Shadow Fiend
Our own data, from ranked matches above 7500 MMR.

Ruína das Sombras
O Shadow Fiend arrasa uma área diretamente à sua frente, causando dano a todas as unidades inimigas na área e dano adicional proporcional à quantidade de almas da Necromaestria . Aplica um efeito negativo cumulativo que faz o inimigo sofrer dano adicional caso seja atingido por outra Ruína das Sombras.

Shadowraze
Shadow Fiend razes the ground a short distance away from him, dealing damage to enemy units in the area, dealing bonus damage based on his current Necromastery Soul amount. Adds a stacking damage amplifier on the target that causes the enemy to take bonus Shadowraze damage per stack.

Shadowraze
Shadow Fiend razes the ground a longer distance away from him, dealing damage to enemy units in the area, dealing bonus damage based on his current Necromastery Soul amount. Adds a stacking damage amplifier on the target that causes the enemy to take bonus Shadowraze damage per stack.

Banquete de Almas
O Shadow Fiend ganha velocidades de movimento e ataque adicionais por 8s. Durante o efeito, a cada 0.5s, ele coleta almas de 2 inimigos em até 600 unidades de distância. Após o término do efeito, ele perde as almas cujos donos ainda estejam vivos, e mantém o resto por 8s — depois disso, as almas coletadas que ultrapassem o limite da Necromaestria também são perdidas. Somente pode coletar almas de cada inimigo uma vez.

Presença do Lorde das Trevas
A presença do Shadow Fiend reduz a armadura de inimigos próximos.

Réquiem das Almas
O Shadow Fiend reúne até 20 almas capturadas para libertá-las em linhas de energia demoníaca. Unidades próximas no momento da libertação podem sofrer o dano de várias linhas de energia. Também amedrontará, reduzirá as velocidades de movimento e a resistência mágica de unidades que sofrerem dano proveniente da habilidade, com duração de 0.6 segundo por cada linha atingida, até um máximo de 2.15. segundos. Linhas de energia são criadas por cada alma capturada pela Necromaestria . O Réquiem das Almas é automaticamente libertado ao morrer, independente do tempo de intervalo entre usos.

Necromaestria
O Shadow Fiend rouba a alma das unidades que matar, ganhando 1.35 de dano adicional por alma. Heróis concedem 4 almas. Ao morrer, ele liberta 30% das almas do seu cativeiro.
Skill order
1
2
3
4
5
6
7
8
9
10- 11
12
13
14- +215
- 16
- +218
- +220
- +225
Talents
- Shadowraze Applies Attack Damage73% · WR 54%25Feast of Souls Grants +30% Cast Speed27% · WR 62%
- +0.2s Requiem Fear per line18% · WR 48%20+5 Necromastery Max Souls82% · WR 50%
- +2 Feast of Souls Souls Collected Per Hero21% · WR 37%15+1.5 Presence Armor Reduction79% · WR 47%
- +30 Feast of Souls Attack Speed72% · WR 46%10+30 Shadowraze Stack Damage28% · WR 40%
Items
Starting kits

222350%

12851%
2


6028%
Starting items
92%44%
74%42%
54%48%
42%39%
30%46%
28%39%
19%32%
13%44%
9%
13%39%
Core, in the order it usually comes out
min 1380%44%-15.3 pp
min 1431%40%-13.8 pp
min 1531%40%-14.6 pp
min 1746%48%
min 1765%49%-13.8 pp
min 2342%50%
min 2740%
min 1934%44%-13.3 pp
min 2249%53%
min 2249%53%
Situational
min 793%45%-13.2 pp
min 1882%
min 2556%46%-13.4 pp
min 1150%48%
min 1430%40%-13.1 pp
min 1830%47%-12.7 pp
min 1530%39%-14.1 pp
min 2930%56%
min 3223%62%
min 3619%66%
min 3217%58%
min 3617%65%
Did this build work for you?
The numbers come from the top of the ladder. What they do not say is whether the build holds up in the games played here. Your answer is what adjusts that.
Trend in the 7500+ ladder
311 ranked matches in the last 8 weeks
- Win rate
- 33.4%
- Pick rate
- 1.9%
Rating: 70% win rate (adjusted for sample size) and 30% pick rate, compared with heroes in the same position. S is the top 10%, A the next 20%, B 30%, C 25%, D the rest.
Shadow Fiend worldwide
OpenDota, public ranked matches from the last 7 days
Win rate
43.6%
342,528 matches
Pick rate
9.7%
of matches include the hero
Pro scene
2 picks and 0 bans
50.0% win rate when picked
Win rate by rank bracket
- Herald and Guardian45.2% · 35,735 matches
- Crusader and Archon44.2% · 124,291 matches
- Legend and Ancient42.2% · 116,879 matches
- Divine and Immortal40.4% · 19,830 matches
Win rate by match duration
- up to 25 min55.7% · 5,768 matches
- 25 to 35 min50.3% · 14,855 matches
- 35 to 45 min49.9% · 6,777 matches
- over 45 min49.7% · 3,573 matches
Advantage in points over Shadow Fiend's usual win rate: negative means trouble. World data comes first (STRATZ and OpenDota); our high-MMR matches take over once the sample is big enough.
Worst counters
Wraith King36.4% win rate in 6,887 games · STRATZ-7.0
Windranger37.5% win rate in 112 games · our high-MMR matches-6.9
Dark Seer37.5% win rate in 104 games · our high-MMR matches-6.9
Winter Wyvern39.2% win rate in 148 games · our high-MMR matches-5.2
Dragon Knight38.5% win rate in 7,526 games · STRATZin our high-MMR matches: 46.4% in 69 games-4.9
Phantom Lancer39.0% win rate in 6,492 games · STRATZin our high-MMR matches: 44.0% in 84 games-4.4
Spirit Breaker39.7% win rate in 8,382 games · STRATZin our high-MMR matches: 41.5% in 94 games-3.7
Outworld Destroyer39.8% win rate in 6,362 games · STRATZin our high-MMR matches: 43.7% in 71 games-3.6
Best synergies
Bounty Hunter56.3% win rate in 119 games · our high-MMR matches+11.9
Wraith King50.2% win rate in 6,228 games · STRATZ+6.8
Ember Spirit51.0% win rate in 104 games · our high-MMR matches+6.5
Phantom Lancer49.3% win rate in 5,334 games · STRATZ+5.9
Visage48.7% win rate in 676 games · STRATZ+5.3
Chaos Knight48.0% win rate in 2,039 games · STRATZ+4.6
Winter Wyvern48.7% win rate in 158 games · our high-MMR matches+4.3
Riki47.4% win rate in 2,016 games · STRATZ+4.0
Shadow Fiend: the power spike only counts if it becomes Roshan
Landing three Razes in a row is the part every Shadow Fiend player practices. Across the hero's 18 lessons in the catalog, almost no game was lost to a missed Raze. They were lost before or after it: in the mana wasted at minute 1, and in the minute when SF, at his peak, picked the wrong objective. Shadow Fiend is a window hero. He gets very strong for a few minutes, and the game depends on what he does with those minutes.
Level 1 mana is the level 3 kill
In the review of a Herald 3 SF, the student leaves base without Magic Wand or Iron Branch, trades Razes with Axe and spends 240 mana to take HP the enemy regains in seconds, while Axe spent 80. When the kill shows up, he is dry. The coach's fix is simple: at levels 1 and 2, Raze is for securing the ranged creep, not for trading. The damage climbs a lot with each level, and level 3 is where it kills.
The same idea shows up at the pro level: in Parivision against Aurora, the coach shows Satanic saving his Raze mana until level 3 against a Timbersaw.
Two details round out the lane:
- Lotus at minute 3. Pull the wave around 2:30 so you are next to the Lotus when it spawns. "Please, take the Lotus every single game of your life."
- Falcon Blade as the lane item, because it fixes the lack of HP and mana regen, which is what holds SF back in the first ten minutes.
Same hero, different mistakes by bracket
| Bracket | Mistake the coach corrected most | What to do |
|---|---|---|
| Herald | Razes on heroes at levels 1 and 2, no mana when the kill comes | Raze the ranged creep, save mana for level 3 |
| Legend | Wipe the enemy team and hit tier 2 without Aegis instead of Roshan | Roshan first, tier 2 with Aegis |
| Unranked | Tormentor instead of Roshan, three minutes of peak wasted | The strongest objective your timing allows |
| Ancient | Trying too much without items, or playing scared with items | Farm until the item and then play with the team |
| Immortal | Late BKB against Mars and Blade Mail | BKB as third item when the game asks for it |
A wipe is not an objective, Roshan is
The mistake that costs SF the most games does not happen in lane. In the review of a Legend 3 SF, the student wipes the enemy team and, instead of going straight to Roshan, goes to hit tier 2. Without Aegis the play does not close, and the coach is curt: "one simple wrong decision you made cost you this game".
In an unranked game, the same pattern with another face: SF finishes Manta, he is at his power spike, and he goes to do Tormentor. "You lost the 3 minutes of power spike when you were strong, to do a Tormentor." Instead of that, Roshan and two tier 2 towers. And SF is the one who should carry the Aegis: he is the hero on the team who hits buildings.
The contrast comes from Yatoro. In Team Spirit against Nigma Galaxy, every finished item turns into something on the map: a tower, a kill, a Roshan. He loses his tier 1, farms the safe side, finishes Manta, calls the team with pings and turns the timing into an objective. Nothing genius, just the right question at every item: what can I do now that I couldn't do a minute ago?
Doing too much and doing too little
At Ancient, SF fails in two opposite ways, and sometimes the same player alternates between them.
In the review of an Ancient 4 SF, the coach sums up the game: "90% of your deaths in this game so far are you trying to do more than you should". The student pressures towers without vision, tries Roshan without items, TPs into fights his team would win without him. Each of those plays takes farm away from a hero who has not reached his timing yet. And the answer against that game's Faceless Void was different: if the game goes long, Void farms six items, so SF wins by taking towers early.
At Ancient 5, the problem is the opposite. The player is afraid to step on the map because Kunkka, Earthshaker and Disruptor can kill him, shows up late to fights and does not use the Tusk who is there precisely to save him. "You don't use your team, you think the lobby is a 1v5."
The way out of both is the same: farm the triangle and safe waves until the item is done, and then play with the team, neither alone up front nor hiding behind.
Items answer the game, not the build
In the Ancient reviews, the item discussion keeps coming back to Manta, BKB and Hurricane Pike. Manta gives a dispel and speeds up farm; BKB is what keeps SF alive against control. Against Magnus, Hurricane Pike was mandatory. Against Kunkka and Disruptor, in another replay, it added no real mobility. Pike is neither good nor bad: it is good when it solves the hero who catches you.
The question before going back to the jungle
When the item is done, before going back to farm, ask: which objective does this item let me take right now? A leftover tier 1, a Roshan with smoke, a tier 2 with Aegis. If the answer is none, farm more. If it is any of them, call your team. The SF who farms well and does not convert loses to the enemy carry who farms slower but uses his own timings.
Locked on exactly what you wanted?
The whole catalog (real match analysis, by hero, position and patch) opens with the subscription.
See plansBuild of the players who play Shadow Fiend the most
7500+ ranked matches our radar captured in the last 60 days: accounts with at least 10 matches on the hero, and the build from their matches only. This is not the pro league: that one is in the hero's pros.
4 specialists, 56 matches, 46% win rate on the hero
Items
Starting kits
2


1155%
2


2
967%
2


850%
Starting items
96%48%
93%48%
91%51%
91%49%
43%54%
39%41%
30%47%
7%
Core, in the order it usually comes out
min 1391%49%
min 1384%49%
min 1582%50%
min 1882%50%
min 2530%65%
min 2963%63%
Situational
min 684%51%
min 2668%55%
min 2230%65%
min 3723%62%
min 4221%58%
min 4221%58%
min 3218%60%
min 3814%
Skill order
Most played17 matches, 47% win rate
1
2
3
4
5
6
7
8
9
10- 11
12
13
14
15- 16
18- 20
- 25
Highest win rate3 matches, 100% win rate
1
2
3
4
5
6
7
8
9
10- 11
12
13
14
15- 16
18- 20
- 25
Who plays Shadow Fiend in pro Dota
Pro leagues since patch 7.33. Recent matches also include 7500+ ranked games our radar watches live.
20 matches shown, 40% win rate
10 min 20 min NW 4.4k 10.7k LH 68 177 DN 18 Lane +1.0k 









3






1 min
3 min
4 min
4 min
7 min
10 min
14 min
14 min
17 min
23 min
27 min
10 min 20 min NW 4.4k 12.3k LH 78 200 DN 2 Lane -0.3k 









2
2





1 min
3 min
4 min
6 min
6 min
11 min
11 min
15 min
15 min
18 min
23 min
27 min
30 min









- +
- +

- +
- +
- +
- +
- +
- +





10 min 20 min NW 4.4k 10.3k LH 71 199 DN 7 Lane +0.3k 










2




1 min
3 min
4 min
5 min
7 min
11 min
15 min
15 min
20 min
26 min
10 min 20 min NW 3.8k - LH 53 - DN 6 Lane -0.5k 









2
2




1 min
3 min
4 min
6 min
7 min
15 min
10 min 20 min NW 3.6k 8.7k LH 66 165 DN 2 Lane +0.4k 









3





2 min
4 min
6 min
9 min
10 min
12 min
17 min
21 min
23 min
24 min
29 min
32 min
37 min
10 min 20 min NW 4.1k 10.2k LH 73 180 DN 13 Lane +0.8k 












2



1 min
3 min
5 min
7 min
12 min
15 min
17 min
18 min
22 min
10 min 20 min NW 5.0k 14.1k LH 86 277 DN 7 Lane +0.7k 











2


3 min
5 min
5 min
7 min
10 min
13 min
16 min
17 min
20 min
21 min
25 min
27 min
35 min
39 min
10 min 20 min NW 3.3k 9.7k LH 88 214 DN 2 Lane -1.1k 










2





2 min
3 min
4 min
8 min
9 min
11 min
13 min
16 min
16 min
19 min
25 min
32 min
37 min









- +
- +

- +
- +
- +
- +
- +





10 min 20 min NW 3.4k 7.7k LH 48 111 DN 2 Lane -0.4k 









2
2





1 min
3 min
4 min
5 min
7 min
12 min
18 min
18 min









- +
- +

10 min 20 min NW 4.0k 5.0k LH 70 70 DN 12 Lane +1.1k 









2
2





1 min
3 min
6 min
7 min
10 min 20 min NW 4.5k 11.6k LH 69 180 DN 6 Lane -0.4k 













2


1 min
2 min
3 min
5 min
7 min
7 min
11 min
16 min
16 min
18 min
23 min
27 min
34 min
42 min
42 min









- +
- +

- +
- +
- +
- +
- +





10 min 20 min NW 4.5k 9.5k LH 57 163 DN 14 Lane +1.6k 









2
2




1 min
3 min
7 min
7 min
8 min
12 min
17 min
17 min
20 min
25 min
30 min
33 min









- +
- +

- +
- +
- +
- +
- +





10 min 20 min NW 3.2k 8.6k LH 50 147 DN 10 Lane -0.1k 










3
4



4 min
6 min
7 min
11 min
17 min
17 min
19 min
32 min









- +
- +

- +
- +
- +
- +
10 min 20 min NW 3.9k 10.8k LH 62 214 DN 14 Lane +0.8k 









3
4



2 min
5 min
6 min
11 min
14 min
14 min
18 min
23 min
31 min
34 min
37 min
47 min









- +
- +

- +
- +
- +
- +
- +





10 min 20 min NW 3.1k - LH 37 - DN 8 Lane -2.3k 















2 min
3 min
5 min
6 min
10 min 20 min NW 1.9k 3.2k LH 24 24 DN 3 Lane +0.1k 









2





3 min
6 min
10 min 20 min NW 3.8k 10.5k LH 63 169 DN 8 Lane +1.1k 









2






1 min
3 min
4 min
7 min
8 min
12 min
14 min
14 min
19 min
25 min
10 min 20 min NW 4.9k 11.0k LH 87 187 DN 4 Lane +0.8k 









2






1 min
3 min
3 min
6 min
8 min
8 min
11 min
15 min
15 min
17 min
24 min
29 min
40 min
42 min
10 min 20 min NW 4.7k 9.8k LH 66 155 DN 7 Lane +2.0k 









4





1 min
3 min
4 min
4 min
6 min
7 min
10 min
15 min
15 min
20 min
25 min
33 min
35 min
40 min
40 min
10 min 20 min NW 2.9k 8.5k LH 52 170 DN 4 Lane +0.0k 

















2 min
6 min
12 min
16 min
16 min
17 min
22 min
30 min
Pro league and 7500+ ranked (seen live by our radar), newest first. In the tree, the lit branch is the talent the player took. In ranked, last hits and denies come from the replay and show up a few hours after the match. Position is estimated from gold per minute within the team.
Patch
Position
Tournaments
Minimum matches
With the item at the end of the match
| # | Player | W x L | Last | |||
|---|---|---|---|---|---|---|
| 1 | jabjik | 35 | 27 x 8 | 77.1% | 63.3% | 10/22/2025 |
| 2 | bOzO | 22 | 18 x 4 | 81.8% | 62.0% | 05/02/2025 |
| 3 | Renarr | 19 | 16 x 3 | 84.2% | 61.8% | 05/03/2024 |
| 4 | Yatoro | 25 | 19 x 6 | 76.0% | 60.7% | 08/23/2026 |
| 5 | m1CKe | 46 | 31 x 15 | 67.4% | 59.8% | 08/22/2026 |
| 6 | Yathest | 21 | 16 x 5 | 76.2% | 59.7% | 06/17/2024 |
| 7 | evis ( choucha prime ) | 28 | 20 x 8 | 71.4% | 59.5% | 10/21/2025 |
| 8 | swap LIKES :3 | 18 | 14 x 4 | 77.8% | 59.4% | 08/23/2025 |
| 9 | scream | 72 | 46 x 26 | 63.9% | 59.3% | 09/19/2025 |
| 10 | PiPi | 15 | 12 x 3 | 80.0% | 59.0% | 09/25/2025 |
| 11 | kiyotaka | 22 | 16 x 6 | 72.7% | 58.8% | 08/10/2026 |
| 12 | BOCKPEC_U_PA3HEC_/7A6JIUK | 17 | 13 x 4 | 76.5% | 58.7% | 07/22/2025 |
| 13 | ryukzak | 17 | 13 x 4 | 76.5% | 58.7% | 08/14/2024 |
| 14 | PICK STUNS | 16 | 12 x 4 | 75.0% | 57.9% | 08/02/2025 |
| 15 | ɹǝʞɹɐd⚔️ | 49 | 31 x 18 | 63.3% | 57.8% | 10/22/2026 |
| 16 | Satanic | 44 | 28 x 16 | 63.6% | 57.7% | 09/30/2026 |
| 17 | 驚き | 15 | 11 x 4 | 73.3% | 57.2% | 08/31/2025 |
| 18 | noob123 | 92 | 55 x 37 | 59.8% | 57.2% | 10/10/2025 |
| 19 | so? | 63 | 38 x 25 | 60.3% | 56.7% | 11/12/2024 |
| 20 | RE.Terax | 42 | 26 x 16 | 61.9% | 56.6% | 02/05/2024 |
| 21 | MidOne | 16 | 11 x 5 | 68.8% | 56.2% | 03/22/2026 |
| 22 | Nightfall | 27 | 17 x 10 | 63.0% | 55.9% | 08/14/2026 |
| 23 | young G | 20 | 13 x 7 | 65.0% | 55.7% | 08/28/2026 |
| 24 | Xm | 22 | 14 x 8 | 63.6% | 55.6% | 08/15/2026 |
| 25 | Timado | 22 | 14 x 8 | 63.6% | 55.6% | 09/12/2026 |
Adjusted pulls the win rate of players with few matches toward the hero's average in these filters (51.1%): 14 wins in 16 does not beat 92 in 134. Position is estimated from gold per minute within the team.