Team groups, counters and map records from Supreme League and Guild Duel matches.
Everything here was produced by an automated analysis of saved tournament results and hero skill descriptions, and no replays were reviewed. Most records come from top-spending players at the top end of the meta, with heavily upgraded rosters, so results may differ with fewer upgrades. More about the data's limits.
Last updated 2026-09-22 · Season 7 records
A team group is a set of related five-hero lineups, named after the heroes they share. Groups are sorted by strength tier, then by appearances, and each column heading explains its figure. See how team groups and common variations are defined and how strength is measured. Each team's profile holds its full ratings and upgrade records.
| Artifacts | Used on | Win rate |
|---|---|---|
| Wing Guard | 56%287 teams | 66%188–99 |
| Vanguard | 38%197 teams | 58%115–82 |
| Valorshield | 2%8 teams | 75%6–2 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
| Heroes | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 300 | 68%(204–96) | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 25 | 64%(16–9) | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 17 | 12%(2–15) | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| + 1 flex | 9 | 67%(6–3) | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 8 | 13%(1–7) | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Exact five-hero teams outside the variations above, including uncommon and losing lineups. Every team here has limited evidence and a concentrated sample, which means few games and one player behind a large share of them.
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Kulu | 65%(253–138) | 51%(62–59) |
| Dunlingr | 68%(243–117) | 47%(72–80) |
| Cyran | 55%(85–69) | 64%(230–128) |
| Aliceth | 64%(96–54) | 60%(219–143) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Valorshield | 74%244 teams | 59%142–100 |
| Frontline | 15%48 teams | 42%20–28 |
| Vanguard | 7%22 teams | 50%11–11 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
| Heroes | ||
|---|---|---|
| 125 | 65%(81–43) | |
Exact five-hero teams outside the variations above, including uncommon and losing lineups. Most teams here have limited evidence and a concentrated sample, which means few games and one player behind a large share of them.
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Perseus | 53%(132–117) | 57%(46–34) |
| Rolan | 63%(121–70) | 41%(57–81) |
| Frieren | 53%(101–90) | 56%(77–61) |
| Zandrok | 48%(60–65) | 58%(118–86) |
| Rowan | 53%(55–49) | 55%(123–102) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Wing Guard | 28%90 teams | 53%40–36 |
| Breakthrough | 28%89 teams | 54%45–39 |
| Surging | 23%72 teams | 58%42–30 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
| Heroes | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 300 | 53%(147–133) | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Exact five-hero teams outside the variations above, including uncommon and losing lineups. Every team here has limited evidence and a concentrated sample, which means few games and one player behind a large share of them.
| Lineup | Players | |||
|---|---|---|---|---|
| 3 | 100%(3–0) | 100%(1–0) | 2 |
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Gerda | 52%(66–62) | 54%(91–76) |
| Aliceth | 59%(72–50) | 49%(85–88) |
| Solise | 41%(30–43) | 57%(127–95) |
| Rolan | 75%(44–15) | 48%(113–123) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Frontline | 88%277 teams | 59%162–114 |
| Vanguard | 5%15 teams | 53%8–7 |
| Breakthrough | 3%8 teams | 38%3–5 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
| Heroes | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 289 | 60%(174–114) | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Zandrok | 61%(93–59) | 55%(88–72) |
| Hepler | 47%(51–57) | 64%(130–74) |
| Cyran | 56%(58–45) | 59%(123–86) |
| Athalia | 52%(29–27) | 59%(152–104) |
| Frieren | 69%(33–15) | 56%(148–116) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Frontline | 45%35 teams | 69%24–11 |
| Wing Guard | 38%30 teams | 73%22–8 |
| Windcall | 5%4 teams | 75%3–1 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Sylphira | 67%(31–15) | 75%(24–8) |
| Mehira | 65%(22–12) | 75%(33–11) |
| Taichi & Agumon | 82%(27–6) | 62%(28–17) |
| Contess | 84%(26–5) | 62%(29–18) |
| Athalia | 76%(19–6) | 68%(36–17) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Magicsurge | 56%20 teams | 55%11–9 |
| Surging | 39%14 teams | 57%8–6 |
| Frontline | 6%2 teams | 0%0–2 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Shadewing | 52%(14–13) | 56%(5–4) |
| Igor | 48%(10–11) | 60%(9–6) |
| Zorya | 60%(9–6) | 48%(10–11) |
| Mehira | 43%(3–4) | 55%(16–13) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Vanguard | 35%86 teams | 56%48–38 |
| Breakthrough | 26%64 teams | 29%18–44 |
| Wing Guard | 15%36 teams | 40%14–21 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
Exact five-hero teams outside the variations above, including uncommon and losing lineups. Most teams here have limited evidence and a concentrated sample, which means few games and one player behind a large share of them.
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Frieren | 42%(73–100) | 38%(27–45) |
| Lily May | 41%(67–96) | 40%(33–49) |
| Himmel | 45%(55–68) | 37%(45–77) |
| Ravion | 36%(36–65) | 44%(64–80) |
| Contess | 45%(19–23) | 40%(81–122) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Magicsurge | 62%91 teams | 30%27–64 |
| Surging | 13%19 teams | 37%7–12 |
| Vine Snare | 9%13 teams | 54%7–6 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Callan | 38%(41–67) | 23%(9–30) |
| Shadewing | 36%(39–69) | 28%(11–28) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Breakthrough | 41%11 teams | 9%1–10 |
| Valorshield | 22%6 teams | 33%2–4 |
| Wing Guard | 22%6 teams | 67%4–2 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Ravion | 42%(5–7) | 27%(4–11) |
| Taichi & Agumon | 17%(2–10) | 47%(7–8) |
| Lyca | 11%(1–8) | 44%(8–10) |
| Solise | 44%(4–5) | 28%(5–13) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Frontline | 64%14 teams | 36%5–9 |
| Breakthrough | 32%7 teams | 71%5–2 |
| Vanguard | 5%1 team | 100%1–0 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
No set of heroes was recorded together eight times, so this group has no common variations.
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Taichi & Agumon | 53%(9–8) | 40%(2–3) |
| Mehira | 43%(6–8) | 63%(5–3) |
| Hepler | 44%(4–5) | 54%(7–6) |
| Sylphira | 57%(4–3) | 47%(7–8) |
| Rolan | 50%(3–3) | 50%(8–8) |
| Rowan | 75%(3–1) | 44%(8–10) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Vanguard | 50%10 teams | 20%2–8 |
| Wing Guard | 30%6 teams | 83%5–1 |
| Vine Snare | 15%3 teams | 33%1–2 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
| Heroes | ||
|---|---|---|
| 9 | 22%(2–7) |
Exact five-hero teams outside the variations above, including uncommon and losing lineups. Every team here has limited evidence and a concentrated sample, which means few games and one player behind a large share of them.
| Lineup | Players | |||
|---|---|---|---|---|
| 3 | 67%(2–1) | 100%(1–0) | 2 |
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Aliceth | 23%(3–10) | 86%(6–1) |
| Ravion | 23%(3–10) | 86%(6–1) |
| Alna | 86%(6–1) | 23%(3–10) |
| Cyran | 75%(3–1) | 38%(6–10) |
| Niru | 100%(3–0) | 35%(6–11) |
| Taichi & Agumon | 33%(1–2) | 47%(8–9) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Breakthrough | 63%12 teams | 42%5–7 |
| Frontline | 26%5 teams | 40%2–3 |
| Valorshield | 11%2 teams | 50%1–1 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Koko | 43%(6–8) | 40%(2–3) |
| Zandrok | 38%(3–5) | 45%(5–6) |
| Kordan | 40%(2–3) | 43%(6–8) |
| Gerda | 67%(2–1) | 38%(6–10) |
| Perseus | 33%(1–2) | 44%(7–9) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Magicsurge | 53%8 teams | 25%2–6 |
| Valorshield | 20%3 teams | 67%2–1 |
| Surging | 13%2 teams | 50%1–1 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
No set of heroes was recorded together eight times, so this group has no common variations.
Exact five-hero teams outside the variations above, including uncommon and losing lineups. Every team here has limited evidence and a concentrated sample, which means few games and one player behind a large share of them.
| Lineup | Players | |||
|---|---|---|---|---|
| 3 | 67%(2–1) | 0%(0–1) | 1 |
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Alna | 22%(2–7) | 50%(3–3) |
| Damian | 29%(2–5) | 38%(3–5) |
| Contess | 50%(3–3) | 22%(2–7) |
| Callan | 67%(2–1) | 25%(3–9) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Valorshield | 53%8 teams | 75%6–2 |
| Vanguard | 33%5 teams | 20%1–4 |
| Enlightening | 7%1 team | 0%0–1 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Rolan | 67%(6–3) | 17%(1–5) |
| Aliceth | 100%(4–0) | 27%(3–8) |
| Contess | 25%(1–3) | 55%(6–5) |
| Kulu | 50%(2–2) | 45%(5–6) |
| Alna | 0%(0–3) | 58%(7–5) |
| Reinier | 0%(0–3) | 58%(7–5) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Breakthrough | 64%9 teams | 100%9–0 |
| Windcall | 29%4 teams | 25%1–3 |
| Magicsurge | 7%1 team | 0%0–1 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
| Heroes | ||
|---|---|---|
| 9 | 100%(9–0) |
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Rolan | 100%(9–0) | 20%(1–4) |
| Baelran | 20%(1–4) | 100%(9–0) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Surging | 52%16 teams | 6%1–15 |
| Magicsurge | 19%6 teams | 17%1–5 |
| Vine Snare | 16%5 teams | 0%0–5 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
Exact five-hero teams outside the variations above, including uncommon and losing lineups. Every team here has limited evidence and a concentrated sample, which means few games and one player behind a large share of them.
| Lineup | Players | |||
|---|---|---|---|---|
| 2 | 0%(0–2) | – | 2 |
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Zorya | 10%(2–19) | 10%(1–9) |
| Daimon | 14%(2–12) | 6%(1–16) |
| Igor | 7%(1–13) | 12%(2–15) |
| Niru | 0%(0–12) | 16%(3–16) |
| Callan | 14%(1–6) | 8%(2–22) |
| Cyran | 17%(1–5) | 8%(2–23) |
| Shadewing | 17%(1–5) | 8%(2–23) |
| Athalia | 20%(1–4) | 8%(2–24) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Vanguard | 32%9 teams | 44%4–5 |
| Valorshield | 32%9 teams | 11%1–8 |
| Breakthrough | 21%6 teams | 50%3–3 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Lily May | 13%(2–13) | 54%(7–6) |
| Solise | 27%(3–8) | 35%(6–11) |
| Dunlingr | 40%(4–6) | 28%(5–13) |
| Lyca | 40%(4–6) | 28%(5–13) |
| Perseus | 50%(3–3) | 27%(6–16) |
| Artifacts | Used on | Win rate |
|---|---|---|
| Breakthrough | 43%6 teams | 33%2–4 |
| Frontline | 43%6 teams | 0%0–6 |
| Wing Guard | 14%2 teams | 0%0–2 |
Every listed hero appears in every team counted. Flex slots are filled by the options inside each row. One team can be counted in more than one row, so the rows can add up to more than the group's appearances.
| Heroes | ||
|---|---|---|
| 8 | 0%(0–8) |
Exact five-hero teams outside the variations above, including uncommon and losing lineups. Every team here has limited evidence and a concentrated sample, which means few games and one player behind a large share of them.
| Lineup | Players | |||
|---|---|---|---|---|
| 2 | 0%(0–2) | 0%(0–1) | 2 |
Overall results with and without each flex hero, across all variations. Teammates, maps, players and opponents can differ, so this does not measure the effect of swapping one hero alone.
| Hero | ||
|---|---|---|
| Zandrok | 9%(1–10) | 33%(1–2) |
| Lily May | 0%(0–10) | 50%(2–2) |
| Dionel | 33%(1–2) | 9%(1–10) |
| Solise | 33%(1–2) | 9%(1–10) |
These teams fall outside the named team groups. Some have been played more than once.
Lineups outside the named groups that beat an established team rated average or better without being ahead in upgrades. Each row is one exact five-hero lineup. Treat each row as a team worth trying out. Most have only a handful of games, so none is proven. Select a map name to see that board.
| Lineup | Beat | ||
|---|---|---|---|
| 67%(2–1) | Alna/ Gwyn/ | ||
| 100%(3–0) | Saida/ |
These comparisons combine all lineups within each team group. They cannot show how much any single hero matters on its own.
Teams are listed by strength, strongest on top. A line joins a team to a team it beats. The arrowhead points at the loser, and the label is the winner's share of those games. Select a line to open that counter's row. Select a team to show only its lines.
A line is drawn where the counter test supports the counter. A grey dashed line marks a possible counter that the test does not confirm. A label is dimmed under eight games. The exact wins and losses show under each label when a team is selected, and in the table below. Line style and label show the raw record, which can include wins helped by an upgrade advantage.
Every supported and possible counter is shown. Two teams with no line between them do not have enough evidence of a counter.
Each row is one team's win rate against each opponent, with wins–losses underneath. Bold figures are the matchups the counter test supports. 155 games with teams from the same group on both sides are left out.
| Gwyn/ | Orion/ | Saida/ | Alna/ | Thador/ | Zorya/ | Rolan/ | Alna/ | Satrana/ | Marcille/ | Thador/ | Alna/ | Dun/ | Orion/ | Saida/ | Frieren/ | Lyca/ | Saida/ | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Gwyn/ | 67%42–21 | 76%50–16 | 35%26–49 | 87%34–5 | 43%20–26 | 67%10–5 | 29%2–5 | 100%4–0 | 75%6–2 | 100%4–0 | 100%4–0 | 100%5–0 | 83%5–1 | 60%3–2 | 63%5–3 | 100%1–0 | 50%1–1 | |
| Orion/ | 33%21–42 | 35%17–31 | 62%28–17 | 73%19–7 | 100%25–0 | 25%4–12 | 80%4–1 | 100%4–0 | 100%1–0 | 100%2–0 | 100%1–0 | 50%1–1 | 50%2–2 | 100%1–0 | 0%0–1 | · | 0%0–3 | |
| Saida/ | 24%16–50 | 65%31–17 | 49%21–22 | 59%22–15 | 100%14–0 | 38%3–5 | 60%3–2 | 100%6–0 | 0%0–1 | 50%2–2 | 33%2–4 | 0%0–2 | 100%1–0 | 100%1–0 | 50%1–1 | 50%1–1 | 50%1–1 | |
| Alna/ | 65%49–26 | 38%17–28 | 51%22–21 | 76%16–5 | 72%13–5 | 17%1–5 | 33%2–4 | 100%3–0 | 75%3–1 | 67%2–1 | · | 67%2–1 | 67%2–1 | 0%0–2 | 100%1–0 | 100%3–0 | · | |
| Thador/ | 13%5–34 | 27%7–19 | 41%15–22 | 24%5–16 | 57%4–3 | 0%0–7 | 75%3–1 | 100%5–0 | 80%4–1 | 67%2–1 | 0%0–3 | · | 0%0–1 | 67%2–1 | 0%0–1 | 100%3–0 | 33%1–2 | |
| Zorya/ | 57%26–20 | 0%0–25 | 0%0–14 | 28%5–13 | 43%3–4 | 0%0–3 | 0%0–2 | 100%1–0 | · | 50%1–1 | 0%0–1 | · | 0%0–1 | · | · | 100%2–0 | · | |
| Rolan/ | 33%5–10 | 75%12–4 | 63%5–3 | 83%5–1 | 100%7–0 | 100%3–0 | · | · | · | 100%4–0 | 100%2–0 | · | 100%1–0 | · | 50%1–1 | · | · | |
| Alna/ | 71%5–2 | 20%1–4 | 40%2–3 | 67%4–2 | 25%1–3 | 100%2–0 | · | · | · | · | · | · | · | · | · | · | · | |
| Satrana/ | 0%0–4 | 0%0–4 | 0%0–6 | 0%0–3 | 0%0–5 | 0%0–1 | · | · | 0%0–2 | · | · | · | · | · | · | · | · | |
| Marcille/ | 25%2–6 | 0%0–1 | 100%1–0 | 25%1–3 | 20%1–4 | · | · | · | 100%2–0 | · | · | · | 0%0–1 | · | · | · | · | |
| Thador/ | 0%0–4 | 0%0–2 | 50%2–2 | 33%1–2 | 33%1–2 | 50%1–1 | 0%0–4 | · | · | · | · | · | · | 100%1–0 | · | · | · | |
| Alna/ | 0%0–4 | 0%0–1 | 67%4–2 | · | 100%3–0 | 100%1–0 | 0%0–2 | · | · | · | · | 0%0–2 | · | 100%1–0 | · | · | · | |
| Dun/ | 0%0–5 | 50%1–1 | 100%2–0 | 33%1–2 | · | · | · | · | · | · | · | 100%2–0 | · | · | · | · | 0%0–1 | |
| Orion/ | 17%1–5 | 50%2–2 | 0%0–1 | 33%1–2 | 100%1–0 | 100%1–0 | 0%0–1 | · | · | 100%1–0 | · | · | · | · | · | · | · | |
| Saida/ | 40%2–3 | 0%0–1 | 0%0–1 | 100%2–0 | 33%1–2 | · | · | · | · | · | 0%0–1 | 0%0–1 | · | · | · | · | · | |
| Frieren/ | 38%3–5 | 100%1–0 | 50%1–1 | 0%0–1 | 100%1–0 | · | 50%1–1 | · | · | · | · | · | · | · | · | · | · | |
| Lyca/ | 0%0–1 | · | 50%1–1 | 0%0–3 | 0%0–3 | 0%0–2 | · | · | · | · | · | · | · | · | · | · | · | |
| Saida/ | 50%1–1 | 100%3–0 | 50%1–1 | · | 67%2–1 | · | · | · | · | · | · | · | 100%1–0 | · | · | · | · |
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, and each game's upgrade gap also counts.
All win rates below are for the team in the first column. Wins–losses appear in parentheses.
| Team | Counter | Verdict | |||||
|---|---|---|---|---|---|---|---|
| Gwyn/ |
Alna/ |
35%(26–49) | 71%(228–94) | 24%(8–25) | −36 | under 1 in 1,000 | supported |
| Saida/ |
Gwyn/ |
24%(16–50) | 65%(124–67) | 21%(7–27) | −41 | under 1 in 1,000 | supported |
| Zorya/ |
Orion/ |
0%(0–25) | 40%(43–64) | 0%(0–7) | −40 | under 1 in 1,000 | supported |
| Gwyn/ |
Zorya/ |
43%(20–26) | 70%(209–89) | 60%(6–4) | −27 | under 1 in 1,000 | supported |
| Orion/ |
Gwyn/ |
33%(21–42) | 62%(124–75) | 26%(6–17) | −29 | under 1 in 1,000 | supported |
| Thador/ |
Gwyn/ |
13%(5–34) | 45%(72–87) | 11%(2–17) | −32 | under 1 in 1,000 | supported |
| Alna/ |
Orion/ |
38%(17–28) | 64%(137–78) | 33%(6–12) | −26 | around 1 in 695 | supported |
| Orion/ |
Saida/ |
35%(17–31) | 60%(127–84) | 15%(2–11) | −25 | around 1 in 194 | supported |
| Rolan/ |
Gwyn/ |
33%(5–10) | 80%(37–9) | 43%(3–4) | −47 | around 1 in 60 | supported |
| Orion/ |
Rolan/ |
25%(4–12) | 60%(128–85) | 33%(1–2) | −35 | around 1 in 34 | supported |
| Orion/ |
Gwyn/ |
17%(1–5) | 45%(5–6) | –(0–0) | −29 | around 1 in 31 | supported |
| Alna/ |
Alna/ |
33%(2–4) | 66%(51–26) | 50%(1–1) | −33 | around 1 in 24 | supported |
| Zorya/ |
Saida/ |
0%(0–14) | 34%(33–64) | 0%(0–3) | −34 | around 1 in 17 | possible, unconfirmed |
| Gwyn/ |
Alna/ |
29%(2–5) | 58%(96–69) | 0%(0–1) | −30 | around 1 in 15 | possible, unconfirmed |
| Dun/ |
Gwyn/ |
0%(0–5) | 57%(8–6) | 0%(0–4) | −57 | around 1 in 11 | possible, unconfirmed |
| Thador/ |
Orion/ |
27%(7–19) | 43%(54–73) | 30%(3–7) | −16 | around 1 in 11 | possible, unconfirmed |
| Alna/ |
Rolan/ |
17%(1–5) | 60%(99–65) | 0%(0–2) | −44 | around 1 in 10 | possible, unconfirmed |
| Thador/ |
Rolan/ |
0%(0–7) | 36%(33–59) | 0%(0–1) | −36 | around 1 in 7 | possible, unconfirmed |
12 matchups have statistical support, and 6 more are possible but unconfirmed. How reliable are these counters?
Which Supreme League maps each team group was used on, and how it did there. A match is played on five maps.
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. Ro4 and Finals allow new teams for each match.
1005 games in 201 matches, played by 115 players.
The league seasons SL1, SL2, SL3, SL4 and SL5 are in the data. See the data limits when comparing results across players and league seasons.
Win percentage (wins–losses) and appearances per map. Old and current arenas have separate columns where a map position changed between league seasons.
| Team | SL map 1 | SL map 2 | SL map 3 (SL1–3) | SL map 3 (SL4–5) | SL map 4 (SL1–3) | SL map 4 (SL4–5) | SL map 5 |
|---|---|---|---|---|---|---|---|
| Gwyn/ | 52%(44–40)84 appearances | 54%(38–33)71 appearances | 56%(5–4)9 appearances | 67%(30–15)45 appearances | 89%(8–1)9 appearances | 61%(25–16)41 appearances | 72%(91–36)127 appearances |
| Orion/ | 55%(52–43)95 appearances | 62%(46–28)74 appearances | 33%(1–2)3 appearances | 56%(19–15)34 appearances | · | 53%(19–17)37 appearances | 56%(14–11)25 appearances |
| Saida/ | 59%(26–18)44 appearances | 71%(24–10)36 appearances | 25%(1–3)4 appearances | 47%(20–23)47 appearances | 50%(4–4)8 appearances | 64%(28–16)54 appearances | 50%(27–27)56 appearances |
| Alna/ | 49%(29–30)59 appearances | 60%(32–21)53 appearances | 75%(3–1)4 appearances | 66%(25–13)38 appearances | · | 59%(26–18)44 appearances | 63%(17–10)27 appearances |
| Thador/ | 32%(8–17)25 appearances | 42%(13–18)31 appearances | 73%(11–4)15 appearances | 38%(18–29)47 appearances | 53%(10–9)19 appearances | 41%(31–44)78 appearances | 24%(5–16)21 appearances |
| Zorya/ | 45%(10–12)22 appearances | 32%(16–34)50 appearances | 50%(1–1)2 appearances | 18%(3–14)17 appearances | · | 47%(9–10)19 appearances | 32%(10–21)31 appearances |
| Rolan/ | 67%(16–8)24 appearances | 75%(3–1)4 appearances | · | 92%(12–1)13 appearances | · | 89%(8–1)9 appearances | 52%(13–12)25 appearances |
| Alna/ | 50%(2–2)4 appearances | 25%(2–6)8 appearances | · | 59%(13–9)22 appearances | · | 100%(1–0)1 appearance | 100%(1–0)1 appearance |
| Satrana/ | 0%(0–5)5 appearances | 0%(0–5)5 appearances | 17%(1–5)6 appearances | 0%(0–1)1 appearance | 0%(0–1)1 appearance | 14%(1–6)7 appearances | 17%(1–5)6 appearances |
| Marcille/ | 33%(2–4)6 appearances | 0%(0–2)2 appearances | · | 0%(0–2)2 appearances | 50%(1–1)2 appearances | 40%(2–3)5 appearances | 36%(4–7)11 appearances |
| Thador/ | 50%(2–2)4 appearances | 25%(1–3)4 appearances | · | 40%(2–3)5 appearances | · | 67%(2–1)3 appearances | 18%(2–9)11 appearances |
| Alna/ | 0%(0–1)1 appearance | · | · | 50%(5–5)10 appearances | · | · | 50%(2–2)4 appearances |
| Dun/ | · | 100%(1–0)1 appearance | · | 56%(5–4)9 appearances | · | 30%(3–7)10 appearances | · |
| Orion/ | 100%(2–0)2 appearances | 14%(1–6)7 appearances | 100%(1–0)1 appearance | 67%(2–1)3 appearances | · | 100%(1–0)1 appearance | 20%(1–4)5 appearances |
| Saida/ | 0%(0–1)1 appearance | 100%(1–0)1 appearance | · | 0%(0–2)2 appearances | · | 67%(2–1)3 appearances | 50%(1–1)2 appearances |
| Frieren/ | 50%(2–2)4 appearances | 33%(1–2)3 appearances | 0%(0–1)1 appearance | 67%(4–2)6 appearances | · | 0%(0–1)1 appearance | · |
| Lyca/ | · | 0%(0–2)2 appearances | 100%(1–0)1 appearance | · | 0%(0–2)2 appearances | 0%(0–4)4 appearances | 20%(1–4)5 appearances |
| Saida/ | · | · | 0%(0–1)1 appearance | 100%(7–0)7 appearances | 25%(1–3)4 appearances | 100%(2–0)2 appearances | · |
Each cell shows win percentage (wins–losses) and appearances. Different league seasons can include different heroes, players and stages, so keep that in mind when comparing them.
| Team | SL1 | SL2 | SL3 | SL4 | SL5 | Total |
|---|---|---|---|---|---|---|
| Gwyn/ | 100%(2–0)2 appearances | 57%(20–15)35 appearances | 69%(9–4)13 appearances | 67%(68–34)102 appearances | 61%(142–92)234 appearances | 386 |
| Orion/ | · | · | 57%(4–3)7 appearances | 52%(35–32)67 appearances | 58%(112–81)194 appearances | 268 |
| Saida/ | 0%(0–2)2 appearances | 59%(13–9)22 appearances | 55%(6–5)11 appearances | 55%(34–28)62 appearances | 57%(77–57)152 appearances | 249 |
| Alna/ | 50%(1–1)2 appearances | 73%(16–6)22 appearances | 50%(3–3)6 appearances | 60%(37–25)62 appearances | 56%(75–58)133 appearances | 225 |
| Thador/ | 100%(1–0)1 appearance | 56%(19–15)34 appearances | 56%(5–4)9 appearances | 39%(26–41)67 appearances | 37%(45–77)125 appearances | 236 |
| Zorya/ | 100%(1–0)1 appearance | 50%(9–9)18 appearances | 33%(1–2)3 appearances | 21%(6–22)28 appearances | 35%(32–59)91 appearances | 141 |
| Rolan/ | · | · | · | 60%(12–8)20 appearances | 73%(40–15)55 appearances | 75 |
| Alna/ | · | · | · | · | 53%(19–17)36 appearances | 36 |
| Satrana/ | · | 9%(1–10)11 appearances | · | 13%(1–7)8 appearances | 8%(1–11)12 appearances | 31 |
| Marcille/ | · | 60%(3–2)5 appearances | 67%(2–1)3 appearances | 33%(3–6)9 appearances | 9%(1–10)11 appearances | 28 |
| Thador/ | · | · | · | 0%(0–1)1 appearance | 35%(9–17)26 appearances | 27 |
| Alna/ | · | · | 0%(0–2)2 appearances | 38%(3–5)8 appearances | 80%(4–1)5 appearances | 15 |
| Dun/ | · | · | · | 47%(8–9)17 appearances | 33%(1–2)3 appearances | 20 |
| Orion/ | · | 0%(0–1)1 appearance | 67%(2–1)3 appearances | 75%(3–1)4 appearances | 27%(3–8)11 appearances | 19 |
| Saida/ | · | · | 50%(1–1)2 appearances | 25%(1–3)4 appearances | 67%(2–1)3 appearances | 9 |
| Frieren/ | 0%(0–1)1 appearance | 0%(0–1)1 appearance | · | · | 54%(7–6)13 appearances | 15 |
| Lyca/ | · | 25%(2–6)8 appearances | · | 0%(0–2)2 appearances | 0%(0–4)4 appearances | 14 |
| Saida/ | · | 20%(1–4)5 appearances | · | 100%(8–0)8 appearances | 100%(1–0)1 appearance | 14 |
Look up a hero’s team groups and results. Each team’s example placement is in the team group profiles.
The three most-used tiles on each current SL map, as a share of Rolan's appearances there. Tiles on the enemy side are mirrored to the ally side.
The three most-used tiles on each current SL map, as a share of Gwyneth's appearances there. Tiles on the enemy side are mirrored to the ally side.
The landing tile is the tile on the enemy half that matches hers after the board is turned halfway around, the same rule 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 formation 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 the blue tile only shows where enemies usually are. Check the actual formation when you know it.
A filled team label means the hero is core in that team group, and an outlined one means it is flex. Pick rate is how often the hero was picked across all recorded games. Both sides can pick the same hero, so the rate can reach 200%. Win-rate gap is how many percentage points the hero's win rate is above or below the average win rate of the players who picked it. Players who picked the hero more often count more toward that average. It cannot show how much the hero matters on its own, because teammates, maps and opponents change from game to game. A dash means fewer than 8 decided games.
| Hero | Teams | Picks | Pick rate | ||
|---|---|---|---|---|---|
| Gwyneth | Gwyn/ |
522 | 43% | 61%(318–204) | +11 |
| Alna | Alna/ |
521 | 43% | 55%(286–234) | +5 |
| Contess | Saida/ |
512 | 43% | 56%(274–212) | +6 |
| Frieren | Orion/ |
509 | 42% | 50%(257–252) | 0 |
| Gunnar | Gwyn/ |
508 | 42% | 62%(313–195) | +11 |
| Dunlingr | Gwyn/ |
470 | 39% | 64%(299–171) | +11 |
| Evie | Orion/ |
449 | 37% | 54%(242–203) | +4 |
| Saida | Saida/ |
414 | 34% | 48%(189–203) | −3 |
| Elijah & Lailah | Saida/ |
406 | 34% | 55%(213–171) | +4 |
| Rolan | Orion/ |
402 | 33% | 66%(264–134) | +16 |
| Kulu | Gwyn/ |
401 | 33% | 65%(260–141) | +10 |
| Athalia | Alna/ |
399 | 33% | 44%(174–225) | −7 |
| Mehira | Alna/ |
393 | 33% | 57%(224–168) | +2 |
| Solise | Saida/ |
390 | 32% | 42%(158–222) | −6 |
| Sylphira | Alna/ |
383 | 32% | 59%(227–155) | +6 |
| Aliceth | Gwyn/ |
372 | 31% | 59%(218–149) | +8 |
| Orion | Orion/ |
346 | 29% | 54%(186–158) | +3 |
| Zandrok | Orion/ |
339 | 28% | 52%(176–162) | +1 |
| Thador | Thador/ |
331 | 28% | 37%(122–206) | −12 |
| Cyran | Gwyn/ |
314 | 26% | 50%(158–156) | +3 |
| Perseus | Orion/ |
303 | 25% | 49%(148–153) | 0 |
| Lily May | Thador/ |
273 | 23% | 36%(98–173) | −14 |
| Hepler | Alna/ |
214 | 18% | 44%(90–115) | −9 |
| Pandora | Zorya/ |
211 | 18% | 35%(73–138) | −13 |
| Ravion | Thador/ |
203 | 17% | 30%(60–143) | −14 |
| Callan | Zorya/ |
192 | 16% | 39%(74–118) | −8 |
| Rowan | Orion/ |
191 | 16% | 49%(91–96) | +1 |
| Zorya | Zorya/ |
191 | 16% | 34%(65–126) | −16 |
| Taichi & Agumon | Rolan/ |
190 | 16% | 56%(105–83) | +2 |
| Gerda | Saida/ |
176 | 15% | 48%(79–87) | −4 |
| Himmel | Thador/ |
149 | 12% | 46%(68–81) | −5 |
| Shadewing | Zorya/ |
146 | 12% | 38%(55–91) | −13 |
| Igor | Alna/ |
126 | 10% | 41%(52–74) | −6 |
| Eryndor | Alna/ |
111 | 9% | 46%(51–60) | −10 |
| Niru | Satrana/ |
72 | 6% | 21%(15–57) | −18 |
| Peggy | · | 63 | 5% | 56%(35–28) | −1 |
| Marcille | Marcille/ |
61 | 5% | 31%(19–42) | −16 |
| Bonnie | Satrana/ |
56 | 5% | 14%(8–48) | −19 |
| Hugin | · | 55 | 5% | 29%(16–39) | −6 |
| Baelran | Saida/ |
53 | 4% | 31%(16–36) | −10 |
| Cryonaia | Dun/ |
47 | 4% | 40%(19–28) | −5 |
| Pang | · | 46 | 4% | 59%(26–18) | +5 |
| Talene | · | 42 | 3% | 35%(14–26) | −8 |
| Satrana | Satrana/ |
42 | 3% | 17%(7–35) | −17 |
| Daimon | Satrana/ |
42 | 3% | 17%(7–35) | −14 |
| Lyca | Marcille/ |
38 | 3% | 18%(7–31) | −27 |
| Koko | Orion/ |
28 | 2% | 39%(11–17) | −13 |
| Eironn | · | 28 | 2% | 18%(5–23) | −22 |
| Kazim | · | 23 | 2% | 30%(7–16) | −23 |
| Kordan | Orion/ |
20 | 2% | 25%(5–15) | −21 |
| Aurora | · | 18 | 1% | 47%(8–9) | −5 |
| Fay | · | 17 | 1% | 59%(10–7) | +7 |
| Damian | Saida/ |
17 | 1% | 18%(3–14) | −24 |
| Zanie | · | 16 | 1% | 25%(4–12) | −10 |
| Scarlita | · | 15 | 1% | 47%(7–8) | −8 |
| Ludovic | · | 15 | 1% | 7%(1–14) | −8 |
| Phraesto | · | 14 | 1% | 21%(3–11) | −28 |
| Shemira | · | 14 | 1% | 7%(1–13) | −6 |
| Tilaya | · | 13 | 1% | 23%(3–10) | −18 |
| Ulmus | · | 11 | 1% | 9%(1–10) | −11 |
| Tasi | · | 11 | 1% | 0%(0–11) | −16 |
| Brutus | · | 10 | 1% | 43%(3–4) | – |
| Reinier | Frieren/ |
10 | 1% | 20%(2–8) | −13 |
| Voracia | · | 10 | 1% | 20%(2–8) | −34 |
| Berial | · | 8 | 1% | 50%(4–4) | −24 |
| Galahad | · | 7 | 1% | 14%(1–6) | – |
| Shakir | · | 7 | 1% | 14%(1–6) | – |
| Lorsan | · | 5 | 0% | 0%(0–5) | – |
| Odie | · | 5 | 0% | 0%(0–5) | – |
| Velara | · | 4 | 0% | 50%(2–2) | – |
| Mikola | · | 4 | 0% | 0%(0–4) | – |
| Yamato & Gabumon | · | 3 | 0% | 67%(2–1) | – |
| Dionel | Lyca/ |
3 | 0% | 33%(1–2) | – |
| Thoran | · | 3 | 0% | 0%(0–3) | – |
| Valka | · | 3 | 0% | 0%(0–3) | – |
| Lucca | · | 2 | 0% | 50%(1–1) | – |
| Cassadee | · | 2 | 0% | 50%(1–1) | – |
| Faramor | · | 2 | 0% | 0%(0–2) | – |
| Lamentis | · | 2 | 0% | 0%(0–2) | – |
| Silvina | · | 2 | 0% | 0%(0–2) | – |
| Antandra | · | 1 | 0% | 0%(0–1) | – |
| Nerion | · | 1 | 0% | 0%(0–1) | – |
| Sonja | · | 1 | 0% | 0%(0–1) | – |
| Harak | · | 1 | 0% | 0%(0–1) | – |
These results come from matches shared by multiple contributors across different leagues. No replays were reviewed, so battle details such as targeting and skill timing have not been checked. Most records come from top-spending players with heavily upgraded rosters, and results may differ with fewer upgrades. This page includes only some of the season's games, and 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.
This page includes only some of the season's games. These games come from many players and districts. A team's record here shows only how it did in these games. It may have done better or worse in games not included here.
Some players and teams appear far more often than others. A team's results can depend on who used it, their opponents and their upgrades. Games where a player had to reuse the same teams show one team choice several times over, so they are related results. Before relying on a team's record, check how many players it comes from. Each team's profile gives that.
Tied maps are left out of win rates. The game gives a tied battle to one player, the defending side, and a board pop-up shows that side as having won on a tie. Winning a map on a tie does not show that the team won the battle, so every win rate leaves ties out.
How missing upgrade levels are read. Paragon levels are recorded in 1069 games and refinement levels in 674. 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 for an upgrade type, a hero with no recorded level for that type is treated as level 0.
The strength explanation covers the upgrade scale and the equal-upgrade ratings. The full ratings in each team's profile show its record when it had the upgrade advantage, when it was behind, and when near level.
Teams are grouped by shared heroes. A hero in 80% or more of a group's appearances counts as core. A hero in 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 every recorded appearance, including ties and the matches where players had to reuse the same teams, so a high share means a hero was picked often, whatever the results were. 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. One team can be counted in more than one row, so the rows can add up to more than the group's appearances. Other recorded lineups lists exact teams outside those rows, including rare picks and losing teams.
The 80% and 15% cutoffs are choices made for this analysis. 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.
Limited evidence means fewer than 30 games excluding ties, fewer than 10 team selections or fewer than 5 players. Concentrated sample means one player contributes at least 35% of appearances, or the two most frequent players contribute at least 60%.
High win rate flags a win rate of at least 60%, excluding ties, with limited evidence. Narrow use marks a group with a win rate below 50% that the counter test still finds to be a counter to another team, named on its label. Worse than it looks and Better than it looks mark a group whose players had clearly more, or clearly fewer, upgrades than their opponents. Once those upgrade differences are accounted for, its expected win rate against an average team moves by at least 5 percentage points, so the unadjusted rating overstates or understates the team. These labels are hints for reading the table. A team's tier is set separately, and no statistical test is behind a label. A winning record may depend on opponents or upgrades. A losing record can still include a useful counter.
The full ratings in each profile show the record of the other players without the most frequent one, and the record near level, where both sides had 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.
Each team group gets a rating worked out 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 in the team groups table. A high rating does not guarantee a favorable matchup. Use the counter results for the opponent you expect.
The rating range in the full ratings comes from redoing the calculation 3,000 times, each time on a random redraw of the recorded matches. Tier 1 means at least 80% of those redraws put the team above a 55% win rate against an average team at equal upgrades (a rating above 1.22). Tier 3 means at least 80% of them put it below 45% (a rating below 0.82). Every other team is Tier 2: close to average, or with too few games to say.
The tiers only compare the recorded teams with each other. Players still field Tier 3 teams. The label means the team rates below the other recorded teams, and it says nothing about teams nobody recorded. The same tiers group the teams in the counter ladder. This rating range is a different figure from the win-rate range shown in the team groups table. The same players appear in many matches, so their results are related, and these ranges do not fully account for that.
The equal-upgrade rating accounts for upgrades. A team can win partly because its heroes have more paragon or EX refinement than the opponent. The equal-upgrade rating includes those differences when estimating team strength.
Both upgrades are counted on one scale, Rivalry points. A paragon level on one hero is 4.5 points, and a fully refined hero is 7.7 points. The equal-upgrade rating uses each game's combined upgrade gap, the number of points one side was ahead by. The size of that effect is estimated from the 866 decided games where the sides were not equal. Between two otherwise evenly matched teams, being one paragon level ahead on one hero moves the expected win rate from 50% to about 55%, and being ahead by one fully refined hero moves it to about 59%.
Players with more upgrades also tend to have better results. So the rating cannot fully tell how much of their results came from skill and how much from upgrades. An equal-upgrade rating below the unadjusted 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 ratings from those redraws. The unadjusted rating is still shown so the two can be compared. 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.
Each diagram shows one of the group's recorded teams as it was actually placed on a Supreme League map. The team is chosen from the group's five-hero teams that include every hero in the group's name, by how widely it was played and how well it did, taken together. Each player who fielded a team counts once, a little more if they picked it again later, but never as much as two players. How well a team did is its chance to beat an average team at equal upgrades, with its opponents' strength taken into account, and a team with few games is judged closer to its group's average. A team only one player fielded can be chosen when its record is strong enough. The same balance then picks the team's exact placement, on whichever map it was used.
The caption counts the games of that exact placement, with its wins and losses. Guild Duel games count when they were played on a map that Supreme League also uses. A group never recorded on a Supreme League map is drawn on the map it was recorded on. The diagram is a recorded example. It does not show the team's best map. Formations recorded on the enemy side are turned around onto the ally half of the board.
The two percentages count this group's boards on the diagram's map, whichever of its teams was on them, one per game. Placed here is how often a unit stood on that tile, out of the boards it was on. On N% of boards is how many of all those boards it was on. 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.
How the test treats each game. The comparison is made map by map and the results are then combined. Losing while far behind on paragon and refinement is weaker evidence of a counter, and winning from behind is stronger evidence. Games with teams from the same group on both sides are excluded. A map where the team faced only one opponent is left out, because there is no other record there to compare against.
Testing many matchups can produce chance findings. 33 pairs of team groups had enough games to test in both directions (66 tests). The counter table labels a result as supported when its p-value is below 0.05 (1 in 20), and as possible, unconfirmed between 0.05 and 0.15 (about 1 in 7). The p-value shows how unusual this record is compared with the team's usual results on the same maps, after accounting for upgrade differences. 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. That stricter threshold is met by 6 of the 12 supported counters. The counter table lists its rows from the strongest evidence down, so those come first.
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. A record built from the same few players says less than its number of games suggests.
The opponents still matter. A team's record against everyone else includes one-off lineups, which can make it look stronger. Comparing results on the same maps and accounting for upgrade gaps does not remove differences in player skill or in the exact teams used. The estimated upgrade effect can also include differences between players. These records cannot tell the two apart.
The upgrade effect is estimated. The counter test uses that estimate as if it were exactly right. Its p-values do not account for that estimate being off, and repeated appearances by the same players and teams also make them less exact. The Near level column helps show whether the losing record also appears when upgrades are similar. Player skill and team differences can still affect it. Treat a supported counter as a matchup worth investigating. The p-value does not say how likely you are to win it, and it is not the probability that a counter is false.
The team grouping rules can also affect the result, because broader or narrower groups change which games count for a team.