Loading the leaderboard…
| # | Player | Town | Elo | Last 10 fights |
|---|
| Rank | Player | Town | Elo | Last 10 | Fights |
|---|
Every conquest fight we have a log for.
| Date | Fight | Attackers | Defenders | Players |
|---|
The best single fights on record, in every fight EldritchBot counts since 2016.
Elo over time
Match history
| Fight | Date | Side | Result | K–D | Damage | Elo | Δ Elo |
|---|
A look at what's coming
Every fight already gets broken down like this on the local build: who actually showed up, who showed up late, and how the war was really won or lost, not just the final result.
Real output from an actual recorded fight.
The elo-over-time chart tracks each side's actual fighting strength minute by minute — sides warping in explain the spikes, deaths explain the overall strength of both sides lowering. Below that, a full presence timeline for every player in the fight — exactly when they showed up, when they died, missed warps, and if they left mid-war.
A battle map is still under development, showing where players are actually present on the map as the fight moves.
Point it at both rosters and it gives you a live win probability before the fight's decided — the same Elo model behind the leaderboard, just running forward instead of backward.
Example output, not a real fight.
It compares both sides' total Elo, scaled for outnumbering, and turns the gap into a percentage. It's most accurate within the first 5 minutes of a fight, looking at whoever's registered to it at that point — and the win % keeps updating dynamically as the fight goes on.
It even picked up on defender advantage on its own — worth about a 10% strength edge in the model.