49.9%51.0%21,769
Aghanim's Shard16% of games · usually at 28:00By 28:00: 57.2%Later: 53.5%55.5%20,525
Essence Distiller16% of games · usually at 15:00By 15:00: 55.7%Later: 48.9%52.7%20,006
Force Staff14% of games · usually at 29:00By 29:00: 52.7%Later: 52.8%52.8%17,177
Spirit Vessel8% of games · usually at 20:00By 20:00: 57.3%Later: 48.2%53.2%9,837
Moon Shard5% of games · usually at 39:00By 39:00: 63.9%Later: 58.3%61.7%6,947
Talents
Each level's two options: how often each is picked over the other, and its win rate.
- +2 Bedlam Attack Targets65% picked · 61.3% won500 Terrorize Impact Damage35% picked · 61.3% wonLevel 25
- +1.5s Shadow Realm Duration58% picked · 58.8% won-7s Bramble Maze CD42% picked · 61.7% wonLevel 20
- +30 Bedlam Damage70% picked · 55.5% won-20s Terrorize Cooldown30% picked · 54.4% wonLevel 15
- +1.0s Bedlam Duration82% picked · 49.8% won+12 Intelligence18% picked · 49.3% wonLevel 10
Last eight weeks
Ranked All Pick win rate by week.
- Oct 1, 202649.2%246,777
- Sep 24, 202649.1%228,929
- Sep 17, 202649.2%232,471
- Sep 10, 202648.9%153,915
- Sep 3, 202648.6%202,516
- Aug 27, 202648.5%291,492
- Aug 20, 202648.4%283,495
- Aug 13, 202648.6%242,088
Dark Willow in short
The quick version of this page, from the last full week of ranked games.
What is Dark Willow's win rate?
49.2% in ranked All Pick games, over 246,584 games. About the same as the week before.
What counters Dark Willow?
Techies, Naga Siren, and Elder Titan beat Dark Willow more often than their overall strength predicts. The clearest is Techies: 4.4 pts better than expected, over 2,526 games.
Which heroes is Dark Willow good against?
Dark Willow beats Largo, Medusa, and Drow Ranger more often than expected. The clearest is Largo: 3.4 pts better than expected, over 1,066 games.
What position is Dark Willow played in?
Mostly soft support (48% of its games), then hard support (47%). See the best soft support heroes
What should I build on Dark Willow?
Most games start with Tango, Blood Grenade, and Iron Branch. The usual boots are Tranquil Boots and Arcane Boots. The most bought core items are Eul's Scepter of Divinity, Blink Dagger, Aghanim's Scepter, and Glimmer Cape.
How we read this
Every number here is our analysis of STRATZ's game counts for the last full week of ranked games: win rates are smoothed toward what's expected, matchups and teammates are compared with both heroes' overall strength, and item timings come from first purchases. STRATZ's own scores aren't used.
Data from STRATZ