How we're planning to measure who actually wins objective games — and why flag captures on their own have been lying to us this whole time.
← Back to dashboardRight now, your objective performance is whatever the game's scoreboard says. For KOTH and Oddball that's fine — your score is seconds you held the hill or the ball, which is a genuinely good measure of what you did.
For CTF, your score is flag captures. Nothing else. And a capture is only the last step of a flag run.
Three people did the work. One of them gets the entire score. If you're the guy who pulls flags all night and dies at mid, the stats say you contributed nothing.
We checked, rather than guessed. Here is every stat-tracked CTF game we have — 120 games across the whole squad:
Only 39% of flag pulls ever become a capture. Roughly six in ten flag runs end with someone dying on the way — and the scoreboard records none of that work.
There were also 949 flag returns — nearly 50% more than there were captures. Every one of those denied the other team a point, and every one of them scored exactly zero.
Instead of scoring the last step, we score the whole chain. Every objective medal the game already tracks gets sorted into one of four jobs:
| Axis | What it means | CTF | KOTH | Oddball |
|---|---|---|---|---|
| Pulls | Getting the objective moving | Flag Taken | seconds held | seconds held |
| Finish | Converting it into a point | Flag Captured | n/a — holding is the score | n/a — holding is the score |
| Recover | Denying theirs / resetting ours | Flag Returned | — | — |
| Enforce | Killing the people who matter | Flag Carrier Kill | Hill Defense + Offense | Ball Carrier Kill |
So CTF scores on four axes, KOTH and Oddball on two. Hold time already captures both the "moving it" and "scoring it" jobs in those modes, so splitting them would be inventing a distinction that isn't there.
Every axis is scored against the rest of the squad, not against some number we made up.
Every line on your profile shows two things: what you actually did per game, and where that ranks you in the squad.
So Returns 1.24 3rd of 13 means you return 1.24 flags a game, and only two people in the squad do it more.
Your score for a gametype is the average of your axes — four numbers for CTF, two for KOTH and Oddball.
CTF is an even split — all four axes count 25% each. We asked around about whether pulling should count more than capping, and got genuinely opposite answers from everyone we asked. So we tested it: even at the most lopsided weighting we tried, nobody moved more than two places on the table and the top three never changed. If the choice barely changes the answer and nobody agrees on it, an even split is the one nobody can fairly object to.
KOTH and Oddball are 75% hold time, 25% kills. Hold time is weighted heavily because seconds on the hill are literally how you win — kills only matter because they help you get those seconds. When we tried it at an even split, 11 of 13 players changed position compared to ranking on hold time alone, and one player who held the hill better than 60% of the squad dropped to 11th of 13 for not fragging enough around it. That's clearly wrong, so hold time leads and kills adjust.
In KOTH, the thing that wins the game (seconds) is also a fair measure of what you personally did — your seconds are your seconds. In CTF, the thing that wins the game (caps) gets credited to whoever happened to touch the flag last. That's the whole problem this page is about. So we lean on the win condition in KOTH, and deliberately don't in CTF.
Doing it this way means we never have to argue about whether a flag cap is "worth" 50 seconds of hill time. We don't compare them at all — we compare you against the squad at flag caps, and you against the squad at hill time, then average the results. The exchange rate problem disappears.
The score itself is on a 0–100 scale rather than a rank, because ranks can't be averaged sensibly — 1st is the best result but the smallest number, and everyone's rank would shift every time a new player joined. So the score is 0–100, and the lines that make it up are shown as plain squad ranks.
It's also measured per game, not as a season total, so turning up to more nights doesn't inflate you past someone who played less but did more.
Because it's a straight average, specialists get under-rated. If you only ever defend, you'll be near-100 on Recover and near-0 on the other three, which averages out to something mediocre. That's why every score comes with a role attached — the number tells you how much, the role tells you what.
Whichever axis you're clearly strongest at becomes your role:
| Role | Means |
|---|---|
| Runner | You pull flags and get them moving. |
| Closer | You're the one who converts — you finish what others start. |
| Anchor | You defend and return. Their offence dies on you. |
| Enforcer | You kill carriers and contest the hill. Objective-aware fragging. |
| Grinder | You sit on the hill or the ball and hold it. KOTH and Oddball only. |
| Allrounder | No single axis stands out — you do a bit of everything. |
Running the real numbers from that night through the system:
Xorvek is the case that makes the point. Under the old scoring he's a rounding error. Under this, he's the squad's best flag defender, and it's obvious at a glance.
Being upfront about this, because it's the most likely thing to annoy someone.
There is no such thing as carry distance in the data. If you pull the flag and run it 80% of the way before dying, you score exactly the same as someone who pulls it and dies immediately. Both count as one pull.
The game simply doesn't record how far or how long you carried it. Nothing we can do about that — but you should know the number can't see it.
Fair question, and we checked it specifically: throwing the flag forward and picking it back up to move faster does not inflate your pulls. Across 23 player-games with flag activity, the average was 2.1 pulls and the most common value was exactly 1. If re-grabbing a thrown flag counted, a single juggled run would put everyone in the dozens. The medal only fires on a real take — flag at rest, or after a reset.
So juggling doesn't cheat the system. But a long run doesn't earn extra credit either. Pulls measure how often you start something, not how far you got.
This is a first pass and all of it is up for debate. The parts most likely to be wrong are the judgement calls, not the maths:
Bring it up in Discord. Nothing here is locked in, and the numbers are cheap to change — the hard part was proving the old score was broken, and that bit's done.