GW2 WvW war score beta 3: what changed and what players say

The third WvW war score beta is live from 11 August to 4 September 2026, and it’s the first one ArenaNet dropped into the middle of a matchup on purpose. On 4 August the Competitive team published what beta 2 actually measured. The numbers are worth reading — and so is the gap between them and what the people who played it are saying.

What beta 2 actually rewarded

The whole point of the war score betas is to move scoring away from things that happen while nobody is playing — owning a tower overnight — and toward things a player does: sieging, capturing, killing. Beta 1 went to the extreme and removed almost all passive income. ArenaNet agreed with the pushback that defending had stopped paying, and beta 2 put some of it back: a little for holding tier 3 objectives, considerably more for tiering an objective up.

Progress ticks on their own accounted for only 5%; the 15% figure is ticks, upgrades and holding combined. The single biggest source of war score in beta 2 was keep captures, with player kills close behind. Note that the three shares above don’t add up to 100 — ArenaNet published the headline categories, not a full ledger, so roughly a third of the score generated sits outside the numbers they shared.

How that compares to normal WvW

The comparison ArenaNet offered is the useful part. Outside the betas, war score looks nothing like the split above.

What changes in beta 3

Four changes carry this one, and three of them are simplifications rather than new systems.

The scaling removal is the one to watch. It was meant to pull players together into bigger fights, but ArenaNet found it too complex to balance and too hard to read — and most score was already coming from objectives whose values climb on their own. The capability stays in the code if they want it back.

The exact numbers from the 11 August patch

The blog described the direction; the update notes give the values. All of these apply only while a war score beta is enabled:

• Defensive player kills: 3 → 5 war score.
• Capturing a sentry: 2 war score. Camp supervisors also award 2 on death.
• Guards: nothing.
• Defensive actions must happen inside the defend or attack event area of an objective — not anywhere in the sector.
• A dolyak-kill bug was fixed that had stopped war score being awarded at all.
• Match history exists now — a Match Statistics tab that shows war score stats during a beta.
• Placement-based skirmish rewards return, and can be granted while offline.

Dolyaks keep their points for now. ArenaNet called supply the lifeblood of the mode and said they’d rather keep tuning the value than zero it out, even though dolyak deliveries were one of the outliers they had to fix after beta 1.

What players are saying

The forum thread that followed the blog splits cleanly in two. Almost nobody argues with the in-game score breakdown — it is the most-welcomed item in the post, and the reason is the same every time: for two betas, players were optimising against a system they couldn’t see.

The scoring changes themselves get a much rougher ride. The sharpest verdict came from a player who preferred beta 1, arguing that beta 2 removed the reason to do anything except farm kills and sit on tier 3 objectives, and that it handed the already-stronger teams a widening lead.

Others were more measured but landed in a similar place: a system that pays well for kills quietly punishes the outnumbered side, because points-per-kill and points-per-downed both scale with the group that was already winning the fight.

Where the data and the players disagree

ArenaNet reports keep captures as the single largest source of war score and presents the beta 2 spread as healthy. Several testers read the same fact as the problem: keeps became the cheapest big prize on the map, flipped back and forth without much resistance, which inflates the capture share without anyone actually contesting anything.

One player described losing a skirmish in Eternal Battlegrounds precisely because his team defended a keep instead of abandoning it — the enemy’s kill income from repeatedly killing the defenders outran everything the defenders gained by holding. His conclusion is the uncomfortable one: they would have won the skirmish by giving the objective up.

And there is a methodological caveat that deserves more weight than it got. In a beta, almost nobody is playing to win the matchup. One forum regular put it bluntly — no one has been trying to win, and people still don’t look at the score. A distribution measured under those conditions describes how testers behaved, not necessarily how a competitive season would play out. Beta 3 starting mid-matchup is at least a partial answer to that.

Who should care about beta 3

Nothing here changes how your class plays. If you came looking for that, our breakdown of the July 14 balance update is the relevant one, and the builds library is where the WvW-tagged builds live.

GW2 WvW war score beta 3, answered

Last updated 17 August 2026