Team groups, counters and drafting advice for Supreme League and Guild Duel.
Everything here, including the written insights, was produced by an automated analysis of saved
tournament results and hero skill descriptions. No person wrote it, and no replays were reviewed, so
explanations of wins and losses have not been checked against what happened during a battle, such as
targeting or ability timing. The records primarily reflect top-spending players at the top end of the
meta, with heavily upgraded rosters. The interpretations can be wrong, and results may differ with
fewer upgrades. More about the data's limits.
A dotted-underlined record or count opens examples; a match or map link opens that match. This version leaves out player names.
Data as of 2026-09-19 · records created 2026-07-08 to 2026-09-19
A team group contains related five-hero lineups that share heroes. Its name identifies the core,
not a fixed team of five. heroes appear most often; heroes vary between lineups.
Icon percentages show how often each hero appears. Every hero meeting the core or flex threshold is shown.
Groups are sorted by strength tier, then by appearances. Strength
estimates performance against an average team with equal upgrades, accounting for the opponents faced and recorded upgrade gaps.
The full ratings include the unadjusted comparison.
excludes ties. The bar below it shows the 95% interval, with a dot for the win rate and a tick at 50%. A wider segment means less certainty; it does not account for differences in players, upgrades or repeated teams. shows how often the most popular artifact is equipped.
Team
Gwyn/Gunnar/KuluCore100%99%84%Flex75%31%26%Artifact55% useAppearances429278–151Win rate65%Players93StrengthTier 172% vs averageEvidence behind these results
429 appearances from 93 players and 345 team selections.
If SL rules require a player to reuse the same teams, their team on each map counts as one selection across those matches. Ro4, Finals and Guild Duel allow changes between matches, so each match counts separately. These selections can still share a strategy.
The most frequent player accounts for
43 appearances (10%).
Other players together went 238–148.
With similar upgrades, the record is 129–69.
These checks do not remove differences in opponents, maps or player skill.
The group's usual lineup, placed as it was most often on SL-Map5:
of this exact placement,
13% of the group's 106 boards there.
How the team and its percentages are chosen.
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. Rows can overlap, so do not add their appearances together.
Heroes
Players
Aliceth
19
Cyran
17
Frieren
9
Igor
8
Contess
8
Taichi & Agumon
12
Peggy ·
3
Mehira ·
5
Aurora ·
3
Athalia ·
2
Eryndor ·
3
Callan ·
–
1
Cassadee ·
–
1
Rolan ·
2
Ravion ·
–
1
Players
CyranKulu ·
5
CyranFrieren ·
–
1
AlicethAthalia ·
1
AlicethCyran ·
–
2
KuluMehira ·
1
AthaliaDunlingr ·
1
AlicethKulu ·
1
FrierenPeggy ·
–
1
KuluZanie ·
1
Other recorded lineups
Exact five-hero teams outside the variations above, including uncommon lineups and those with poor results.
Lineup
Players
·
5
·
–
4
·
3
·
4
·
1
·
–
1
·
–
1
·
1
·
1
·
2
·
1
·
1
·
1
·
3
·
1
·
–
1
·
2
·
1
·
1
·
–
1
·
–
1
·
1
·
–
1
·
2
·
1
·
1
·
–
1
·
–
1
·
1
·
–
1
·
–
1
·
–
1
·
–
1
·
1
·
–
1
·
1
·
–
1
·
1
·
–
1
·
–
1
·
–
1
·
–
1
Hero results across this team group
Overall results with and without each supporting hero, across all variations. Teammates, maps, players and opponents can differ; this does not measure the effect of swapping one hero alone.
Hero
Dunlingr
Aliceth
Cyran
Evie/OrionCore98%97%Flex72%57%53%39%31%Artifact69% useAppearances308160–146Win rate52%Players90StrengthTier 166% vs averageEvidence behind these results
308 appearances from 90 players and 252 team selections.
If SL rules require a player to reuse the same teams, their team on each map counts as one selection across those matches. Ro4, Finals and Guild Duel allow changes between matches, so each match counts separately. These selections can still share a strategy.
The most frequent player accounts for
30 appearances (10%).
Other players together went 142–135.
With similar upgrades, the record is 47–51.
These checks do not remove differences in opponents, maps or player skill.
The group's usual lineup, placed as it was most often on SL-Map1:
of this exact placement,
8% of the group's 105 boards there (28 of them Guild Duel games on the same ground).
How the team and its percentages are chosen.
Core
Evie 98%, Orion 97%
Artifacts
Valorshield · 69% of teams 57% win rate (121–90)
Frontline · 17% of teams 43% win rate (22–29)
Breakthrough · 6% of teams 32% win rate (6–13)
RolanSupportplaced here 17% · on 71% of boards
ZandrokWarriorplaced here 69% · on 33% of boards
EvieSupportplaced here 10%
PerseusWarriorplaced here 16% · on 61% of boards
OrionWarriorplaced here 66%
PAurelianPhantimalplaced here 15%
Beats: Zorya/Athalia/Pandora, Alna/Syph/Mehira
Loses to: Gwyn/Gunnar/Kulu, Saida/Contess/Twins, Rolan/Twins
Common variations
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. Rows can overlap, so do not add their appearances together.
Heroes
Players
Zandrok ·
10
Rowan ·
7
Rolan
10
Hepler ·
3
Taichi & Agumon ·
4
Kazim ·
–
3
Ravion ·
2
Alna ·
–
1
Contess ·
1
Cyran ·
–
1
Lily May ·
–
1
Players
Kordan ·
–
1
Zandrok ·
1
Gerda ·
2
Perseus ·
1
Other recorded lineups
Exact five-hero teams outside the variations above, including uncommon lineups and those with poor results.
Lineup
Players
12
17
8
·
7
·
6
·
4
·
–
1
·
–
1
·
1
·
1
·
1
·
–
2
·
1
·
–
1
·
2
·
–
1
·
1
·
1
·
2
·
2
·
–
1
·
–
1
·
1
·
–
1
·
–
1
·
1
·
1
·
–
1
·
1
·
–
1
·
1
·
–
1
·
–
1
·
1
·
1
·
–
1
·
–
1
·
–
1
·
–
1
·
–
1
Hero results across this team group
Overall results with and without each supporting hero, across all variations. Teammates, maps, players and opponents can differ; this does not measure the effect of swapping one hero alone.
Hero
Perseus
Frieren
Rolan
Zandrok
Rowan
Alna/Syph/MehiraCore99%99%93%Flex50%33%31%19%17%Artifact87% useAppearances292163–128Win rate56%Players69StrengthTier 165% vs averageEvidence behind these results
292 appearances from 69 players and 244 team selections.
If SL rules require a player to reuse the same teams, their team on each map counts as one selection across those matches. Ro4, Finals and Guild Duel allow changes between matches, so each match counts separately. These selections can still share a strategy.
The most frequent player accounts for
30 appearances (10%).
Other players together went 138–124.
With similar upgrades, the record is 78–62.
These checks do not remove differences in opponents, maps or player skill.
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. Rows can overlap, so do not add their appearances together.
Heroes
Players
HeplerZandrok
15
CyranHepler
15
DunlingrZandrok
7
AlicethZandrok ·
3
FrierenZandrok ·
3
AlicethCyran ·
3
CyranZandrok ·
6
AthaliaFrieren ·
2
EvieZandrok ·
4
AthaliaZandrok ·
3
AthaliaHepler ·
6
CyranEryndor ·
4
CyranDunlingr ·
–
2
AlicethAthalia ·
3
FrierenHepler ·
3
CyranFrieren ·
2
Lily MayZandrok ·
1
NiruSatrana ·
1
AthaliaNiru ·
1
CyranLily May ·
1
CyranNiru ·
1
EryndorThador ·
3
HeplerKordan ·
–
1
RolanZandrok ·
2
Taichi & AgumonZandrok ·
3
VelaraZandrok ·
1
ContessHepler ·
1
ContessZandrok ·
2
EvieHepler ·
1
HeplerSatrana ·
1
PeggyZandrok ·
1
AlicethVoracia ·
1
AthaliaCyran ·
1
AthaliaGalahad ·
1
AthaliaRolan ·
1
AthaliaZorya ·
1
DunlingrFrieren ·
1
EryndorRavion ·
1
FrierenSolise ·
1
GerdaZandrok ·
1
GwynethHepler ·
–
1
HeplerPeggy ·
1
HeplerSolise ·
1
NiruZandrok ·
1
SoliseZandrok ·
–
1
Other recorded lineups
Exact five-hero teams outside the variations above, including uncommon lineups and those with poor results.
Lineup
Players
·
1
·
2
·
–
1
·
2
·
–
1
·
–
1
·
1
·
1
·
1
·
–
1
·
1
·
1
·
1
·
–
1
·
–
1
·
1
Hero results across this team group
Overall results with and without each supporting hero, across all variations. Teammates, maps, players and opponents can differ; this does not measure the effect of swapping one hero alone.
Hero
Zandrok
Hepler
Cyran
Athalia
Frieren
Rolan/TwinsCore100%100%Flex67%49%46%46%26%Artifact48% useAppearances6145–16Win rate74%Players17StrengthTier 176% vs averageEvidence behind these results
61 appearances from 17 players and 40 team selections.
If SL rules require a player to reuse the same teams, their team on each map counts as one selection across those matches. Ro4, Finals and Guild Duel allow changes between matches, so each match counts separately. These selections can still share a strategy.
The most frequent player accounts for
16 appearances (26%).
Other players together went 31–14.
With similar upgrades, the record is 8–5.
These checks do not remove differences in opponents, maps or player skill.
Taichi & AgumonWarriorplaced here 86% · on 41% of boards
MehiraMageplaced here 35%
Elijah & LailahSupportplaced here 100%
RolanSupportplaced here 94%
SylphiraRogueplaced here 100%
Elijah & LailahCompanionplaced here 100%
PMidnight HunterPhantimalplaced here 100%
Beats: Evie/Orion
Loses to: Gwyn/Gunnar/Kulu
Common variations
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. Rows can overlap, so do not add their appearances together.
Heroes
Players
Taichi & Agumon ·
4
Kazim ·
1
Aliceth ·
2
Zandrok ·
2
Gerda ·
2
Athalia ·
–
1
Players
Athalia ·
–
3
Sylphira ·
2
Yamato & Gabumon ·
1
Frieren ·
–
1
Kulu ·
–
1
Other recorded lineups
Exact five-hero teams outside the variations above, including uncommon lineups and those with poor results.
Lineup
Players
·
1
·
–
1
·
2
·
–
2
·
–
1
·
–
1
·
–
1
Hero results across this team group
Overall results with and without each supporting hero, across all variations. Teammates, maps, players and opponents can differ; this does not measure the effect of swapping one hero alone.
Hero
Sylphira
Taichi & Agumon
Contess
Mehira
Athalia
Saida/Contess/TwinsCore100%97%96%Flex40%40%26%19%Artifact29% useAppearances299151–130Win rate54%Players83StrengthTier 260% vs averageEvidence behind these results
299 appearances from 83 players and 249 team selections.
If SL rules require a player to reuse the same teams, their team on each map counts as one selection across those matches. Ro4, Finals and Guild Duel allow changes between matches, so each match counts separately. These selections can still share a strategy.
The most frequent player accounts for
19 appearances (6%).
Other players together went 139–123.
With similar upgrades, the record is 61–57.
These checks do not remove differences in opponents, maps or player skill.
Elijah & LailahSupportplaced here 83% · on 86% of boards
ContessSupportplaced here 79%
GerdaTankplaced here 38% · on 75% of boards
AlicethMarksmanplaced here 50% · on 79% of boards
SaidaRogueplaced here 81%
Elijah & LailahCompanionplaced here 83%
PMidnight HunterPhantimalplaced here 18%
Beats: Evie/Orion
Loses to: Gwyn/Gunnar/Kulu
Common variations
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. Rows can overlap, so do not add their appearances together.
Heroes
Players
AlicethGerda
10
AlicethRolan
11
GerdaSolise
20
GerdaRowan
8
AlicethEvie
7
PangSolise ·
2
AlicethHepler ·
3
AthaliaPang ·
–
1
BrutusHepler ·
3
GerdaRolan ·
2
HeplerSolise ·
–
2
AthaliaGerda ·
2
GerdaHepler ·
3
AlnaSolise ·
2
AlicethScarlita ·
–
1
RolanSolise ·
–
2
AlicethLily May ·
–
2
AlicethRowan ·
–
1
AthaliaEvie ·
3
AthaliaLily May ·
1
AthaliaShadewing ·
–
1
AthaliaSolise ·
1
BaelranTalene ·
1
CallanRolan ·
2
CallanTalene ·
–
1
EvieGerda ·
3
GerdaPang ·
1
AlicethPang ·
2
AlnaAthalia ·
–
1
DunlingrGerda ·
–
1
EvieSolise ·
–
2
KuluSolise ·
1
Lily MayRolan ·
1
Lily MayScarlita ·
1
AlicethBaelran ·
–
–
1
AthaliaRolan ·
–
1
AuroraRolan ·
–
1
BaelranRolan ·
1
BrutusCyran ·
–
1
CallanSolise ·
1
CyranPang ·
1
DunlingrSolise ·
–
1
GerdaLily May ·
1
GerdaScarlita ·
–
1
HeplerRowan ·
–
1
Lily MaySolise ·
–
1
LudovicRowan ·
–
1
PeggyRowan ·
1
RavionSolise ·
1
Players
Gerda
20
Pang ·
2
Hepler ·
–
2
Alna ·
2
Rolan ·
–
2
Athalia ·
1
Evie ·
–
2
Kulu ·
1
Callan ·
1
Dunlingr ·
–
1
Lily May ·
–
1
Ravion ·
1
Other recorded lineups
Exact five-hero teams outside the variations above, including uncommon lineups and those with poor results.
Lineup
Players
·
2
·
1
·
1
·
1
·
–
1
·
1
·
–
1
·
1
·
1
·
–
1
·
1
·
–
1
·
–
1
·
–
1
·
1
Hero results across this team group
Overall results with and without each supporting hero, across all variations. Teammates, maps, players and opponents can differ; this does not measure the effect of swapping one hero alone.
Hero
Gerda
Aliceth
Solise
Rolan
Zorya/Athalia/PandoraCore100%95%94%Flex72%70%Artifact57% useAppearances12744–83Win rate35%Players51StrengthTier 246% vs averageEvidence behind these results
127 appearances from 51 players and 102 team selections.
If SL rules require a player to reuse the same teams, their team on each map counts as one selection across those matches. Ro4, Finals and Guild Duel allow changes between matches, so each match counts separately. These selections can still share a strategy.
The most frequent player accounts for
16 appearances (13%).
Other players together went 36–75.
With similar upgrades, the record is 15–27.
These checks do not remove differences in opponents, maps or player skill.
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. Rows can overlap, so do not add their appearances together.
Heroes
Other recorded lineups
Exact five-hero teams outside the variations above, including uncommon lineups and those with poor results.
Lineup
Players
·
5
4
·
4
·
1
·
1
·
–
1
·
1
·
–
2
·
–
2
·
–
2
·
–
2
·
–
1
·
–
1
·
1
·
–
1
·
–
1
·
–
1
·
1
·
–
1
·
–
1
·
–
1
·
1
Hero results across this team group
Overall results with and without each supporting hero, across all variations. Teammates, maps, players and opponents can differ; this does not measure the effect of swapping one hero alone.
Hero
Shadewing
Callan
Athalia/Alna/EryndorCore97%91%89%Flex77%57%43%20%Artifact51% useAppearances3517–18Win rate49%Players13StrengthTier 262% vs averageEvidence behind these results
35 appearances from 13 players and 20 team selections.
If SL rules require a player to reuse the same teams, their team on each map counts as one selection across those matches. Ro4, Finals and Guild Duel allow changes between matches, so each match counts separately. These selections can still share a strategy.
The most frequent player accounts for
7 appearances (20%).
Other players together went 13–15.
With similar upgrades, the record is 6–1.
These checks do not remove differences in opponents, maps or player skill.
This group of similar teams still needs review; some teams may belong in another group.
The group's usual lineup, placed as it was most often on SL-Map3 (SL4–5):
of this exact placement,
40% of the group's 20 boards there.
How the team and its percentages are chosen.
Core
Athalia 97%, Alna 91%, Eryndor 89%
Artifacts
Magicsurge · 51% of teams 50% win rate (9–9)
Surging · 40% of teams 57% win rate (8–6)
Vine Snare · 6% of teams 0% win rate (0–2)
AlnaTankplaced here 45%
EryndorMageplaced here 40%
ShadewingRogueplaced here 69% · on 65% of boards
IgorWarriorplaced here 89% · on 45% of boards
AthaliaRogueplaced here 95%
PNecrodrakonPhantimalplaced here 40%
Common variations
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. Rows can overlap, so do not add their appearances together.
Heroes
Players
Igor ·
5
Zorya ·
2
Other recorded lineups
Exact five-hero teams outside the variations above, including uncommon lineups and those with poor results.
Lineup
Players
·
2
·
–
1
·
–
1
·
1
·
–
1
·
–
1
Hero results across this team group
Overall results with and without each supporting hero, across all variations. Teammates, maps, players and opponents can differ; this does not measure the effect of swapping one hero alone.
Hero
Shadewing
Igor
Zorya
Mehira
Gwyn/Cyran/HuginCore100%88%84%Flex66%34%19%19%16%16%Artifact47% useAppearances327–25Win rate22%Players25StrengthTier 231% vs averageEvidence behind these results
32 appearances from 25 players and 32 team selections.
If SL rules require a player to reuse the same teams, their team on each map counts as one selection across those matches. Ro4, Finals and Guild Duel allow changes between matches, so each match counts separately. These selections can still share a strategy.
The most frequent player accounts for
3 appearances (9%).
Other players together went 5–24.
With similar upgrades, the record is 1–5.
These checks do not remove differences in opponents, maps or player skill.
This group of similar teams still needs review; some teams may belong in another group.
The group's usual lineup, placed as it was most often on SL-Map3 (SL4–5):
of this exact placement,
33% of the group's 6 boards there.
How the team and its percentages are chosen.
Core
Gwyneth 100%, Cyran 88%, Hugin 84%
Artifacts
Wing Guard · 47% of teams 13% win rate (2–13)
Vanguard · 44% of teams 29% win rate (4–10)
Vine Snare · 3% of teams 0% win rate (0–1)
GwynethMarksmanplaced here 33%
CyranMageplaced here 80% · on 83% of boards
GunnarTankplaced here 33%
HuginSupportplaced here 100% · on 83% of boards
AlnaTankplaced here 67% · on 50% of boards
PAurelianPhantimalplaced here 50%
Common variations
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. Rows can overlap, so do not add their appearances together.
Heroes
Players
Alna ·
–
3
Aliceth ·
2
Callan ·
–
2
Contess ·
–
1
Ravion ·
–
1
Rolan ·
–
1
Zanie ·
–
1
Other recorded lineups
Exact five-hero teams outside the variations above, including uncommon lineups and those with poor results.
Lineup
Players
·
2
·
1
·
2
·
–
1
·
–
1
·
–
1
·
1
·
–
1
·
–
1
·
–
1
·
–
1
·
–
1
·
1
·
–
1
·
–
1
Hero results across this team group
Overall results with and without each supporting hero, across all variations. Teammates, maps, players and opponents can differ; this does not measure the effect of swapping one hero alone.
Hero
Gunnar
Alna
Callan
Zanie
Igor
Ravion
Marcille/Ravion/ThadorCore97%97%93%Flex57%40%33%33%20%Artifact30% useAppearances309–21Win rate30%Players13StrengthTier 231% vs averageEvidence behind these results
30 appearances from 13 players and 23 team selections.
If SL rules require a player to reuse the same teams, their team on each map counts as one selection across those matches. Ro4, Finals and Guild Duel allow changes between matches, so each match counts separately. These selections can still share a strategy.
The most frequent player accounts for
10 appearances (33%).
Other players together went 7–13.
With similar upgrades, the record is 5–8.
These checks do not remove differences in opponents, maps or player skill.
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. Rows can overlap, so do not add their appearances together.
Heroes
Players
Solise ·
–
1
Pang ·
–
1
Aliceth ·
–
1
Alna ·
–
1
Dunlingr ·
1
Lyca ·
–
1
Perseus ·
1
Players
Perseus ·
2
Dunlingr ·
3
Lily May ·
–
1
Other recorded lineups
Exact five-hero teams outside the variations above, including uncommon lineups and those with poor results.
Lineup
Players
·
–
1
·
1
·
1
·
1
·
1
·
–
1
Hero results across this team group
Overall results with and without each supporting hero, across all variations. Teammates, maps, players and opponents can differ; this does not measure the effect of swapping one hero alone.
Hero
Lily May
Solise
Dunlingr
Lyca
Perseus
Alna/AthaliaCore100%100%Flex76%62%33%33%29%Artifact67% useAppearances219–12Win rate43%Players6StrengthTier 252% vs averageEvidence behind these results
21 appearances from 6 players and 19 team selections.
If SL rules require a player to reuse the same teams, their team on each map counts as one selection across those matches. Ro4, Finals and Guild Duel allow changes between matches, so each match counts separately. These selections can still share a strategy.
The most frequent player accounts for
12 appearances (57%).
Other players together went 3–6.
With similar upgrades, the record is 2–5.
These checks do not remove differences in opponents, maps or player skill.
This group of similar teams still needs review; some teams may belong in another group.
The group's usual lineup, placed as it was most often on SL-Map3 (SL4–5):
of this exact placement,
67% of the group's 9 boards there.
How the team and its percentages are chosen.
Core
Alna 100%, Athalia 100%
Artifacts
Frontline · 67% of teams 29% win rate (4–10)
Breakthrough · 29% of teams 83% win rate (5–1)
Vanguard · 5% of teams 0% win rate (0–1)
MehiraMageplaced here 67%
AlnaTankplaced here 67%
HeplerTankplaced here 100% · on 67% of boards
Taichi & AgumonWarriorplaced here 78%
AthaliaRogueplaced here 89%
PMidnight HunterPhantimalplaced here 78%
Common variations
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. Rows can overlap, so do not add their appearances together.
No required-hero combination has eight recorded appearances. The lineups below include every recorded team.
Other recorded lineups
Exact five-hero teams outside the variations above, including uncommon lineups and those with poor results.
Lineup
Players
·
2
·
–
1
·
1
·
1
·
1
·
–
1
·
1
Hero results across this team group
Overall results with and without each supporting hero, across all variations. Teammates, maps, players and opponents can differ; this does not measure the effect of swapping one hero alone.
Hero
Taichi & Agumon
Mehira
Hepler
Rolan
Sylphira
Eryndor/LM/ThadorCore100%100%100%Flex63%42%32%16%16%Artifact47% useAppearances194–15Win rate21%Players8StrengthTier 236% vs averageEvidence behind these results
19 appearances from 8 players and 11 team selections.
If SL rules require a player to reuse the same teams, their team on each map counts as one selection across those matches. Ro4, Finals and Guild Duel allow changes between matches, so each match counts separately. These selections can still share a strategy.
The most frequent player accounts for
5 appearances (26%).
Other players together went 3–11.
With similar upgrades, the record is 0–1.
These checks do not remove differences in opponents, maps or player skill.
This group of similar teams still needs review; some teams may belong in another group.
Taichi & AgumonWarriorplaced here 100% · on 89% of boards
PBlightshroomPhantimalplaced here 89%
Common variations
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. Rows can overlap, so do not add their appearances together.
Heroes
Players
Lyca ·
–
2
Solise ·
–
2
Other recorded lineups
Exact five-hero teams outside the variations above, including uncommon lineups and those with poor results.
Lineup
Players
·
–
1
·
–
1
·
1
·
–
1
·
–
1
Hero results across this team group
Overall results with and without each supporting hero, across all variations. Teammates, maps, players and opponents can differ; this does not measure the effect of swapping one hero alone.
Hero
Taichi & Agumon
Lyca
Solise
Cyran
Pang
Dun/Pandora/CryonaiaCore100%94%83%Flex61%61%39%22%17%17%Artifact44% useAppearances189–9Win rate50%Players7StrengthTier 254% vs averageEvidence behind these results
18 appearances from 7 players and 14 team selections.
If SL rules require a player to reuse the same teams, their team on each map counts as one selection across those matches. Ro4, Finals and Guild Duel allow changes between matches, so each match counts separately. These selections can still share a strategy.
The most frequent player accounts for
5 appearances (28%).
Other players together went 6–7.
With similar upgrades, the record is 3–4.
These checks do not remove differences in opponents, maps or player skill.
The group's usual lineup, placed as it was most often on SL-Map4 (SL4–5):
of this exact placement,
67% of the group's 9 boards there.
How the team and its percentages are chosen.
Core
Dunlingr 100%, Pandora 94%, Cryonaia 83%
Artifacts
Vanguard · 44% of teams 25% win rate (2–6)
Wing Guard · 33% of teams 83% win rate (5–1)
Vine Snare · 17% of teams 33% win rate (1–2)
CryonaiaMageplaced here 100% · on 67% of boards
AlicethMarksmanplaced here 100%
PandoraSupportplaced here 100%
DunlingrTankplaced here 100%
RavionRogueplaced here 100%
PMidnight HunterPhantimalplaced here 100%
Common variations
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. Rows can overlap, so do not add their appearances together.
Heroes
Other recorded lineups
Exact five-hero teams outside the variations above, including uncommon lineups and those with poor results.
Lineup
Players
·
1
·
2
·
1
·
–
1
Hero results across this team group
Overall results with and without each supporting hero, across all variations. Teammates, maps, players and opponents can differ; this does not measure the effect of swapping one hero alone.
Hero
Aliceth
Ravion
Alna
Cyran
Niru
Taichi & Agumon
Rowan/Saida/Contess/EvieCore100%100%93%93%Flex64%36%Artifact64% useAppearances1410–4Win rate71%Players2StrengthTier 260% vs averageEvidence behind these results
14 appearances from 2 players and 7 team selections.
If SL rules require a player to reuse the same teams, their team on each map counts as one selection across those matches. Ro4, Finals and Guild Duel allow changes between matches, so each match counts separately. These selections can still share a strategy.
The most frequent player accounts for
12 appearances (86%).
Other players together went 2–0.
With similar upgrades, the record is 3–3.
These checks do not remove differences in opponents, maps or player skill.
This group of similar teams still needs review; some teams may belong in another group.
The group's usual lineup, placed as it was most often on SL-Map3 (SL4–5):
of this exact placement,
100% of the group's 7 boards there.
How the team and its percentages are chosen.
Core
Rowan 100%, Saida 100%, Contess 93%, Evie 93%
Artifacts
Breakthrough · 64% of teams 100% win rate (9–0)
Windcall · 29% of teams 25% win rate (1–3)
Magicsurge · 7% of teams 0% win rate (0–1)
RolanSupportplaced here 100%
RowanSupportplaced here 100%
EvieSupportplaced here 100%
SaidaRogueplaced here 100%
ContessSupportplaced here 100%
PMidnight HunterPhantimalplaced here 100%
Common variations
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. Rows can overlap, so do not add their appearances together.
Heroes
Other recorded lineups
Exact five-hero teams outside the variations above, including uncommon lineups and those with poor results.
Lineup
Players
·
1
·
1
Hero results across this team group
Overall results with and without each supporting hero, across all variations. Teammates, maps, players and opponents can differ; this does not measure the effect of swapping one hero alone.
Hero
Rolan
Baelran
Thador/SoliseCore98%93%Flex68%66%49%45%16%Artifact33% useAppearances22189–129Win rate41%Players73StrengthTier 342% vs averageEvidence behind these results
221 appearances from 73 players and 164 team selections.
If SL rules require a player to reuse the same teams, their team on each map counts as one selection across those matches. Ro4, Finals and Guild Duel allow changes between matches, so each match counts separately. These selections can still share a strategy.
The most frequent player accounts for
24 appearances (11%).
Other players together went 75–119.
With similar upgrades, the record is 38–60.
These checks do not remove differences in opponents, maps or player skill.
The group's usual lineup, placed as it was most often on SL-Map4 (SL4–5):
of this exact placement,
25% of the group's 52 boards there (7 of them Guild Duel games on the same ground).
How the team and its percentages are chosen.
Core
Thador 98%, Solise 93%
Artifacts
Vanguard · 33% of teams 58% win rate (43–31)
Breakthrough · 26% of teams 27% win rate (15–41)
Wing Guard · 14% of teams 41% win rate (12–17)
FrierenMageplaced here 79%
SoliseSupportplaced here 83%
HimmelWarriorplaced here 91% · on 88% of boards
ThadorTankplaced here 92%
Lily MayRogueplaced here 43% · on 67% of boards
PBlightshroomPhantimalplaced here 35%
Loses to: Gwyn/Gunnar/Kulu
Common variations
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. Rows can overlap, so do not add their appearances together.
Heroes
Players
Callan
5
Lily May ·
2
Alna ·
–
1
Daimon ·
–
1
Other recorded lineups
Exact five-hero teams outside the variations above, including uncommon lineups and those with poor results.
Lineup
Players
14
9
·
1
·
1
·
1
·
–
1
·
1
·
2
·
–
1
·
1
·
1
·
–
1
·
2
·
1
·
1
·
1
·
1
·
1
·
–
1
·
–
1
·
–
1
·
–
1
·
–
1
·
1
·
–
1
·
1
·
–
1
·
1
·
–
1
·
–
1
·
1
·
–
1
·
–
1
·
–
1
·
–
1
·
–
1
Hero results across this team group
Overall results with and without each supporting hero, across all variations. Teammates, maps, players and opponents can differ; this does not measure the effect of swapping one hero alone.
Hero
Frieren
Lily May
Himmel
Ravion
Contess
Satrana/BonnieCore100%100%Flex67%48%44%33%26%22%19%Artifact52% useAppearances273–24Win rate11%Players18StrengthTier 316% vs averageEvidence behind these results
27 appearances from 18 players and 23 team selections.
If SL rules require a player to reuse the same teams, their team on each map counts as one selection across those matches. Ro4, Finals and Guild Duel allow changes between matches, so each match counts separately. These selections can still share a strategy.
The most frequent player accounts for
7 appearances (26%).
Other players together went 2–18.
With similar upgrades, the record is 0–9.
These checks do not remove differences in opponents, maps or player skill.
The group's usual lineup, placed as it was most often on SL-Map4 (SL1–3):
of this exact placement,
100% of the group's 1 boards there.
How the team and its percentages are chosen.
Core
Bonnie 100%, Satrana 100%
Artifacts
Surging · 52% of teams 7% win rate (1–13)
Magicsurge · 22% of teams 17% win rate (1–5)
Vine Snare · 15% of teams 0% win rate (0–4)
ZoryaWarriorplaced here 100%
BonnieMarksmanplaced here 100%
AthaliaRogueplaced here 100%
SatranaMageplaced here 100%
NiruSupportplaced here 100%
PNecrodrakonPhantimalplaced here 100%
Common variations
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. Rows can overlap, so do not add their appearances together.
Heroes
Players
IgorShadewing ·
–
2
NiruZorya ·
1
AthaliaZorya ·
–
1
CallanContess ·
–
1
CallanIgor ·
1
CallanShadewing ·
–
1
IgorLorsan ·
1
IgorNiru ·
1
IgorZorya ·
–
1
ShadewingZorya ·
–
1
Players
Athalia ·
3
Callan ·
1
Daimon ·
1
Igor ·
1
Other recorded lineups
Exact five-hero teams outside the variations above, including uncommon lineups and those with poor results.
Lineup
Players
·
–
1
·
1
·
–
1
·
–
1
·
–
1
·
1
Hero results across this team group
Overall results with and without each supporting hero, across all variations. Teammates, maps, players and opponents can differ; this does not measure the effect of swapping one hero alone.
Hero
Zorya
Igor
Daimon
Niru
Callan
Shadewing
Cyran
Insights
Place Gwyneth behind Gunnar and include Dunlingrhigh confidence
Gwyneth stands directly behind Gunnar on 448 of 454 boards fielding both. Gwyn/Gunnar/Kulu is with Dunlingr and without him.
What to do Keep Gwyneth behind Gunnar and prioritize Dunlingr when the opponent may use Saida/Contess/Twins or Thador/Solise.
Gunnar gives the ally directly behind him extra range, attack and steadfast. That placement is part of the skill, so keep her behind him when changing the formation. Dunlingr adds a block on enemy ultimates at battle start: his Bell has one mode chosen before battle, and players usually use Spellbind. Gwyneth relies on normal attacks while many opposing teams need their first ultimates.
With Dunlingr, Gwyn/Gunnar/Kulu is against Saida/Contess/Twins and against Thador/Solise. Different players and opponents can affect those records, but his ultimate block gives a reason to favor this version against teams that need early casts.
Alna/Syph/Mehira needs the Winter Warrior buff on Sylphira and a path to the enemymedium confidence
Alna/Syph/Mehira is with Zandrok in Supreme League, against without him; the Guild Duel records are and .
What to do Place Sylphira so she receives Winter Warrior. Favor Zandrok where his Haste buff and wall destruction help the team reach the enemy.
Alna slows her own allies as well as the enemy. Her Winter Warrior escapes the Haste reduction and gains max HP: the closest ally in her diagonal row, ties to the frontmost unit. Across 471 Alna boards, Sylphira receives it 294 times, going 169–124 with a tie. The thirteen boards with nobody in the row went 3–10.
Zandrok helps the rest of the melee team with opening Haste and max HP, and his charge destroys breakable walls. That clears a route to the enemy and can remove a wall placed by Kulu. Permanent blocks remain. A battle-start charm from Mehira can interrupt his roar, so when both teams use Alna, check whether Mehira can interrupt Zandrok at battle start.
Evie/Orion benefits from Rolan but struggles when enemies stop its ultimatesmedium confidence
Evie/Orion is alongside Rolan, against without him.
What to do Use allies that cast ultimates frequently and keep Orion and Perseus safe so they can deal damage. Adding Rolan does not make the team reliable against Contess or Dunlingr.
Rolan supplies temporary buffs for Orion and benefits from frequent allied ultimates. Those casts also protect Evie: each ally ultimate lowers her Detection meter, keeping her concealed longer. Contess and Saida interfere with the casting that makes this work, while Dunlingr can prevent the opening ultimates. Orion is 20–52 with two ties against boards containing Contess, and 11–37 against boards containing Dunlingr. The Rolan split supports the build, but it does not turn its bad matchups into favorable ones.
Check the format before adding Rolan to Saida/Contess/Twinsmedium confidence
The Rolan version is in Supreme League but in Guild Duel.
What to do Practice the Rolan build, but place it where it is less likely to face Gwyn/Gunnar/Kulu and check its Guild Duel evidence separately.
Saida/Contess/Twins punishes enemy ultimate casts while supplying its own healing and Energy. Rolan can support that team, but its strong Supreme League record does not establish the same advantage in Guild Duel. Check which opposing teams it would face and whether another team in your set needs Rolan more.
Check Gerda’s partner in Saida/Contess/Twinsmedium confidence
With Gerda and Solise, the exact team is across 20 players. Its frequent use is not a reason to copy it.
What to do If Gerda is part of this team, test Aliceth as her partner before copying the common Solise version.
Keeping Gerda and using Aliceth instead, the team is across 10 players. With similar upgrades, the Solise version is and the Aliceth version is .
These results compare different players, opponents and formations, so they do not prove Aliceth causes the improvement. They do make the Aliceth version a more promising starting point to test. Check whether Gwyn/Gunnar/Kulu already needs her before assigning the full set.
Rolan and the twins: choose supporting heroes your other teams can sparemedium confidence
Rolan/Twins boards vary their supporting heroes, and the usual ones are also needed by other main teams.
What to do Check the full roster before using its record, and reserve the shared heroes before choosing the other teams.
Sylphira is on 67% of its boards, Taichi & Agumon on 49%, Contess and Mehira on 46% each, and Athalia on 26%. A Sylphira version competes directly with Alna/Syph/Mehira for heroes; a Contess version competes with Saida/Contess/Twins. Choose the version as part of your full set.
In Supreme League the boards with Contess are , against without her; the boards with Sylphira are , against without her. Different players and opponents sit behind those splits, and the counter records pool every version, so check which supporting heroes the recorded wins used before copying one.
Avoid sending Rolan/Twins into an expected Gwyn/Gunnar/Kulu on SL map 5medium confidence
Rolan/Twins is on SL map 5: against Gwyn/Gunnar/Kulu, but against other teams.
What to do Check the opposing team before assigning Rolan/Twins to map 5. If Gwyn/Gunnar/Kulu is likely there, try another matchup and recheck the rest of your set.
The losing map record is concentrated in one matchup. It does not establish that the arena itself is bad for Rolan. These are small samples with different players, upgrades and supporting heroes; moving the same lineup elsewhere does not guarantee a win.
Thador/Solise struggles against Gwyn/Gunnar/Kulu even with the usual formationmedium confidence
Keep Frieren, Thador and Solise each within one tile of Himmel to form his party. That formation still needs a favorable opponent.
What to do Keep Frieren, Thador and Solise each within one tile of Himmel. Avoid placing Thador/Solise where it may face Gwyn/Gunnar/Kulu, even if your heroes have more upgrades.
Across boards fielding all four, the party is formed on 101 boards at 50–51; the six boards that break it are 0–6. The skill requires that placement; the six losses alone cannot show how much it improves results. Lily May can interrupt an early opposing ultimate, but cannot stop Gwyneth’s normal attacks. See the counter results before assigning this team to an opponent.
Try Frieren with Alna, Sylphira, Mehira and Zandrok if she is availablelow confidence
This exact lineup is , including with similar upgrades, but only 3 players used it.
What to do Test the Frieren version against your expected opponent if your other teams can spare her. Keep Sylphira receiving Alna’s Winter Warrior buff.
One player accounts for 11 of its 16 appearances; the other two went 4–1. The result is worth testing, but opponents, placement and player differences could explain part of it.
With Hepler in Frieren’s place, the same four heroes are across 15 players, and with similar upgrades. That is a broader sample, not a controlled hero swap. Frieren may also be needed alongside Himmel and Thador/Solise in your set.
How team groups and common variations are defined
Teams are grouped by shared heroes.
A hero on 80% or more of a group's boards counts as core; 15% to below 80% counts as flex.
Every qualifying hero is shown, in order of usage, with no limit on the number of icons.
These shares count recorded appearances, including ties and matches where players must reuse the same teams;
they do not measure strength or independent draft choices. Rarer picks stay in the linked recorded teams,
but are omitted from the core and flex lists. Artifacts do not determine the groups.
For common variations, we first find heroes often used together, then count only teams containing
every listed hero. A row needs at least eight matching appearances in the team group. +1 flex means
one other hero completes the team; expanding a row shows the actual heroes or combinations filling those slots.
Rows can overlap, so their counts cannot be added together. Other recorded lineups keeps every exact
team outside those rows, including rare picks and losing teams.
These cutoffs are analysis choices, not game rules. Broader or narrower groups can change a
group's record. Each profile includes the results for its common variations, so you can compare
recorded teams without listing every small variation as a separate group.
Flex comparisons can also differ in players, upgrades and opponents. A better record with a
hero is worth testing, but does not prove that hero caused the improvement.
How the evidence labels work
Limited evidence means fewer than 30 games with a winner, 10 team selections or 5 players.
Concentrated sample means one player contributes at least 35% of appearances, or the two most frequent players contribute at least 60%.
Promising flags a win rate of at least 60% with limited evidence. Poor results so far flags a win rate of 40% or less.
These are reading aids, not statistical confidence tests or changes to the tiers. A winning record may depend on opponents or upgrades; a losing record can still include a useful counter.
The evidence in each profile shows the record without one most frequent player and at similar upgrades. When SL rules require the same teams for several matches, they count as one team selection per player and map. Separate selections can still share a strategy.
Guild membership is not recorded, so these checks cannot identify teams copied within a guild. No team group is excluded for having a losing record.
How strength is measured, including upgrade differences
The model estimates a rating for each team group from its results against the other groups. A team rated
r is expected to beat one rated s about r/(r+s) of the time. An average team is rated
1.00. These ratings compare teams with equal upgrades. This gives the expected win rate against an average team shown above.
The rating range in the table below comes from 3,000 repeated samples of 171 groups of related matches.
Every match stays together. Matches where a player was required to reuse the same five teams also stay in one group.
Each sample refits both team strength and the upgrade effect. Tier 1 requires the entire range to
imply a win rate above 55% against an average team at equal upgrades (a rating above 1.22).
Tier 3 means the entire rating range is below 1.00; all other teams are Tier 2.
The tiers compare the recorded teams with each other: a Tier 3 team is still
one that players field, rated below the others in this data rather than below every possible team. The
same tiers group the teams in the counter ladder.
This rating range is different from the observed win-rate interval in the main table.
The same players can still recur in different stages, so these intervals do not remove every source of dependence.
The adjusted rating accounts for upgrades. A team can win partly because its heroes
have more paragon or EX refinement than the opponent. The adjusted rating includes those differences
when estimating team strength.
Both upgrades use one scale, Rivalry points. A paragon level is
4.5 points, the Inspiration/Intimidation it grants (each net point moves healing, shield value, extra
Energy gains and buff effects by 1%). A refinement level is 0.9 points, its stat gains
measured against a paragon level's, and the skill effect unlocked at R2 and the Rivalry-mode stats
added at R4 are 2.1 points each, so a fully refined hero is worth
7.7 points against an unrefined one, about 1.7 paragon levels. The equal-upgrade rating
uses each game's combined upgrade gap. The model estimates the effect of that gap, multiplying
the advantaged side's odds by ×1.05 per net point (×1.24 for
one paragon level on one hero, ×1.44 for one fully refined hero,
×1.53 for a 9-point gap), estimated from the 749 decided
games where the sides were not equal. These are changes in odds, not percentage-point increases in win rate.
Players with more upgrades also tend to have better results. The model cannot fully separate their
skill from their upgrades. A lower adjusted rating suggests that upgrade advantages explain part of
the raw record. A higher one suggests the team performed well despite having fewer upgrades.
Tiers use the equal-upgrade rating and its resampled interval. The unadjusted rating remains a comparison.
Near level is the team's record
in games where the combined upgrade gap was below 4.5 points, equivalent to one paragon level on
one hero. The recorded games do not show a winning advantage for gaps that small.
Is refinement weighted right? The same model run with paragon and refinement weighted separately
prefers a refinement point at 1.21 of a paragon point,
from the 481 decided games that carried a refinement gap. That is a point estimate, with no
interval for the exchange rate. Allowing separate effects improves the log-likelihood, a measure of fit to
these recorded outcomes, by 0.19. That is little improvement, so these results do not give a strong reason to change the agreed scale.
Both fits still mix upgrade effects with differences between players and their teams.
Team
Equal-upgrade rating
Unadjusted rating
Expected vs average
Range
Raw win rate
Opponent strength
vs strong
vs the rest
Verdict
Rolan/Twins
3.15
3.73
76%
1.85–7.75
74%
62%(8–5)
1.22
59%(17–12)
90%(27–3)
Tier 1
Gwyn/Gunnar/Kulu
2.60
2.78
72%
1.91–3.53
65%
65%(129–69)
1.17
56%(72–56)
77%(159–48)
Tier 1
Evie/Orion
1.92
1.70
66%
1.46–2.60
52%
48%(47–51)
1.42
38%(42–69)
66%(86–45)
Tier 1
Alna/Syph/Mehira
1.90
2.00
65%
1.44–2.64
56%
56%(78–62)
1.42
53%(62–56)
62%(75–46)
Tier 1
Athalia/Alna/Eryndor
1.63
1.15
62%
0.58–3.50
49%
86%(6–1)
1.23
50%(7–7)
47%(8–9)
Tier 2
Saida/Contess/Twins
1.53
1.65
60%
1.15–2.04
54%
52%(61–57)
1.35
43%(66–88)
70%(74–31)
Tier 2
Rowan/Saida/Contess/Evie
1.53
2.41
60%
0.46–8.43
71%
50%(3–3)
1.01
75%(3–1)
70%(7–3)
Tier 2
Dun/Pandora/Cryonaia
1.16
1.43
54%
0.45–2.83
50%
43%(3–4)
1.38
33%(3–6)
67%(6–3)
Tier 2
Alna/Athalia
1.09
1.12
52%
0.48–2.63
43%
29%(2–5)
1.39
0%(0–6)
60%(9–6)
Tier 2
Zorya/Athalia/Pandora
0.84
0.71
46%
0.49–1.47
35%
36%(15–27)
1.52
31%(23–52)
37%(14–24)
Tier 2
Thador/Solise
0.71
0.82
42%
0.47–0.99
41%
39%(38–60)
1.31
15%(12–69)
59%(58–41)
Tier 3
Eryndor/LM/Thador
0.57
0.44
36%
0.27–1.54
21%
0%(0–1)
1.49
0%(0–9)
40%(4–6)
Tier 2
Marcille/Ravion/Thador
0.46
0.51
31%
0.16–1.02
30%
38%(5–8)
1.28
19%(3–13)
42%(5–7)
Tier 2
Gwyn/Cyran/Hugin
0.45
0.35
31%
0.16–1.00
22%
17%(1–5)
1.38
27%(3–8)
16%(3–16)
Tier 2
Satrana/Bonnie
0.19
0.15
16%
0.06–0.38
11%
0%(0–9)
1.04
0%(0–8)
16%(3–16)
Tier 3
Opponent strength is the average rating of the opponents a team met:
above 1.00 means stronger opponents than average. vs strong is the record against the
four teams
in the top tier, vs the rest against everyone else including one-off line-ups, which rate
0.52 as a group with equal upgrades (0.55 unadjusted).
A high rating does not guarantee a favorable matchup. Use the counter results
for the opponent you expect.
Each team's record with the upgrade advantage (its side a paragon level or more ahead in
upgrade points), behind (the other side that far ahead), and near level. Neither rating fully separates
player skill from team strength, so compare the raw and adjusted ratings with these records.
Team
With the upgrade advantage
Behind
Near level
Gwyn/Gunnar/Kulu
76%(109–34)
45%(40–48)
65%(129–69)
Evie/Orion
77%(65–19)
39%(48–76)1 tied
48%(47–51)1 tied
Saida/Contess/Twins
77%(70–21)7 tied
28%(20–52)3 tied
52%(61–57)8 tied
Alna/Syph/Mehira
77%(62–19)
33%(23–47)
56%(78–62)1 tied
Thador/Solise
65%(40–22)
19%(11–47)3 tied
39%(38–60)
Zorya/Athalia/Pandora
33%(10–20)
35%(19–36)
36%(15–27)
Rolan/Twins
92%(34–3)
27%(3–8)
62%(8–5)
Athalia/Alna/Eryndor
57%(4–3)
33%(7–14)
86%(6–1)
Gwyn/Cyran/Hugin
57%(4–3)
11%(2–17)
17%(1–5)
Marcille/Ravion/Thador
44%(4–5)
0%(0–8)
38%(5–8)
Satrana/Bonnie
33%(1–2)
13%(2–13)
0%(0–9)
Alna/Athalia
60%(3–2)
44%(4–5)
29%(2–5)
Eryndor/LM/Thador
50%(3–3)
8%(1–11)
0%(0–1)
Dun/Pandora/Cryonaia
50%(4–4)
67%(2–1)
43%(3–4)
Rowan/Saida/Contess/Evie
88%(7–1)
n/a
50%(3–3)
How to read the formation diagrams and their percentages
Each diagram shows the group's usual lineup as it was actually placed on a ground in the Supreme League
rotation. The team is the five-hero line-up the group fielded most often on those grounds, with each
line-up's uses weighted by how common its heroes are within the group (the Core and Flex percentages
multiplied together), so the board shows the popular flex picks rather than a placement one player
repeated with an unusual pick. Guild Duel games count when they were played on a ground that is also a
Supreme League map (GD-Map1 and GD-Map2 are); Guild Duel's other grounds are left out, because
Guild Duel is played far more often and would otherwise decide the board. The ground shown is the one
where that team was placed the same way most often, named by its Supreme League map position, and the
caption counts uses of that exact placement. It is an example of placement, not a recommendation to
choose that map. A team group never recorded on a Supreme League ground is drawn from all of its boards.
Enemy-side formations are mirrored so both sides can be compared.
Placed here is how often a unit used that tile; on N% of boards is how often it
appeared. Both refer to the team group on the diagram's map. The core and flex percentages in the main
table instead cover all maps.
The phantimal is summoned automatically from the team's factions. You choose its tile, which
can affect targeting and artifact effects, but do not select it independently of the heroes.
These teams fall outside the named team groups; some have been played more than once.
Line-ups outside the named groups that beat an established team rated average or
better. Each row is one exact five-hero build. Treat a row as a lead worth testing, not a proven
answer. Most have a handful of games. Pay particular attention to wins with similar or fewer upgrades.
Select a record to see sample games the line-up played against named teams, or a map name to see that whole match.
Line-up
Artifact
Beat
Alna, Athalia, Elijah & Lailah, Frieren, Himmel
Gwyn/Gunnar/Kuluon near level
Alna/Syph/Mehiraon near level
Gwyn/Gunnar/Kuluon near level
Athalia, Baelran, Elijah & Lailah, Gerda, Rolan
Rolan/Twinson with a +49-point upgrade advantage
Evie/Orionon near level
Alna/Syph/Mehiraon near level
Evie, Frieren, Rolan, Solise, Taichi & Agumon
Alna/Syph/Mehiraon
Saida/Contess/Twinson near level
Contess, Cryonaia, Niru, Pandora, Saida
Evie/Orionon with a +22.5-point upgrade advantage
Saida/Contess/Twinson with a +14-point upgrade advantage
Aliceth, Dunlingr, Frieren, Himmel, Rolan
Gwyn/Gunnar/Kuluon with a +18.5-point upgrade advantage
Gwyn/Gunnar/Kuluon near level
Aliceth, Alna, Contess, Frieren, Sylphira
Alna/Syph/Mehiraon near level
Gwyn/Gunnar/Kuluon near level
Athalia, Contess, Gerda, Rolan, Saida
Evie/Orionon with a +13.2-point upgrade advantage
Evie/Orionon with a +5.5-point upgrade advantage
Gunnar, Gwyneth, Perseus, Rowan, Zandrok
Rowan/Saida/Contess/Evieon near level
Alna, Hepler, Koko, Orion, Zandrok
Gwyn/Gunnar/Kuluon with a +7-point upgrade advantage
Cryonaia, Evie, Pandora, Peggy, Rowan
Gwyn/Gunnar/Kuluon with a +17.5-point upgrade advantage
Artifact choices can help identify an opponent's team, but several teams use the same artifact.
Choose yours with the full formation in mind; an artifact's overall win rate also reflects the teams
and players using it. Each team group profile lists its artifact choices and results.
Team groups are stacked by equal-upgrade strength rating, best on top, in their rating tiers. A line on the right runs from a
higher-rated team down to a lower-rated team it beats; a line on the left runs upward, from a lower-rated team
to a higher-rated team it beats. The arrowhead points at the team that loses, and the number on each line is the
winner's record in those games. Select a line for the recorded evidence; select
a team to show only its lines, and click anywhere else to show every line again.
43–2425–032–146–1440–1117–1628–188–229–1711–317–6
Tier 1
Tier 2
Tier 3
← left: beats a higher-rated teamright: beats a lower-rated team →
80% or more observed wins65–79% observed winsunder 65% observed winspossible, not establishedwhen a team is selected, the lines of the teams that beat it
Connections use the test accounting for maps and upgrades. Line styles and win–loss labels describe the raw records, so they can also reflect upgrade advantages.
All supported counter relationships are shown. Missing lines mean insufficient evidence.
Read across a row for that team's win rate against each opponent; wins–losses appear underneath.
Green is a winning matchup, red a losing one. Bright bold percentages meet the statistical test below;
the rest are dimmed. 147 games used teams from the same group on both sides (mirrors) and are left out.
A losing record alone does not establish a counter. The team may struggle against many opponents or
on that particular map. The test compares its record against this opponent with its record
against everyone else on the same maps, then combines those comparisons. Each game's upgrade gap also
affects the expected outcome: losing while far behind on paragon and refinement is less evidence of a counter,
and winning from behind is more. Games with teams from the same group on both sides are excluded. A map contributes no
comparison if the team faced only one opponent there.
All win rates below are for the team in the first column; wins–losses appear in parentheses.
Team
Counter
Verdict
Gwyn/Gunnar/Kulu
Alna/Syph/Mehira
−41
under 1 in 1,000
supported
Zorya/Athalia/Pandora
Evie/Orion
−42
under 1 in 1,000
supported
Thador/Solise
Gwyn/Gunnar/Kulu
−44
under 1 in 1,000
supported
Saida/Contess/Twins
Gwyn/Gunnar/Kulu
−41
under 1 in 1,000
supported
Evie/Orion
Gwyn/Gunnar/Kulu
−40
under 1 in 1,000
supported
Gwyn/Gunnar/Kulu
Zorya/Athalia/Pandora
−27
under 1 in 1,000
supported
Alna/Syph/Mehira
Evie/Orion
−22
around 1 in 290
supported
Rolan/Twins
Gwyn/Gunnar/Kulu
−59
around 1 in 108
supported
Evie/Orion
Saida/Contess/Twins
−20
around 1 in 70
supported
Evie/Orion
Rolan/Twins
−36
around 1 in 51
supported
Thador/Solise
Evie/Orion
−17
around 1 in 8
possible, not established
Near level keeps only games with an upgrade gap below 4.5 points, combining
paragon and refinement. It helps show whether the losing record also appears when upgrades are similar.
Player skill and team differences can still affect it. Win-rate gap is the difference between the
two win rates, in percentage points.
The p-value describes how unusual a result this extreme would be if there were no
special matchup effect after accounting for maps and upgrade gaps. Smaller values provide stronger evidence under the
test's assumptions. It is not the probability that a counter is false.
Counter results by map, format, player and team version
Alna/Syph/Mehira beats Gwyn/Gunnar/Kulu
On the map slots where they met, Gunnar / Gwyneth / Kulu wins
77% of its games against everyone else but
36% against Alna/Syph/Mehira: 24 wins, where the maps and upgrade
gaps would lead to about 39 without a special matchup effect. The test accounting for maps and
upgrade gaps gives a p-value of under 1 in 1,000. Counted by whole matches instead of games it
is 39–20 with 3 split to Alna/Syph/Mehira, from
62 separate matches between 51 pairs of players.
Gwyn/Gunnar/Kulu's record
GD-Map1
GD-Map2
GD-Map3 (GD1 to GD3-2)
GD-Map3 (GD3-4 to GD5)
SL-Map1
SL-Map2
SL-Map3 (SL1–3)
SL-Map3 (SL4–5)
SL-Map4 (SL4–5)
SL-Map5
vs Alna/Syph/Mehira
vs everyone else, same slot
65%(11–6)
88%(29–4)
80%(4–1)
75%(3–1)
66%(29–15)
63%(27–16)
100%(3–0)
81%(22–5)
71%(12–5)
88%(60–8)
Recorded in both formats: SL 13–21 against it, 153–49 against everyone else; GD 11–22 against it, 47–12 against everyone else.
Upgrade check: 35 of the cell's games carried a gap of a paragon level or more and the side
holding the upgrade advantage went 22–13 in them; with those removed the cell is still
24–8 to Alna/Syph/Mehira, so it also has a winning record with similar upgrades. 12 of the gap games were level on paragon and apart on refinement alone; the refined side went 7–5 there.
Common variations against this opponent: Gunnar/Gwyneth/Kulu/Dunlingr + 1 flex ; Gunnar/Gwyneth/Dunlingr/Cyran/Alna . Every listed hero is required; rows can overlap.
Evie/Orion beats Zorya/Athalia/Pandora
On the map slots where they met, Zorya / Pandora / Athalia wins
42% of its games against everyone else but
0% against Evie/Orion: 0 wins, where the maps and upgrade
gaps would lead to about 9 without a special matchup effect. The test accounting for maps and
upgrade gaps gives a p-value of under 1 in 1,000. Counted by whole matches instead of games it
is 23–0 to Evie/Orion, from
23 separate matches between 22 pairs of players.
Zorya/Athalia/Pandora's record
GD-Map3 (GD3-4 to GD5)
SL-Map1
SL-Map2
SL-Map3 (SL4–5)
SL-Map4 (SL4–5)
SL-Map5
vs Evie/Orion
vs everyone else, same slot
0%(0–2)
62%(8–5)
40%(12–18)
17%(2–10)
63%(5–3)
44%(8–10)
Recorded in both formats: SL 0–24 against it, 35–46 against everyone else; GD 0–1 against it, 0–2 against everyone else.
Upgrade check: 17 of the cell's games carried a gap of a paragon level or more and the side
holding the upgrade advantage went 9–8 in them; with those removed the cell is still
8–0 to Evie/Orion, so it also has a winning record with similar upgrades. Three of the gap games were level on paragon and apart on refinement alone; the refined side went 2–1 there.
Gwyn/Gunnar/Kulu beats Thador/Solise
On the map slots where they met, Thador / Solise wins
47% of its games against everyone else but
3% against Gwyn/Gunnar/Kulu: 1 wins, where the maps and upgrade
gaps would lead to about 11 without a special matchup effect. The test accounting for maps and
upgrade gaps gives a p-value of under 1 in 1,000. Counted by whole matches instead of games it
is 32–1 to Gwyn/Gunnar/Kulu, from
33 separate matches between 33 pairs of players.
Thador/Solise's record
GD-Map2
SL-Map1
SL-Map2
SL-Map3 (SL1–3)
SL-Map3 (SL4–5)
SL-Map4 (SL1–3)
SL-Map4 (SL4–5)
SL-Map5
vs Gwyn/Gunnar/Kulu
vs everyone else, same slot
100%(2–0)
27%(4–11)
53%(9–8)
90%(9–1)
46%(12–14)
62%(8–5)
41%(19–27)
33%(5–10)
Recorded in both formats: SL 1–27 against it, 66–76 against everyone else; GD 0–5 against it, 2–0 against everyone else.
Upgrade check: 15 of the cell's games carried a gap of a paragon level or more and the side
holding the upgrade advantage went 9–6 in them; with those removed the cell is still
17–1 to Gwyn/Gunnar/Kulu, so it also has a winning record with similar upgrades. Four of the gap games were level on paragon and apart on refinement alone; the refined side went 4–0 there.
Common variations against this opponent: Solise/Thador/Frieren/Himmel/Lily May ; Solise/Lily May/Saida/Thador/Contess ; Ravion/Thador/Taichi & Agumon/Solise + 1 flex . Every listed hero is required; rows can overlap.
Gwyn/Gunnar/Kulu beats Saida/Contess/Twins
On the map slots where they met, Contess / Saida / Elijah & Lailah wins
64% of its games against everyone else but
23% against Gwyn/Gunnar/Kulu: 14 wins, where the maps and upgrade
gaps would lead to about 27 without a special matchup effect. The test accounting for maps and
upgrade gaps gives a p-value of under 1 in 1,000. Counted by whole matches instead of games it
is 42–11 with 3 split to Gwyn/Gunnar/Kulu, from
56 separate matches between 51 pairs of players.
Saida/Contess/Twins's record
GD-Map1
GD-Map2
GD-Map3 (GD3-4 to GD5)
SL-Map1
SL-Map2
SL-Map3 (SL4–5)
SL-Map4 (SL1–3)
SL-Map4 (SL4–5)
SL-Map5
vs Gwyn/Gunnar/Kulu
vs everyone else, same slot
69%(11–5)
63%(5–3)
33%(3–6)
75%(21–7)
72%(21–8)
45%(14–17)
75%(3–1)
68%(27–13)
69%(20–9)
Recorded in both formats: SL 10–27 against it, 106–55 against everyone else; GD 4–19 against it, 19–14 against everyone else.
Upgrade check: 27 of the cell's games carried a gap of a paragon level or more and the side
holding the upgrade advantage went 20–7 in them; with those removed the cell is still
27–6 to Gwyn/Gunnar/Kulu, so it also has a winning record with similar upgrades. Five of the gap games were level on paragon and apart on refinement alone; the refined side went 3–2 there.
Common variations against this opponent: Contess/Elijah & Lailah/Saida + 2 flex ; Contess/Saida/Elijah & Lailah/Solise + 1 flex . Every listed hero is required; rows can overlap.
Gwyn/Gunnar/Kulu beats Evie/Orion
On the map slots where they met, Evie / Orion wins
61% of its games against everyone else but
22% against Gwyn/Gunnar/Kulu: 11 wins, where the maps and upgrade
gaps would lead to about 24 without a special matchup effect. The test accounting for maps and
upgrade gaps gives a p-value of under 1 in 1,000. Counted by whole matches instead of games it
is 36–9 with 2 split to Gwyn/Gunnar/Kulu, from
47 separate matches between 42 pairs of players.
Evie/Orion's record
GD-Map1
GD-Map2
GD-Map3 (GD1 to GD3-2)
GD-Map3 (GD3-4 to GD5)
SL-Map1
SL-Map2
SL-Map3 (SL4–5)
SL-Map4 (SL4–5)
SL-Map5
vs Gwyn/Gunnar/Kulu
vs everyone else, same slot
67%(8–4)
25%(1–3)
0%(0–1)
50%(7–7)
66%(33–17)
68%(27–13)
64%(14–8)
58%(18–13)
53%(8–7)
Recorded in both formats: SL 6–26 against it, 100–58 against everyone else; GD 5–14 against it, 16–15 against everyone else.
Upgrade check: 30 of the cell's games carried a gap of a paragon level or more and the side
holding the upgrade advantage went 19–11 in them; with those removed the cell is still
17–4 to Gwyn/Gunnar/Kulu, so it also has a winning record with similar upgrades. 10 of the gap games were level on paragon and apart on refinement alone; the refined side went 5–5 there.
Common variations against this opponent: Evie/Orion/Perseus/Frieren + 1 flex ; Hepler/Orion/Evie/Koko + 1 flex . Every listed hero is required; rows can overlap.
Zorya/Athalia/Pandora beats Gwyn/Gunnar/Kulu
On the map slots where they met, Gunnar / Gwyneth / Kulu wins
75% of its games against everyone else but
48% against Zorya/Athalia/Pandora: 16 wins, where the maps and upgrade
gaps would lead to about 24 without a special matchup effect. The test accounting for maps and
upgrade gaps gives a p-value of under 1 in 1,000. Counted by whole matches instead of games it
is 16–15 to Zorya/Athalia/Pandora, from
31 separate matches between 26 pairs of players.
Gwyn/Gunnar/Kulu's record
GD-Map2
SL-Map1
SL-Map2
SL-Map3 (SL4–5)
SL-Map4 (SL4–5)
SL-Map5
vs Zorya/Athalia/Pandora
vs everyone else, same slot
82%(32–7)
63%(30–18)
65%(24–13)
79%(23–6)
70%(14–6)
86%(57–9)
Recorded in both formats: SL 15–16 against it, 148–52 against everyone else; GD 1–1 against it, 32–7 against everyone else.
Upgrade check: 23 of the cell's games carried a gap of a paragon level or more and the side
holding the upgrade advantage went 7–16 in them; with those removed the cell is still
3–7 to Zorya/Athalia/Pandora, so the games with similar upgrades do not show a winning record for this counter. Five of the gap games were level on paragon and apart on refinement alone; the refined side went 0–5 there.
Common variations against this opponent: Gunnar/Gwyneth/Kulu/Dunlingr + 1 flex ; Evie/Gunnar/Gwyneth + 2 flex ; Gunnar/Gwyneth/Dunlingr/Cyran/Alna . Every listed hero is required; rows can overlap.
Evie/Orion beats Alna/Syph/Mehira
On the map slots where they met, Alna / Sylphira / Mehira wins
62% of its games against everyone else but
39% against Evie/Orion: 18 wins, where the maps and upgrade
gaps would lead to about 26 without a special matchup effect. The test accounting for maps and
upgrade gaps gives a p-value of around 1 in 290. Counted by whole matches instead of games it
is 25–16 with 1 split to Evie/Orion, from
42 separate matches between 41 pairs of players.
Alna/Syph/Mehira's record
GD-Map1
GD-Map2
GD-Map3 (GD1 to GD3-2)
GD-Map3 (GD3-4 to GD5)
SL-Map1
SL-Map2
SL-Map3 (SL4–5)
SL-Map4 (SL4–5)
SL-Map5
vs Evie/Orion
vs everyone else, same slot
29%(5–12)
55%(6–5)
83%(15–3)
100%(8–0)
56%(15–12)
58%(21–15)
67%(18–9)
60%(18–12)
69%(11–5)
Recorded in both formats: SL 13–21 against it, 83–53 against everyone else; GD 5–7 against it, 34–20 against everyone else.
Upgrade check: 27 of the cell's games carried a gap of a paragon level or more and the side
holding the upgrade advantage went 18–9 in them; with those removed the cell is still
12–7 to Evie/Orion, so it also has a winning record with similar upgrades. Eight of the gap games were level on paragon and apart on refinement alone; the refined side went 4–4 there.
Gwyn/Gunnar/Kulu beats Rolan/Twins
On the map slots where they met, Elijah & Lailah / Rolan wins
79% of its games against everyone else but
20% against Gwyn/Gunnar/Kulu: 2 wins, where the maps and upgrade
gaps would lead to about 5 without a special matchup effect. The test accounting for maps and
upgrade gaps gives a p-value of around 1 in 108. Counted by whole matches instead of games it
is 7–2 to Gwyn/Gunnar/Kulu, from
9 separate matches between 8 pairs of players.
Rolan/Twins's record
SL-Map1
SL-Map2
SL-Map5
vs Gwyn/Gunnar/Kulu
vs everyone else, same slot
81%(13–3)
100%(2–0)
73%(8–3)
Only observed in SL 2–8 against it, 23–6 against everyone else.
Upgrade check: 5 of the cell's games carried a gap of a paragon level or more and the side
holding the upgrade advantage went 4–1 in them; with those removed the cell is still
4–1 to Gwyn/Gunnar/Kulu, so it also has a winning record with similar upgrades. One of the gap game was level on paragon and apart on refinement alone; the refined side went 0–1 there.
Common variations against this opponent: Elijah & Lailah/Rolan/Sylphira/Mehira + 1 flex ; Contess/Elijah & Lailah/Rolan/Taichi & Agumon + 1 flex . Every listed hero is required; rows can overlap.
Saida/Contess/Twins beats Evie/Orion
On the map slots where they met, Evie / Orion wins
57% of its games against everyone else but
37% against Saida/Contess/Twins: 17 wins, where the maps and upgrade
gaps would lead to about 24 without a special matchup effect. The test accounting for maps and
upgrade gaps gives a p-value of around 1 in 70. Counted by whole matches instead of games it
is 29–17 to Saida/Contess/Twins, from
46 separate matches between 39 pairs of players.
Evie/Orion's record
GD-Map1
GD-Map2
GD-Map3 (GD3-4 to GD5)
SL-Map1
SL-Map2
SL-Map3 (SL4–5)
SL-Map4 (SL4–5)
SL-Map5
vs Saida/Contess/Twins
vs everyone else, same slot
50%(8–8)
29%(2–5)
42%(5–7)
60%(29–19)
65%(30–16)
55%(11–9)
63%(17–10)
47%(7–8)
Recorded in both formats: SL 12–22 against it, 94–62 against everyone else; GD 5–7 against it, 15–20 against everyone else.
Upgrade check: 33 of the cell's games carried a gap of a paragon level or more and the side
holding the upgrade advantage went 22–11 in them; with those removed the cell is still
11–2 to Saida/Contess/Twins, so it also has a winning record with similar upgrades. 11 of the gap games were level on paragon and apart on refinement alone; the refined side went 8–3 there.
Rolan/Twins beats Evie/Orion
On the map slots where they met, Evie / Orion wins
57% of its games against everyone else but
21% against Rolan/Twins: 3 wins, where the maps and upgrade
gaps would lead to about 7 without a special matchup effect. The test accounting for maps and
upgrade gaps gives a p-value of around 1 in 51. Counted by whole matches instead of games it
is 11–3 to Rolan/Twins, from
14 separate matches between 13 pairs of players.
Evie/Orion's record
GD-Map3 (GD3-4 to GD5)
SL-Map1
SL-Map3 (SL4–5)
SL-Map4 (SL4–5)
SL-Map5
vs Rolan/Twins
vs everyone else, same slot
50%(8–8)
63%(32–19)
64%(14–8)
59%(19–13)
40%(8–12)
Recorded in both formats: SL 3–10 against it, 73–52 against everyone else; GD 0–1 against it, 8–8 against everyone else.
Upgrade check: 12 of the cell's games carried a gap of a paragon level or more and the side
holding the upgrade advantage went 11–1 in them; with those removed the cell is still
2–0 to Rolan/Twins, so it also has a winning record with similar upgrades. Five of the gap games were level on paragon and apart on refinement alone; the refined side went 5–0 there.
Evie/Orion beats Thador/Solisepossible, not established
On the map slots where they met, Thador / Solise wins
43% of its games against everyone else but
26% against Evie/Orion: 6 wins, where the maps and upgrade
gaps would lead to about 9 without a special matchup effect. The test accounting for maps and
upgrade gaps gives a p-value of around 1 in 8. Counted by whole matches instead of games it
is 14–4 with 2 split to Evie/Orion, from
20 separate matches between 19 pairs of players.
Thador/Solise's record
SL-Map1
SL-Map2
SL-Map3 (SL1–3)
SL-Map3 (SL4–5)
SL-Map4 (SL4–5)
vs Evie/Orion
vs everyone else, same slot
33%(4–8)
56%(9–7)
80%(8–2)
33%(10–20)
39%(17–27)
Only observed in SL 6–17 against it, 48–64 against everyone else.
Upgrade check: 14 of the cell's games carried a gap of a paragon level or more and the side
holding the upgrade advantage went 7–7 in them; with those removed the cell is still
6–3 to Evie/Orion, so it also has a winning record with similar upgrades. Two of the gap games were level on paragon and apart on refinement alone; the refined side went 1–1 there.
Common variations against this opponent: Solise/Thador/Frieren/Himmel/Lily May ; Solise/Lily May/Saida/Thador/Contess ; Ravion/Thador/Taichi & Agumon/Solise + 1 flex . Every listed hero is required; rows can overlap.
How reliable are these counters?
Testing many matchups can produce chance findings. 29 pairs of team groups
had enough games to test in both directions (58 tests). The counter table labels results with
p-values below 0.05 as supported. A stricter threshold divides 0.05 by the number of tests to reduce
the chance of finding a counter just because many comparisons were made. The counter insights
identify which matchups meet that threshold.
Repeated games are not independent examples. The same players can appear many times, sometimes because SL rules require them to reuse the same teams for several matches. Each matchup's details include a recount by whole match, the number of player pairs,
and a breakdown when one player accounts for a large share of the games. Those checks help show
how much the result depends on repeated games or particular players; they do not remove those effects.
The opponents still matter. A team's record against everyone else includes one-off line-ups,
which can make it look stronger. Comparing results on the same maps and pricing upgrade gaps does not remove
differences in player skill or the exact teams used. The fitted upgrade effect can also include differences
between players; these records do not separate those effects.
The upgrade effect is estimated. The counter test treats that fitted value as fixed. Its p-values
do not include uncertainty in the upgrade estimate, and repeated appearances by the same players and teams also limit their precision. Statistical support is a reason to investigate a matchup, not a calibrated chance of winning it.
The team grouping rules can also affect the result.
Every supported counter was also checked with broader and narrower groups.
Insights
Six counters have stronger statistical evidencehigh confidence
Alna/Syph/Mehira over Gwyn/Gunnar/Kulu, Gwyn/Gunnar/Kulu over Saida/Contess/Twins, Gwyn/Gunnar/Kulu over Evie/Orion, Gwyn/Gunnar/Kulu over Thador/Solise, Evie/Orion over Zorya/Athalia/Pandora and Zorya/Athalia/Pandora over Gwyn/Gunnar/Kulu meet the stricter threshold used for the 58 comparisons. Evie/Orion over Alna/Syph/Mehira, Gwyn/Gunnar/Kulu over Rolan/Twins, Saida/Contess/Twins over Evie/Orion and Rolan/Twins over Evie/Orion clear the ordinary 0.05 line but not the stricter one. The test counts each game by its upgrade gap, so a win taken while behind on upgrades counts for more than one taken with an edge.
What to do Use these counters when choosing teams for a likely matchup. Rolan/Twins still needs more games against Evie/Orion with similar upgrades.
The counter table gives the records and near-level comparisons. Zorya/Athalia/Pandora over Gwyn/Gunnar/Kulu meets the stricter threshold at the fitted upgrade effect, but only the ordinary threshold at the lower end of the effect's resampled interval. Evie/Orion against Zorya/Athalia/Pandora is , with of those games near level. Rolan/Twins against Evie/Orion is but only with similar upgrades, so its strong overall result still needs particular care.
The confirmed mechanics support these directions. Alna slows the Gunnar pair while Sylphira attacks its clustered backline; Zandrok helps the melee team reach it. Dunlingr prevents the opening casts Saida/Contess/Twins and Thador/Solise need, and his block also starves Evie’s team of the ultimates that keep her hidden while Gwyneth’s normal attacks are unaffected: Orion is 11–37 against boards with Dunlingr and 14–38 against boards with Gwyneth. Contess punishes casts and Saida drains Energy, undermining the allied ultimates that keep Evie concealed. Zorya does nothing until her Energy fills, so a team that attacks early can remove her support pieces before she wakes. A statistical label still describes recorded evidence, not a guaranteed result.
Evidence Threshold after accounting for all comparisons: 0.05 / 58 = 0.00086
Evie/Orion leads against Alna/Syph/Mehira, but the support is still limitedmedium confidence
Evie/Orion is against Alna/Syph/Mehira and with similar upgrades; the comparison accounting for maps and upgrades clears the 0.05 line but not the stricter threshold.
What to do Use Evie as a candidate against Alna, and compare it with a practiced mirror rather than assuming a proven counter exists.
Evie can begin inside the enemy half while invincible, which gives a reason to test the pairing. The recorded lead remains too uncertain to treat it as a dependable answer. Compare the exact formation with a practiced mirror (see the mirror insight), and check the map and upgrade differences.
Zorya/Athalia/Pandora beats Gwyn/Gunnar/Kulu while behind on upgrades, but struggles badly against other teamslow confidence
Zorya/Athalia/Pandora is against Gwyn/Gunnar/Kulu, and in the games where it was behind on upgrades, so the test that counts each game by its upgrade gap strongly supports the counter. With similar upgrades it is only , so the evidence rests on games where the Gwyn/Gunnar/Kulu side had more upgrades.
What to do Use Zorya only for a specific expected matchup with a team you have practiced. Check whether the opponent can respond with Evie/Orion or Saida/Contess/Twins.
Its Gwyn/Gunnar/Kulu results also differ by map: 5–3 on SL map 1, 6–5 on SL map 2 and 3–1 across the current SL maps 3 and 4, but 2–6 on SL map 5. Practice the intended formation and arena before using it as an alternative to Alna/Syph/Mehira.
It is against Evie/Orion and against Saida/Contess/Twins. Before choosing Zorya against Gwyn/Gunnar/Kulu, check whether the set can still win if the opponent changes the order and Zorya faces one of those teams.
Choose five teams with twenty-five distinct heroes before assigning maps. Players must reuse their chosen teams throughout Group and Swiss, for both Ro48 matches, and through Ro16 and Ro8. Ro4 and Finals allow new teams for each match.
A five-map starting set to practice
low confidence
Start practice with these twenty-five distinct heroes. Thador/Solise on map 4 is the weakest choice; adapt the teams and map order to the opponent before submitting them.
Starting map order
Map 1Evie/OrionRolan + Rowan
Map 2Saida/Contess/TwinsTwins + Gerda + Aliceth
Map 3Alna/Syph/MehiraZandrok + Hepler
Map 4Thador/SoliseWeakest choice; test carefully
Map 5Gwyn/Gunnar/KuluDunlingr + Kulu + Cyran
Evidence and reasoning
This map order and the allocation of heroes below both appear in the records. The order’s combined results include different team versions and opponents, so they do not predict how this exact set will do.
Five heroes per map
Map
Heroes
1
Evie, Orion, Perseus, Rolan, Rowan
2
Saida, Contess, Elijah & Lailah, Gerda, Aliceth
3
Alna, Sylphira, Mehira, Zandrok, Hepler
4
Thador, Solise, Frieren, Himmel, Lily May
5
Gwyneth, Gunnar, Dunlingr, Kulu, Cyran
Why this order
Evie/Orion starts on the open map 1. Saida/Contess/Twins takes map 2, Alna/Syph/Mehira takes map 3, and Gwyn/Gunnar/Kulu stays on map 5, where it has its largest winning sample. Thador/Solise fills map 4 with Frieren and Himmel, but its losing record there makes it the first team to reconsider. Compare the map results when testing another order.
Keep the teams compatible
Rolan and Rowan support Evie/Orion; Zandrok stays with Alna/Syph/Mehira; Dunlingr stays with Gwyn/Gunnar/Kulu. Replacing a team can require changing another: a Rolan/Twins team would take heroes from both Evie/Orion and Alna/Syph/Mehira in this set. Use the team group profiles for formations and flex choices.
Check how an opponent could beat it
A rearranged set built from recorded public teams can put Gwyn/Gunnar/Kulu against Evie/Orion, Evie/Orion against Alna/Syph/Mehira, Rolan/Twins against Thador/Solise, and an Athalia/Alna/Eryndor board against Gwyn/Gunnar/Kulu. The first two have statistical support, though Evie/Orion against Alna/Syph/Mehira is less certain. Thador/Solise is 0–7 against Rolan/Twins, with only one of those games near level; that small record does not establish a counter. The last pairing rests on four games. The exact formations still need testing on their new maps. Before submitting your teams, test changed enemy orders and teams chosen to beat yours. If two maps look likely to lose, the other three must all win; try a different team or map order and recheck the full set.
This team order on the current arenas: 1 set-up, 3–2 games · This exact allocation of heroes: 1 recorded set
Maps, results and data coverage
880 games in 176 matches, covering 104 players.
SL1, SL2, SL3, SL4 and SL5 are in the data. See the data limits
when comparing results across players and seasons.
Map1· every season, and the same arena as GD map 1: wide, open, no wallsMap2· every season: the widest gap between deployment zones of any board hereMap3 (SL1–3)· replaced after SL3Map3 (SL4–5)· tile 21 grants attack speed in a two-tile area enclosed by walls, open only toward the enemy half; check that a hero placed there can help without nearby alliesMap4 (SL1–3)· replaced after SL3Map4 (SL4–5)· the same arena as GD map 2: eight breakable wallsMap5· every season: breakable walls, with an attack-speed tile on each side
Win percentage (wins–losses) and appearances per map. Old and current arenas have separate columns where a map
position changed between seasons.
Team
SL-Map1
SL-Map2
SL-Map3 (SL1–3)
SL-Map3 (SL4–5)
SL-Map4 (SL1–3)
SL-Map4 (SL4–5)
SL-Map5
Gwyn/Gunnar/Kulu
57%(40–30)70 appearances
60%(31–21)52 appearances
60%(3–2)5 appearances
78%(25–7)32 appearances
89%(8–1)9 appearances
59%(16–11)27 appearances
73%(83–31)114 appearances
Evie/Orion
56%(44–35)79 appearances
56%(39–31)70 appearances
50%(2–2)4 appearances
56%(18–14)32 appearances
·
56%(20–16)37 appearances
43%(10–13)23 appearances
Saida/Contess/Twins
60%(25–17)42 appearances
70%(23–10)35 appearances
25%(1–3)4 appearances
44%(16–20)38 appearances
50%(4–4)8 appearances
67%(30–15)53 appearances
52%(26–24)52 appearances
Alna/Syph/Mehira
48%(26–28)54 appearances
55%(30–25)55 appearances
60%(3–2)5 appearances
63%(19–11)30 appearances
·
58%(22–16)38 appearances
62%(13–8)21 appearances
Thador/Solise
28%(5–13)18 appearances
42%(10–14)24 appearances
73%(11–4)15 appearances
39%(16–25)41 appearances
53%(10–9)19 appearances
41%(28–41)72 appearances
25%(5–15)20 appearances
Zorya/Athalia/Pandora
42%(8–11)19 appearances
33%(15–30)45 appearances
50%(1–1)2 appearances
13%(2–13)15 appearances
·
42%(5–7)12 appearances
43%(12–16)28 appearances
Rolan/Twins
79%(15–4)19 appearances
67%(2–1)3 appearances
·
100%(11–0)11 appearances
·
86%(6–1)7 appearances
47%(9–10)19 appearances
Athalia/Alna/Eryndor
50%(2–2)4 appearances
20%(2–8)10 appearances
·
60%(12–8)20 appearances
·
100%(1–0)1 appearances
·
Gwyn/Cyran/Hugin
14%(1–6)7 appearances
30%(3–7)10 appearances
0%(0–1)1 appearances
14%(1–6)7 appearances
·
·
0%(0–1)1 appearances
Marcille/Ravion/Thador
33%(2–4)6 appearances
0%(0–3)3 appearances
·
0%(0–2)2 appearances
50%(1–1)2 appearances
40%(2–3)5 appearances
33%(4–8)12 appearances
Satrana/Bonnie
0%(0–4)4 appearances
0%(0–4)4 appearances
17%(1–5)6 appearances
0%(0–1)1 appearances
0%(0–1)1 appearances
20%(1–4)5 appearances
17%(1–5)6 appearances
Alna/Athalia
0%(0–1)1 appearances
·
·
50%(5–5)10 appearances
·
·
33%(1–2)3 appearances
Eryndor/LM/Thador
100%(1–0)1 appearances
·
·
25%(1–3)4 appearances
·
50%(1–1)2 appearances
8%(1–11)12 appearances
Dun/Pandora/Cryonaia
·
100%(1–0)1 appearances
·
63%(5–3)8 appearances
·
33%(3–6)9 appearances
·
Rowan/Saida/Contess/Evie
·
·
0%(0–1)1 appearances
100%(7–0)7 appearances
25%(1–3)4 appearances
100%(2–0)2 appearances
·
Team group usage and records by league season
Each cell shows win percentage (wins–losses) and appearances. Seasons can have different heroes, players and stages represented; compare their results with that context.
Choose an order for three teams with fifteen distinct heroes. Guild Duel allows a fresh draft for each match.
Start with Contess, Gwyn/Gunnar/Kulu and Alna
medium confidence
Use this order as a starting point when you have no strong read on the opponent’s map choices. Choose fifteen distinct heroes and practice the full formations.
Starting map order
Map 1Saida/Contess/Twinsdecide whether Rolan goes here
Map 2Gwyn/Gunnar/KuluDunlingr
Map 3Alna/Syph/MehiraZandrok
Evidence and reasoning
This order keeps Gwyn/Gunnar/Kulu on its best-supported Guild Duel map and Alna/Syph/Mehira on map 3. The arrangement results compare it with other recorded choices.
Why these maps
Gwyn/Gunnar/Kulu has the largest winning sample on map 2. Alna/Syph/Mehira has winning records on both map 3 arenas, and Zandrok can clear breakable walls on the current one. Saida/Contess/Twins completes the set on map 1. Check the map records for the available alternatives.
When to change it
Evie/Orion is another map 1 candidate against an expected Alna, but it struggles against Gwyn/Gunnar/Kulu. Choose its supporting heroes together with the other teams so none is needed twice. Test an opponent putting Gwyn/Gunnar/Kulu against Contess and Alna against Gwyn/Gunnar/Kulu: losing those two maps would cost the match even if your map 3 team wins.
Against Contess/Gwyn/Alna, try Gwyn/Alna/Evie
medium confidence
Try this order when you expect Contess/Gwyn/Alna. Recheck the matchups if the opponent may change that order.
Against Contess / Gwyn / Alna
Map 1Gwyn/Gunnar/Kulutarget Contess
Map 2Alna/Syph/Mehiratarget Gwyn/Gunnar/Kulu
Map 3Evie/Orionuncertain against Alna
Evidence and reasoning
The counter recommendations support Gwyn/Gunnar/Kulu against Saida/Contess/Twins and Alna/Syph/Mehira against Gwyn/Gunnar/Kulu. Evie/Orion against Alna/Syph/Mehira is less certain.
Why try it
The first two pairings are the main reason to choose this order. Alna is a recorded answer to Gwyn/Gunnar/Kulu on map 2; simply replacing Gwyn/Gunnar/Kulu with Saida/Contess/Twins there has gone poorly. Keep Zandrok available for Alna and check whether Evie needs Rolan.
Example and limits
This order won all three maps against Contess/Gwyn/Alna in a recorded GD4 round 5 match, with upgrade advantages of 11.6, 11.6 and 7.7 points. That does not establish the same result with similar upgrades. The opponent can also change order to put Contess or another team that stops ultimates against Evie.
Maps, results and data coverage
198 games in 66 matches, covering 26 players.
GD1, GD2, GD3, GD4 and GD5 are in the data; GD3 round 3 is unavailable.
Older and current versions of map 3 have separate records. See the
upgrade recording assumptions when comparing events.
Map1· the same arena as SL Map1 in the current map order: wide, open, no wallsMap2· the same arena as SL Map4 in the current map order: eight breakable wallsMap3 (GD1 to GD3-2)· Guild Duel only: a permanent wall with two passesMap3 (GD3-4 to GD5)· Guild Duel only, map 3’s replacement ground from GD3 round 4: no permanent walls, seven breakable ones
Win percentage (wins–losses) and appearances per map. Tied games are listed separately.
Team
Map1
Map2
Map3 (GD1 to GD3-2)
Map3 (GD3-4 to GD5)
All GD
Gwyn/Gunnar/Kulu
61%(19–12)31 appearances
73%(40–15)55 appearances
38%(10–16)26 appearances
38%(3–5)8 appearances
60%(72–48)120 appearances
Alna/Syph/Mehira
31%(8–18)26 appearances
54%(7–6)1 tied14 appearances
70%(23–10)33 appearances
75%(12–4)16 appearances
57%(50–38)1 tied89 appearances
Saida/Contess/Twins
60%(15–10)1 tied26 appearances
30%(8–19)3 tied30 appearances
0%(0–1)1 appearances
30%(3–7)10 appearances
41%(26–37)4 tied67 appearances
Evie/Orion
48%(13–14)1 tied28 appearances
27%(3–8)11 appearances
33%(1–2)3 appearances
48%(10–11)21 appearances
44%(27–35)1 tied63 appearances
Thador/Solise
50%(2–2)4 appearances
29%(2–5)7 appearances
·
0%(0–1)1 appearances
33%(4–8)12 appearances
Alna/Athalia
0%(0–1)1 appearances
50%(3–3)6 appearances
·
·
43%(3–4)7 appearances
Zorya/Athalia/Pandora
0%(0–1)1 appearances
50%(1–1)2 appearances
·
0%(0–3)3 appearances
17%(1–5)6 appearances
Gwyn/Cyran/Hugin
33%(2–4)6 appearances
·
·
·
33%(2–4)6 appearances
Rolan/Twins
·
·
·
100%(2–0)2 appearances
100%(2–0)2 appearances
one-off line-ups
67%(6–3)9 appearances
0%(0–7)7 appearances
0%(0–5)5 appearances
67%(2–1)3 appearances
33%(8–16)24 appearances
For Supreme League results on the shared arenas, see the SL map table.
Arrangement results and team group usage by season
Each arrangement counts one three-team set per player per match, along with its game results. These totals describe recorded choices; they do not adjust for player skill or hero upgrades.
Arrangement
GD1 sets
GD1 results
GD2 sets
GD2 results
GD3 sets
GD3 results
GD4 sets
GD4 results
GD5 sets
GD5 results
All sets
All GD results
Gwyn/Gunnar/Kulu on map 2 and Alna/Syph/Mehira on map 3
6
71%(12–5)1 tied
13
74%(29–10)
5
53%(8–7)
10
73%(22–8)
3
78%(7–2)
37
71%(78–32)1 tied
Alna/Syph/Mehira on map 3, something else on map 2
1
100%(3–0)
4
42%(5–7)
4
75%(9–3)
·
·
3
56%(5–4)
12
61%(22–14)
Alna/Syph/Mehira opens on map 1
7
28%(5–13)3 tied
12
25%(9–27)
3
33%(3–6)
2
50%(3–3)
2
67%(4–2)
26
32%(24–51)3 tied
Other arrangements
12
47%(16–18)2 tied
1
67%(2–1)
12
44%(16–20)
18
37%(20–34)
14
40%(17–25)
57
42%(71–98)2 tied
Matchup records by map
Matchup (results for first team)
Map1
Map2
Map3 (GD1 to GD3-2)
Map3 (GD3-4 to GD5)
All GD
Alna/Syph/MehiravsGwyn/Gunnar/Kulu
33%(2–4)6 appearances
50%(4–4)8 appearances
80%(12–3)15 appearances
100%(4–0)4 appearances
67%(22–11)33 appearances
Gwyn/Gunnar/KuluvsSaida/Contess/Twins
60%(3–2)5 appearances
88%(15–2)17 appearances
·
100%(1–0)1 appearances
83%(19–4)23 appearances
Gwyn/Gunnar/KuluvsEvie/Orion
78%(7–2)9 appearances
80%(4–1)5 appearances
50%(1–1)2 appearances
67%(2–1)3 appearances
74%(14–5)19 appearances
Saida/Contess/TwinsvsEvie/Orion
60%(3–2)1 tied6 appearances
100%(2–0)2 appearances
·
40%(2–3)5 appearances
58%(7–5)1 tied13 appearances
Alna/Syph/MehiravsEvie/Orion
20%(1–4)5 appearances
50%(1–1)2 appearances
100%(1–0)1 appearances
50%(2–2)4 appearances
42%(5–7)12 appearances
Alna/Syph/MehiravsSaida/Contess/Twins
43%(3–4)7 appearances
50%(1–1)1 tied3 appearances
100%(1–0)1 appearances
100%(2–0)2 appearances
58%(7–5)1 tied13 appearances
Team group usage by Guild Duel season
Each cell shows win percentage (wins–losses) and appearances.
The three most-used tiles on each current SL map, as a share of Rolan's appearances there. Enemy-side positions are mirrored.
SL map 1SL map 2SL map 3SL map 4SL map 5
How his skills work
A Celestial support built around a team Morale meter. Battle starts at 20% Morale (more when the team uses heroes from one faction, most of all Lightbearers or Celestials, who also take 50% less damage for 8s); every 20% Morale gives every ally ATK worth 10% of Rolan's; every ally ultimate adds 10% (Elijah & Lailah's ultimate counts twice), each kill adds 10%, each death costs 10%, and if Rolan dies the meter is wiped.
At 60% Morale
His blessing on the two highest-damage allies (control-immune for 10s, +ATK) adds 50 Haste, and his heal also stuns the densest enemy cluster for 3s.
At 100% Morale
A radiant ward: each ally ultimate spawns a clone that casts it again, Morale can no longer drop while he lives, and allies inside are immune to stat reductions.
Ultimate
Cleanses allies, equalizes their HP percentages, grants +150% max HP decaying over 12s and a large heal, plus 15% Morale.
Skills from the app's text for Rolan; numbers rounded. The insights below explain how these skills affect team choices.
Placement advice
Placing Rolan farther back may help, but does not guarantee a winmedium confidence
Back-row placements have winning records, but the direct comparison between teams that both used Rolan is too close to establish that farther back is better.
What to do Place Rolan where his allies can protect him from the enemy’s opening attacks. The back corner is not always safest.
His death removes Morale, so protection matters. The two Rolans can also have different teammates. A corner can expose him to enemies that target the rearmost hero; check those skills and his allies’ protection before moving him farther back.
Evidence Rows from center: 4 rows 82–38–1t; 3 rows 102–40–1t; 2 rows 38–26; 1 row 1–2; center row 11–12 · Both teams used Rolan: Rolan farther back won 14 of 26
A mirror needs a practiced build and an upgrade checkmedium confidence
The player with the higher overall win rate won 102 of 147 decided mirrors within named team groups. That overall win rate includes these same games, so this is not an independent measure of skill. Where a mirror had an upgrade gap of at least one level, the advantaged side went 63–16 with three ties.
What to do Practice the full formation and check investment before choosing a mirror as the answer.
Using the same team group as the opponent is an option when no counter is established, but those observations do not make it an automatic win for the stronger player. Build differences, placement and battle-start interactions remain. Private 1v1 tests can help compare a flex change under otherwise similar conditions; they do not replace tournament evidence or establish the set’s minimum win chance.
Gwyneth: skills, placement and the best version of her team #
The three most-used tiles on each current SL map, as a share of Gwyneth's appearances there. Enemy-side positions are mirrored.
SL map 1SL map 2SL map 3SL map 4SL map 5
How her skills work
A Lightbearer marksman whose damage comes from her normal attacks, which run in a cycle of three. The second arrow deals heavy fire damage and burns the target, lowering its Vitality; the third silences the target for 1s and splashes lightning over every enemy within 2 tiles, stunning them briefly. With her Exclusive Equipment the first arrow carries both effects and all her normal attacks deal 20–30% more; at its third level that arrow also strikes for 10% of the target's max HP and ignores most of its physical defense.
Ultimate
Rains arrows over the tile with the most enemies and every tile within 2 of it, immobilizing them for 2s. Her charged normal attack also hits harder the more attack speed and Haste she has.
Distance
Hero Focus gives her attack speed during battle and more of it when no enemy has been within 1 tile of her for 3s; her Enhance Force awakening then shortens the gap between her normal attacks by 40% under the same condition. Everything in her kit rewards standing where enemies cannot reach her quickly, and Gunnar's bonus range to the ally directly behind him lets her hit from there.
Skills from the app's text for Gwyneth; numbers rounded.
Which version of the team does best?
The most-used exact five-hero line-ups of Gwyn/Gunnar/Kulu, out of 67 recorded. Near level keeps games with less than one paragon level's worth of upgrade difference. Players counts different people. Set-ups counts times they chose the lineup: if SL rules require the same teams for five matches (for example, in Group or Swiss), that counts as one set-up. Many games but few set-ups means fewer separate decisions to use the lineup.
Line-up
Players
19
47
Example: SL5 map 5
17
37
Example: SL5 map 5
9
33
Example: SL5 map 1
8
18
Example: SL5 map 4
8
26
Example: GD5 map 2
12
16
Example: SL5 map 5
3
16
Example: SL5 map 2
3
9
Example: SL5 map 1
5
11
Example: SL5 map 4
5
11
Example: SL5 map 1
–
4
8
Example: SL5 map 2
5
8
Example: SL5 map 5
Team and placement advice
Gwyneth, Gunnar, Kulu and Dunlingr with Aliceth is the best-supported versionhigh confidence
The Aliceth line-up is across 19 players and with similar upgrades. The Cyran line-up is , and with similar upgrades.
What to do Start from Gwyneth, Gunnar, Kulu, Dunlingr and Aliceth. Use Cyran as the fifth hero when Aliceth is needed elsewhere, and favor Dunlingr against teams that need early ultimates.
Favor Dunlingr against teams that need opening ultimates: across both formats the team group is with him and without him. These raw records do not isolate his contribution from upgrades, teammates or opponents. Kulu is on most boards and in both leading line-ups, and his wall changes the enemy’s paths, but the records do not require him: in Supreme League the team group is with Kulu and without him.
Aliceth has the largest sample and the widest use. Peggy is , but only three players used her and just is near level. That higher raw win rate does not make her the safer choice. Contess () and Frieren () were used mostly in Guild Duel; check the exact-lineup table when choosing alternatives.
Evidence Line-ups compared: the 12 most-used exact line-ups, out of 67 recorded
Keep Gwyneth in the back two rows, directly behind Gunnarhigh confidence
Inside Gwyn/Gunnar/Kulu she is in the back two rows and one row further forward.
What to do Place Gunnar in the third row from the center line or deeper, with Gwyneth directly behind him. Bringing her forward to reach a target costs the attack speed her kit is built on.
Hero Focus always grants attack speed. Its extra bonus and her awakening’s 40% shorter time between attacks require 3 seconds without an enemy within 1 tile. A protected back-row tile helps her keep those bonuses. Gunnar gives the ally directly behind him extra range, attack and steadfast, which is why she stands there on almost every board. When you move the pair, move them together.
Evidence Directly behind Gunnar: 420 of 425 boards inside the team group
Evie: where to start her on each Supreme League map #
How her skills work
When the battle starts, Intel Chase teleports Evie to the tile on the enemy half that mirrors her own and hides her there, invincible. She immediately gathers intel on every enemy hero within 1 tile of where she lands, then on the rest one at a time from nearest to farthest, 4.5s each, gaining 150 Energy per hero. Her Detection meter fills at 120 a second and every ally ultimate takes 500 off it; at 1000 she is revealed where she stands and loses her invincibility. Once every enemy is investigated she teleports back to her own tile.
Why the landing tile matters
Without ally ultimates she has about 8s hidden. Enemies next to her landing tile are investigated together in the first 4.5s; each enemy farther away costs another 4.5s. Every investigated hero loses 20% Magic DEF for the battle, and investigating all of them while hidden takes another 25% off every enemy, makes them deal 25–35% less damage through her Exclusive Equipment, and with her awakening adds a second healing quill. Landing next to the enemy line is how she finishes in time.
Active skills
Her ultimate interrogates the enemy with the most damage dealt, immobilizing it for 6–8s and silencing it if already investigated; Pointed Proof dashes to the nearest enemy for two dagger strikes; Foretold Favor's quill heals and buffs the ally dealing the most damage.
Skills from the app's text for Evie; numbers rounded. The landing tile is the point-symmetric tile on the enemy half, the same mapping this report uses to mirror enemy positions.
Amber tiles are her three most-used starts on each map, with her record from each. The
blue tile is the start whose landing tile has had the most enemy heroes next to it, averaged over
every enemy deployment recorded on that map, whatever team Evie's side played; it is drawn as a blue
border when it is also one of her starts. The opponent's team changes where enemies stand, so treat the
blue tile as the usual case, not a rule.
SL map 1· 81 Evie boards. Blue: tile 7, next to 2.9 enemies on average over 352 recorded enemy deployments.SL map 2· 93 Evie boards. Blue: tile 7, next to 3.2 enemies on average over 352 recorded enemy deployments.SL map 3· 49 Evie boards. Blue: tile 4, next to 3.1 enemies on average over 296 recorded enemy deployments.SL map 4· 40 Evie boards. Blue: tile 4, next to 3.0 enemies on average over 296 recorded enemy deployments.SL map 5· 41 Evie boards. Blue: tile 7, next to 2.8 enemies on average over 352 recorded enemy deployments.
Placement advice
Start Evie where her landing tile touches the enemy linemedium confidence
Across all recorded enemy formations on each SL arena, tile 4’s landing spot touches 2.7, 3.0 and 3.1 enemies on maps 1, 2 and 3. These averages include games without Evie. On map 5, tile 3 touches only 1.4 enemies on average.
What to do On maps 1, 2 and 3 keep her on tile 4, or on tile 7 on maps 1 and 2 when the formation allows it. On map 4 prefer tile 4 to tile 2, and on map 5 test tile 7 instead of tile 3.
She investigates nearby enemies together, then the remaining enemies one at a time. Each investigation takes 4.5 seconds. Allied ultimates lower her Detection meter and help her stay hidden, but their timing matters. Landing beside more enemies reduces the time she needs; it does not guarantee she finishes before being revealed.
On map 4, the same all-formation comparison gives tile 2 an average of 1.3 nearby enemies, against 3.0 for tile 4. Her landing spot is the matching tile on the enemy half after rotating the board halfway around. Check the actual enemy positions when available; the averages cannot account for a different formation.
Evidence Start whose landing tile touches the most enemies: map 1 tile 7 (2.9); map 2 tile 7 (3.2); map 3 tile 4 (3.1); map 4 tile 4 (3.0); map 5 tile 7 (2.8), averaged over every recorded enemy deployment on the arena
Her results by start tile do not separate the tiles yetlow confidence
Tile 4 is on map 1, on map 2 and on map 3. The second choices, tile 5 on maps 1 and 2 ( and ) and tile 6 on map 3 (), have better records from smaller samples that land next to fewer enemies.
What to do Choose the tile by the enemies it will touch, then check that map’s records in the table above.
Those records mix players, opponents, upgrades and the rest of her team, and the smaller samples are too few to show that the second choices are better tiles. On map 4, tile 2 is despite landing far from the enemy, and the three boards that started her on the forward tile 14 all won: those wins do not show whether another tile would have worked better. Use nearby-enemy counts to choose a starting placement, then test it against the intended opponent.
Evidence Map 4 forward tile 14: 3 boards, 3–0, landing next to no enemy
The full hero table
Solid chips mean core, outlined chips mean flex.
Presence is the share of the 1078 games they appear in, and since both sides can field the
same hero the ceiling is 200%. Lift is their win rate minus the average win rate of the players
who brought them, a rough way to tell a strong hero from a hero that strong players like.
These results include matches from multiple contributors and leagues, analyzed without replays. Many matchups
have few games. Player skill, hero upgrades and the rules used to group similar teams can affect the
results. Ties are excluded from win rates. Read these limits before relying on a particular number.
The caveats in full
Players and teams are represented unevenly. A team's results can depend on who used it,
their opponents and their upgrades. Games where a player must reuse the same teams are related results, not separate tests of a new team choice.
Check the matchup's player and format breakdowns before applying its overall record to another situation.
Tied maps are left out of win rates. The game gives a tied battle to one player, the
defending side, and the match lists color a tied map for that player. Winning a map on a tie does
not show that the team won the battle, so every win rate leaves ties out.
Missing upgrade levels use match-wide rules. Paragon levels are recorded in
944 games and refinement levels in 549.
If a match has no recorded levels for one upgrade type, every hero is treated as level 4 for that type.
If any levels are recorded, a hero without a row is level 0.
The strength section explains the upgrade scale and adjusted ratings, and
holds each team's record with the upgrade advantage, behind it and near level.