For eight months this was a game people described as dishonest. Not badly balanced, not poorly mapped — dishonest. You would round a corner, take cover, and die a full third of a second after the wall was between you and the shooter. The forums filled with the usual accusations and the usual defences, and almost nobody was talking about the number that explains all of it.
Then the studio raised the tick rate from twenty to sixty, changed nothing else, and the complaints stopped.
What a tick is, and why twenty is not enough
A server resolving the world twenty times a second is deciding where everyone is once every fifty milliseconds. Between those decisions the game does not know anything; it is interpolating, guessing, and drawing a plausible picture on your screen.
Add the round-trip time of the connection and you get the total window in which two players can hold contradictory beliefs about where you are standing. On a good line at twenty ticks we measured a hundred and ten milliseconds. That is not lag in the colloquial sense — our ping was thirty-one — it is the update rate refusing to arbitrate quickly enough.
| Configuration | Tick interval | Our median RTT | Peek advantage |
|---|---|---|---|
| Launch, near region | 50 ms | 31 ms | 110 ms |
| Launch, far region | 50 ms | 96 ms | 178 ms |
| Current, near region | 16.7 ms | 29 ms | 34 ms |
| Current, far region | 16.7 ms | 94 ms | 101 ms |
| Current, throttled tether | 16.7 ms | 187 ms | 192 ms |
Note the third row against the first. Same region, same hardware, same connection quality. The entire improvement is arithmetic on the tick interval, and it is the difference between a game people accuse of cheating and one they praise.
The natural experiment we have been waiting for
This is the cleanest test our desk has ever had access to, because almost nothing else moved. The patch notes list two weapon adjustments, both to items outside the top ten by usage, and a map lighting change. Everything that determines how a firefight resolves stayed where it was.
Four hundred and twenty of our matches were logged before the change and four hundred and eighty after, on the same accounts, from the same four connections, on the same maps. Deaths where the killer was already fully behind cover on our screen fell from thirty-one per cent to seven. Deaths our testers could explain rose from fifty-eight per cent to ninety-two.
Nobody patched the design. They stopped asking the server to guess, and a game that felt like it was lying started telling the truth.
What the game does with the extra fidelity
Verge Squad's design was always built for a higher update rate than it shipped with, which is why the change lands as well as it does. Weapon recoil patterns are deterministic and tight; the movement has a short acceleration curve; the maps are full of the head-height sight lines that only work if the server can tell who saw whom first.
At sixty ticks all of that becomes legible. Our testers reported that recoil control started feeling like a skill rather than a suggestion, and the peeking duels that define the game's best moments became winnable by whoever was actually faster.
Corrections, and the absence of an indicator
One point two visible corrections per match, which is respectable but not excellent, and the game does not tell you when one has occurred. Something moves, you flinch, and there is no way to distinguish a rollback from your own mistake.
Rooftop Duel and Slipstream Nine both show an explicit indicator and both feel more trustworthy for it. This is the main thing keeping Verge Squad out of the nines, and it is a small piece of interface work rather than an engineering problem.
| Measure | Verge Squad | Archive median |
|---|---|---|
| Tick rate | 60 Hz | 28 Hz |
| Hosting regions | 7 | 4 |
| Visible corrections per match | 1.2 | 3.6 |
| Correction indicator | None | None |
| Matchmaking latency ceiling | 140 ms | None |
Seven regions, and a ceiling that works
Seven hosting locations against an archive median of four, and a matchmaking ceiling of a hundred and forty milliseconds that the game enforces rather than advertises. From our two deliberately distant test positions we waited longer for matches and got better ones, which is the correct trade and one Banner Fields refuses to make.
The ceiling has one flaw: when it cannot fill a lobby it says so vaguely, and players in thin regions reasonably conclude the game is broken rather than being protected. Say the number.
The short version
- Tick rate raised from twenty to sixty with no design changes.
- Peek advantage fell from a hundred and ten milliseconds to thirty-four.
- Deaths behind cover fell from thirty-one per cent to seven.
- Explained deaths rose from fifty-eight per cent to ninety-two.
- Seven regions and an enforced hundred-and-forty-millisecond ceiling.
- No correction indicator, which is the one thing still missing.
Verdict
Eight point nine, and the most useful review we have published because of what it proves rather than what it recommends. Every studio shipping a twenty-tick shooter should read the before-and-after table and understand that the game they are about to be criticised for is not the game they designed.