Behind on support.
Gold recovery is not the whole role.
Support operates on the lowest private income and the highest dependence on opportunity. Personal recovery is valuable early, but team conversions dominate the longer horizon.
Compare every role →Support can recover the game without recovering the scoreboard.
Early personal recovery was directionally powerful, but support gold bands are not directly comparable with farming roles. Over eight minutes, fights, team gold, teammate economy, bottom-lane recovery, turrets, objectives, and cross-map conversion all produced supported gains.
A support deficit should be interpreted through access to wards, durability, engage or peel thresholds, and the ability to move with teammates. The next study must measure vision, roam, save, engage, and reset opportunities directly.
at 10:00
113 similar team situations.
from here
Likely range: 29.1%–38.5%
later in the game
What happened in comparable games.
How the forecast changes from here
Remain in the selected absolute deficit band.
Move one band farther behind every two minutes.
Move one band toward even every two minutes.
Every percentage forecasts the entire match.
What changed the next eight minutes?
See which kinds of recovery showed up most often in winning comebacks.
| What changed | Chance to win | Observed gain | Adjusted gain | Likely range | Similar situations |
|---|---|---|---|---|---|
| Fight swingkill differential improves by at least two | 74.0% | +24.0 pp | +32.9 pp | 30.7–35.2 pp | 27664% of games |
| Team gold recoveryteam gold differential improves by more than 500g | 69.4% | +21.9 pp | +0.1 pp | 0.0–0.2 pp | 10424% of games |
| Teammate economynon-support teammate gold differential improves by more than 500g | 68.7% | +20.9 pp | +13.8 pp | 11.9–16.0 pp | 13431% of games |
| Bottom-lane gainbot-duo gold differential improves by more than 500g | 58.8% | +15.1 pp | +6.4 pp | 5.7–7.1 pp | 32274% of games |
| Turret gainrelative structure count improves by at least one | 69.3% | +14.7 pp | +7.4 pp | 6.3–8.3 pp | 24958% of games |
| Personal recoverysupport gold differential improves by more than 250g | 54.7% | +12.6 pp | -0.0 pp | -0.0–0.0 pp | 17540% of games |
| Objective gainrelative neutral-objective count improves by at least one | 63.5% | +11.8 pp | +3.1 pp | 2.5–3.7 pp | 29067% of games |
| Cross-map dividendsupport does not materially recover while teammates, objectives, structures, or fights improve | 57.0% | +9.5 pp | +8.7 pp | 7.0–10.4 pp | 8419% of games |
| Mid-lane gainmid gold differential improves by more than 300g | 51.0% | +9.2 pp | +0.5 pp | 0.3–0.7 pp | 431100% of games |
What does adjusted gain mean?
Adjusted gain tries to separate this change from other things happening at the same time. It is still a pattern in the data, not proof that the action alone caused the win.
See every recovery path
Hold, bleed, recover, reverse
Results with fewer than 50 similar situations are early signals only.
| After | Path | Ending gap | Chance to win | Change | Situations | Confidence |
|---|---|---|---|---|---|---|
| +2 min | bleeding | -1,067g | 31.0% | -2.7 pp | 81 | exploratory |
| +2 min | recovered | -321g | 51.0% | +13.2 pp | 47 | suppressed |
| +2 min | reversed | +174g | 16.2% | +0.0 pp | 7 | suppressed |
| +2 min | stabilized | -704g | 32.4% | -2.6 pp | 224 | exploratory |
| +4 min | bleeding | -1,188g | 23.9% | -8.2 pp | 152 | exploratory |
| +4 min | recovered | -350g | 49.8% | +13.6 pp | 43 | suppressed |
| +4 min | reversed | +266g | 57.6% | +21.4 pp | 14 | suppressed |
| +4 min | stabilized | -680g | 35.9% | -0.9 pp | 150 | exploratory |
| +6 min | bleeding | -1,318g | 29.3% | -7.2 pp | 153 | exploratory |
| +6 min | recovered | -315g | 51.5% | +5.5 pp | 43 | suppressed |
| +6 min | reversed | +307g | 57.4% | +20.1 pp | 20 | suppressed |
| +6 min | stabilized | -684g | 38.4% | +2.5 pp | 108 | exploratory |
| +8 min | bleeding | -1,405g | 27.8% | -8.8 pp | 168 | exploratory |
| +8 min | recovered | -254g | 50.6% | +7.9 pp | 40 | suppressed |
| +8 min | reversed | +394g | 61.1% | +20.1 pp | 25 | suppressed |
| +8 min | stabilized | -709g | 42.4% | +4.3 pp | 82 | exploratory |
See the full Support win-chance table
Chance to win by minute and deficit
Average result for teams in each 250g deficit range.
| Minute | -1,500g | -1,250g | -1,000g | -750g | -500g | -250g |
|---|---|---|---|---|---|---|
| 3:00 | — | — | — | — | 34.8% | 48.5% |
| 4:00 | — | — | — | — | 40.9% | 46.2% |
| 6:00 | — | — | — | 31.6% | 40.6% | 45.7% |
| 8:00 | — | — | 37.7% | 33.7% | 38.2% | 44.6% |
| 10:00 | — | — | 33.8% | 34.4% | 39.4% | 44.8% |
| 12:00 | — | 40.4% | 27.9% | 32.1% | 40.8% | 46.6% |
| 14:00 | 22.9% | 31.5% | 31.5% | 32.6% | 41.0% | 45.0% |
| 16:00 | 28.4% | 30.9% | 31.1% | 34.5% | 40.5% | 50.4% |
| 18:00 | 31.7% | 29.0% | 33.3% | 38.0% | 41.6% | 45.4% |
| 20:00 | 33.5% | 32.3% | 32.2% | 36.2% | 43.8% | 46.9% |
How we calculated this
How to read the deeper tables
The first comparison shows everything that changed along the observed path. The adjusted comparison tries to avoid giving one fight or tower credit for the entire comeback. We only show cells with at least 50 similar situations.
support-win-path-v1 · patch 16.17 · EUW1 · KR · NA1