Q2 LEAGUES

Server Activity Heatmap

Maps played by hour of day and day of week, across all recorded history. Darker = fewer games, brighter = more activity.

Player Elo Growth Over Time

Number of human players who have reached each Elo threshold at any point in their history. Tracks peak Elo — once a player crosses a threshold they stay counted. Growth is expected to continue as new players join the pool.

Elo Equilibrium

Every Elo point lost by players below 1000 is gained by players above 1000. Both sides balance — the total deficit equals the total surplus. The percentage in each bar represents that range's share of the total Elo distance from 1000 across all players.

New Player Arrivals

Weekly count of players who played their very first map on the server. Brighter bars indicate higher activity weeks.

Current Map Rotation

Number of bots: 6

Techs enabled: Yes

Power Ups: No

Mods

We're currently weighing two community mods for the server: Q2Eaks, a lightweight set of gameplay tweaks, and MuffMode, a more comprehensive competitive ruleset overhaul. Both bring useful quality-of-life features, and we want whichever we adopt to fit the spirit of this server before rolling it out. No decision has been made yet — stay tuned for updates.

Player Retirement System - new

Starting August 29, 2026, multiple players on the same account have been consolidated. Only a single player per account is now allowed in the Rankings — if an account had multiple names ("alts"), one has been kept as that account's official player, and the rest have been retired: hidden from Rankings and from the Server tab's growth charts, though their full historical stats and Elo Evolution chart remain visible on their own player page. Any game played on a retired account from now on, even under a brand-new name never seen before, is attributed entirely to that account's official player.

Each retired player's Elo was reset to 1000, and the difference between their old Elo and 1000 was redistributed evenly across the bots — the bots have effectively inherited all of the Elo that was accumulated by the retired players, so the total Elo in the pool is unchanged.

Update — September 1, 2026: a few merging edge cases caused minor Elo changes after the corrections above were applied. All of these have been resolved, and every Elo change has been reviewed and confirmed correct.

Recent Database Corrections - new

September 3, 2026 — A further "collision" issue was found affecting a handful of players: in rare cases, two different people's stats could get merged onto the same account during a live session. Fixing this properly meant recalculating the database from scratch rather than patching around it, so some players' Elo may have shifted slightly as a result.

Some bugs are worth fixing fully at the root rather than just patched over — this was one of them. With this correction applied, the vast majority of known issues have now been ironed out, and the Elo values you see today reflect the actual values players have earned.

September 5, 2026 — Another edge-case issue arose that caused a similar effect to the previous one, but from a different underlying cause. A couple players' Elo were corrected and some others also slightly shifted.

No foul play was detected — the Quake 2 server has some oddities in how it reports match data, which can make this kind of data tricky to collect accurately. As before, every frag is accounted for on each player's record, and the numbers you see reflect the up-to-date corrections.

Thanks for your patience and understanding while these corrections were made. Keep on fragging — cheers!

Elo System

The Elo rating system is a method originally designed for ranking chess players based on head-to-head results. Each player starts at 1000 points. When you frag someone, you gain points and they lose points. The amount depends on the rating difference — beating a higher-rated player earns more points than beating a lower-rated one, and vice versa.

Applying Elo to a multiplayer free-for-all deathmatch is more challenging than a simple 1v1 scenario. In a typical map, a strong player faces multiple opponents simultaneously, which can amplify rating swings in ways that don't always reflect true skill. Balancing this is an ongoing process.

A spread cap limits how far apart two players' Elos can appear to each other for the expected-score calculation. For example, if a 1500-rated player frags a 1000-rated player, the actual ratings are used because the outcome was expected. But when the 1000 player upsets the 1500, each player sees the opponent capped within the spread cap distance — limiting the swing rather than producing an outsized gain or loss.

Elo V2 — Changes effective August 19, 2026

To improve rating stability and better reflect performance in busy lobbies, the following changes take effect on August 19, 2026. All maps before that date continue to use the original V1 rules.

K-Factor (Elo swing per frag)

Matchup	V1	V2	Effect
Player vs Player	K=16	K=8	Slower convergence, more stable rankings
Player vs Bot	K=2	K=4	Bot frags matter slightly more
Bot vs Player	K=2	K=4	Dying to a bot costs more
Bot vs Bot	K=16	K=4	Bot-on-bot fights barely move ratings

Halving K from 16 to 8 for player-vs-player frags has a few interrelated effects:

Slower convergence. A new player at 1000 Elo who is actually playing at a 1400 level climbs there relatively quickly under K=16 — each frag against a similarly-rated opponent swings roughly 8 points (since the expected score is ~0.5, the gain is approximately K×0.5). At K=8 that same frag only moves ~4 points. It takes roughly twice as many frags — meaning twice as many maps — for players to reach their true rating.

More stable rankings. The flip side is that established players will not see their Elo bounce around as much session to session. A bad night will not crater someone's rating the way it can at K=16. The leaderboard will feel more earned and less volatile, which rewards consistency. Top players will hold their positions more firmly.

Upsets matter less. With K=16, a lower-rated player fragging a higher-rated one gets a meaningful reward (and the higher-rated player takes a meaningful hit). At K=8, those upset moments are half as impactful. This makes it harder for improving players to break into the upper ranks, and it protects top players from losing ground to up-and-comers.

Spread Cap (bots and players)

Real players in the map	V1	V2
1 – 4 players	200	200
5+ players	200	100

In busier lobbies (5 or more real players), the tighter cap dampens the outsized swings that can happen when many skill levels mix in a free-for-all. With fewer players, the original cap is preserved so ratings still converge at a reasonable pace.

Contact the Admin

Need help or want to report an issue while in-game? Type ADMIN: your message here in the chat. This triggers an instant notification to the server administrator, who will attempt to get online as soon as possible.

If you do not wish to participate in this server's ranking system, mention this to the admin by typing something like ADMIN: please remove me from the web site in the chat. Your stats will be removed from the site. Please note that you will be expected to no longer log into the Q2 server.

Fair Play

Cheating, exploiting, and spamming are not tolerated. The server actively monitors for spam and other disruptive behavior. Hopefully this won't be too much of an issue — Q2 Leagues is built around fun and good sportsmanship. Any player found cheating or deliberately disrupting gameplay will be removed from the server and ousted from this site's rankings.

Forum

This feature is currently being worked on and should become available shortly.