You dashed behind the wall. The kill cam shows you standing in the open. That’s favor-the-shooter: Overwatch’s netcode credits the attacker’s version of events, so every spike turns deaths you escaped into deaths the server awarded. 70 ms is the practical line — and stability matters more than the number. Run the 25-second test below; it counts the spikes the in-game overlay forgets.
Under 70 ms and steady. That’s the threshold LagScope scores Overwatch 2 against in its per-game readiness chips — comfortably under it, the game feels honest: cooldowns land when you press them, and deaths match what you saw on screen. Between 70 and about 105 ms you’re playable but donating a slice of every duel to the netcode — you’ll notice Suzu and Lamp going off “after” you died. Past that, you’re playing the kill cam, not the match. The average is only half the story, though: Overwatch 2 punishes variance more than altitude. A steady 85 ms line plays better than a 40 ms line that spikes to 300 every fight, because the spike is what rewinds your position after you’ve already made the play. That’s why the test above reports spikes, jitter and loss next to the average — the four numbers together are the verdict.
| Average ping | Verdict for Overwatch 2 |
|---|---|
| Under 70 ms | Competitive-ready. The server’s version of events matches yours. Dying behind cover becomes rare instead of nightly. |
| 70–105 ms | Playable, not ideal. Defensive cooldowns feel late. You start losing the “I was already behind the wall” arguments with the kill cam. |
| Over 105 ms | Rough. The server keeps believing the other player. Fix the connection before grinding ranked. |
Short answer: the shooter’s client told the server about the shot before your client told it about your dodge — and the server believes the shooter. This is deliberate design, not a bug: Blizzard built Overwatch’s netcode to favor the shooter so that firing feels responsive, which means the defender’s latency is where deaths behind cover are born. Your own latency and, worse, your spikes decide how often the server’s replay contradicts your screen. Packet loss makes it weirder: lost updates arrive as teleports and snapped-back positions. Blizzard’s own connection troubleshooting guide points at the same home-side suspects we do — Wi-Fi, saturated lines, background traffic. To see what the game sees, open the in-game network graph (Ctrl+Shift+N on PC) and watch RTT during a team fight — then compare it with the timestamped log from the full monitor below, which remembers every spike the overlay already forgot.
Check your own line first — 25 seconds above. Flat baseline, no spikes, no loss, while the whole lobby is rubber-banding? That’s server-side; watch Blizzard’s channels and wait it out. Spikes or loss at the exact moments you felt them? That’s your connection — usually Wi-Fi or a loaded line, not Blizzard.
Usually one of four: Wi-Fi interference, bufferbloat when the line is busy, bad ISP routing to Blizzard, or a far region’s server because of who you grouped with. The test above counts spikes and loss — not just average ping — so you can see which pattern matches.
It’s a gamble, not a setup. Wi-Fi averages look fine, but interference causes exactly the short spikes and loss that a favor-the-shooter netcode converts into deaths behind cover. One Ethernet cable is the highest-winrate change most players can make — measure your link and see.
The 25-second test is a snapshot. The full LagScope monitor watches your connection for as long as you play — lagspikes, outages, bufferbloat, 24-hour history — and builds a timestamped evidence report you can show your ISP. Free, no signup, runs in your browser.
Open the full monitor