Gem Fort Stats & Draft Guide
Gem Fort is a Gem Grab map with 1,118 Ranked battles of top players in the last 21 days. The best measured pick right now is Mr. P at 66% over 117 matches. Below: who wins here, who is safe to pick early, who to keep for later and who to avoid, each with the matches behind the number.
- Ranked battles
- 1,118 · last 21 days
- In Ranked rotation
- Yes · last battle 2026-11-15
- Best measured pick
- Mr. P · 66% over 117 matches
- Most picked
- Rico · in 50% of battles
What the data says about Gem Fort
Over the last 21 days we counted 1,118 Ranked battles of top players on Gem Fort. 43 brawlers have at least 30 matches here, so they get a win rate; 56 more were picked too rarely to judge. The map is in the current Ranked rotation (last battle 2026-11-15).
Counted from Ranked matches of top players over the last 21 days, last updated 2026-11-15. Matches played before the latest balance change count for less. A brawler needs at least 30 matches on this map before we show a win rate.
Best Brawlers on Gem Fort by win rate
The strongest measured pick on Gem Fort is Mr. P: 66% over 117 matches (fairly reliable), picked in 11% of battles, 7 points above its Gem Grab average of 59%. Mr. P (66%), Lou (59%), Surge (59%), Bo (54%) combine a winning record here with a high blind-pick safety, so they are the safest early picks. Ash (58%) wins here too but is easy to counter: keep it for a late pick. Stu (46% over 193 matches) is among the most picked brawlers here yet loses more than it wins.
| Brawler | Win rate on Gem Fort | Matches | How sure | Picked in | vs Gem Grab | Draft verdict |
|---|---|---|---|---|---|---|
| Mr. P controller | 66% | 117 | fairly reliable | 11% of battles | +7 pts | safe early pick |
| Lou controller | 59% | 167 | fairly reliable | 16% of battles | +4 pts | safe early pick |
| Surge fighter | 59% | 129 | fairly reliable | 17% of battles | +1 pts | safe early pick |
| Ash tank | 58% | 118 | fairly reliable | 15% of battles | +4 pts | strong, pick late |
| Lumi fighter | 59% | 64 | early signal | 6% of battles | +9 pts | promising |
| Amber fighter | 55% | 262 | reliable | 31% of battles | +3 pts | strong pick |
| Tara support | 60% | 55 | early signal | 5% of battles | +4 pts | promising |
| Alli fighter | 56% | 108 | fairly reliable | 10% of battles | +9 pts | strong pick |
| Lily assassin | 60% | 40 | early signal | 4% of battles | +5 pts | promising |
| Max support | 58% | 46 | early signal | 14% of battles | +11 pts | promising |
| Shade assassin | 57% | 65 | early signal | 11% of battles | +6 pts | promising |
| Bo controller | 54% | 166 | fairly reliable | 15% of battles | +1 pts | safe early pick |
Showing the 12 strongest of 43 measured brawlers. "Picked in" is the share of Gem Fort battles with the brawler on either team. "vs Gem Grab" is the win rate here minus the same brawler's win rate across all Gem Grab maps.
Safe picks and risky picks on Gem Fort
Safe means winning here with a good sample and hard to punish when picked blind; risky means the numbers say it loses, or it wins but is easy to counter once the enemy sees it. The pick-order guide explains when to lock what.
Safe early picks
Winning record, a fairly reliable sample or better, and a blind-pick safety of at least 6/10 in the coaching model.
- Mr. P — 66% over 117 matches (fairly reliable), picked in 11% of battles, 7 points above its Gem Grab average of 59%. Blind-pick safety 6/10.
- Lou — 59% over 167 matches (fairly reliable), picked in 16% of battles, 4 points above its Gem Grab average of 55%. Blind-pick safety 7/10.
- Surge — 59% over 129 matches (fairly reliable), picked in 17% of battles, in line with its Gem Grab average. Blind-pick safety 6/10.
- Bo — 54% over 166 matches (fairly reliable), picked in 15% of battles, in line with its Gem Grab average. Blind-pick safety 7/10.
- Emz — 53% over 236 matches (reliable), picked in 21% of battles, in line with its Gem Grab average. Blind-pick safety 6/10.
- Squeak — 52% over 113 matches (fairly reliable), picked in 10% of battles, 5 points above its Gem Grab average of 47%. Blind-pick safety 6/10.
Strong, but keep for a late pick
Win here but are easy to counter: lock them in once the enemy draft is visible, after checking which counters are still open.
- Ash — 58% over 118 matches (fairly reliable), picked in 15% of battles, 4 points above its Gem Grab average of 54%. Blind-pick safety 4/10.
Also winning here
Winning record without a strong pick-order signal, or with a sample that is still small.
- Lumi — 59% over 64 matches (early signal), picked in 6% of battles, 9 points above its Gem Grab average of 50%. Blind-pick safety 5/10.
- Amber — 55% over 262 matches (reliable), picked in 31% of battles, 3 points above its Gem Grab average of 52%. Blind-pick safety 5/10.
- Tara — 60% over 55 matches (early signal), picked in 5% of battles, 4 points above its Gem Grab average of 56%. Blind-pick safety 7/10.
- Alli — 56% over 108 matches (fairly reliable), picked in 10% of battles, 9 points above its Gem Grab average of 48%. Blind-pick safety 5/10.
- Lily — 60% over 40 matches (early signal), picked in 4% of battles, 5 points above its Gem Grab average of 55%. Blind-pick safety 5/10.
- Max — 58% over 46 matches (early signal), picked in 14% of battles, 11 points above its Gem Grab average of 47%. Blind-pick safety 7/10.
7 more measured brawlers carry this verdict on this map; only the strongest 6 are listed.
Popular but losing
Among the 8 most picked brawlers on Gem Fort, yet losing more than winning.
- Stu — 46% over 193 matches (fairly reliable), picked in 21% of battles, 5 points below its Gem Grab average of 51%. Blind-pick safety 6/10.
Map effect: who changes on Gem Fort
A gap of 4 points or more to the brawler's own Gem Grab average, with at least 30 matches on both sides. Compare with the live tier list: a modest overall tier can hide a Gem Fort specialist.
Better here than in Gem Grab overall
The Gem Fort specialists.
- Colt — 52% over 33 matches (early signal), picked in 3% of battles, 18 points above its Gem Grab average of 34%. Blind-pick safety 5/10.
- Max — 58% over 46 matches (early signal), picked in 14% of battles, 11 points above its Gem Grab average of 47%. Blind-pick safety 7/10.
- Rosa — 57% over 35 matches (early signal), picked in 3% of battles, 9 points above its Gem Grab average of 48%. Blind-pick safety 6/10.
- Lumi — 59% over 64 matches (early signal), picked in 6% of battles, 9 points above its Gem Grab average of 50%. Blind-pick safety 5/10.
- Alli — 56% over 108 matches (fairly reliable), picked in 10% of battles, 9 points above its Gem Grab average of 48%. Blind-pick safety 5/10.
- Mr. P — 66% over 117 matches (fairly reliable), picked in 11% of battles, 7 points above its Gem Grab average of 59%. Blind-pick safety 6/10.
Worse here than in Gem Grab overall
This layout works against them.
- Starr Nova — 43% over 54 matches (early signal), picked in 6% of battles, 10 points below its Gem Grab average of 53%. Blind-pick safety 5/10.
- Mortis — 42% over 224 matches (reliable), picked in 20% of battles, 7 points below its Gem Grab average of 49%. Blind-pick safety 4/10.
- Moe — 43% over 53 matches (early signal), picked in 5% of battles, 6 points below its Gem Grab average of 50%. Blind-pick safety 5/10.
- Wendy — 44% over 33 matches (early signal), picked in 18% of battles, 6 points below its Gem Grab average of 50%. Blind-pick safety 5/10.
- Edgar — 46% over 148 matches (fairly reliable), picked in 13% of battles, 6 points below its Gem Grab average of 51%. Blind-pick safety 4/10.
- Gale — 45% over 187 matches (fairly reliable), picked in 17% of battles, 5 points below its Gem Grab average of 51%. Blind-pick safety 5/10.
Brawlers to avoid on Gem Fort
Clearly losing records on Gem Fort; the match count says how much weight to give each one.
- Rico — 43% over 435 matches (reliable), picked in 50% of battles.
- Mortis — 42% over 224 matches (reliable), picked in 20% of battles.
- Spike — 44% over 175 matches (fairly reliable), picked in 16% of battles.
- Colette — 44% over 132 matches (fairly reliable), picked in 18% of battles.
- Gale — 45% over 187 matches (fairly reliable), picked in 17% of battles.
- Starr Nova — 43% over 54 matches (early signal), picked in 6% of battles.
Which roles win on Gem Fort
The best performing role on Gem Fort is controller: 56% over 792 matches, 2 points above the Gem Grab average for that role. The weakest is marksman at 45% over 277 matches.
| Role | Win rate on Gem Fort | Matches | How sure | vs Gem Grab |
|---|---|---|---|---|
| controller | 56% | 792 | reliable | +2 pts |
| tank | 54% | 357 | reliable | +2 pts |
| support | 50% | 331 | reliable | +2 pts |
| fighter | 50% | 1,918 | reliable | 0 pts |
| thrower | 48% | 143 | fairly reliable | +3 pts |
| assassin | 47% | 970 | reliable | -2 pts |
| sharpshooter | 46% | 704 | reliable | -1 pts |
| marksman | 45% | 277 | reliable | -3 pts |
Roles are the coaching model's class per brawler.
Most picked on Gem Fort
How often top players take a brawler onto Gem Fort, and how it went. Popular is not the same as good.
- Rico — picked in 50% of battles, 43% over 435 matches (reliable) (avoid).
- Amber — picked in 31% of battles, 55% over 262 matches (reliable) (strong pick).
- Gus — picked in 29% of battles, 50% over 139 matches (fairly reliable) (coin flip).
- Chester — picked in 27% of battles, 50% over 284 matches (reliable) (coin flip).
- Emz — picked in 21% of battles, 53% over 236 matches (reliable) (safe early pick).
- Stu — picked in 21% of battles, 46% over 193 matches (fairly reliable) (popular but losing).
- Mortis — picked in 20% of battles, 42% over 224 matches (reliable) (avoid).
- Colette — picked in 18% of battles, 44% over 132 matches (fairly reliable) (avoid).
How to read these numbers
A win rate only means something when enough matches sit behind it. Next to every number we say how sure we are:
- reliable — 200 matches or more: a solid basis for a verdict.
- fairly reliable — 80 to 199 matches: enough to spot a real pattern.
- early signal — 30 to 79 matches: a hint, not a verdict.
- too few matches — fewer than 30 matches, so we do not show a win rate yet.
A 60% win rate over 40 matches can easily be luck; a 54% win rate over 800 matches is a real edge. Brawlers with fewer than 30 matches are listed without a win rate on purpose.
The coaches' view on Gem Fort
The coaching model's view of the closed layout of Gem Fort, next to the measured numbers above. When the two disagree, trust the one with more matches behind it.
Coaches' early picks: Barley combine a positive map fit with high blind-pick safety.
Coaches' later picks: Squeak, Emz fit the map but are easier to counter, so keep them for the middle or last pick.
Mode context
Gem Fort is a Gem Grab map. Read the mode guide for the win condition and role requirements, then apply the map-specific picks above.
Other Gem Grab maps
Same mode, different layout. Compare before you build your Gem Grab pool, or start from the Gem Grab mode guide.
Hard Rock MineGem Grab · open layout · picks and win rates
UndermineGem Grab · mixed layout · picks and win rates
Double SwooshGem Grab · mixed layout · picks and win rates
Rustic ArcadeGem Grab · mixed layout · picks and win rates
Crystal ArcadeGem Grab · limited data
On a RollGem Grab · limited data
Whisper ValeGem Grab · limited data
Let Me DanceGem Grab · limited dataRavenous RapidsGem Grab · limited dataDon't Leaf MeGem Grab · limited data
Bottom TwoGem Grab · limited dataEmotional DetachmentGem Grab · limited data
Dungeon TrainGem Grab · limited data
Deathcap TrapGem Grab · limited data
Local RestaurantsGem Grab · limited dataStardust StormGem Grab · limited data
Open SpaceGem Grab · limited data
PicturesqueGem Grab · limited data
Lilygear LakeGem Grab · limited dataSapphire PlainsGem Grab · limited dataThe cooler Hard RockGem Grab · limited dataRoyal FlushGem Grab · limited dataDeep DinerGem Grab · limited dataEscape VelocityGem Grab · limited data
Ancestral RootsGem Grab · limited data
Snake PitGem Grab · limited dataFour SquaredGem Grab · limited dataData confidence
The measured numbers on this page come from 1,118 Ranked battles of top players on Gem Fort in the last 21 days. 12 brawlers have enough matches here to show a win rate; 56 more were picked too rarely. Matches played before the latest balance change count for less, and no win rate is shown below 30 matches.