Starlink ping for gaming: the published numbers and what your game will show

Every published number, with dates

Starlink and Ookla are the only two sources with US-wide figures, and they measure different trips. Starlink collects round-trip times from its routers to the internet. Ookla uses Speedtests that people run on their own phones and computers to several test servers, so home Wi-Fi is included. Read each row with its method.

The trend is the encouraging part. Starlink’s own median fell to 33 ms by its March 2024 report and to 25.7 ms by June 2025. Ookla’s device-based figure moved the same way, from 45 ms in Q1 2025 to 39 ms in Q1 2026. Alaska and Hawaii remain the exceptions: Starlink warns that remote places including Alaska see higher latency.

Why does your game show a higher number?

Starlink’s app is clear about what its speed test does: by default it times the link from your router out to the internet, and Starlink adds that lower numbers mean less lag in games. So the app’s latency stops at the router’s connection to the internet. Your game’s number covers more:

• Your house. Wi-Fi adds delay and the odd lost packet, more through brick, stone or several walls. The app’s Device to Router test shows the Wi-Fi part; if its download speed is lower than the router-to-internet test’s, Wi-Fi is your limit.
• The server’s distance. From the ground station, traffic rides ordinary fiber to the game’s data center. A home in Northern New England matched to a US West server pays for the whole country, on Starlink or on fiber. Starlink doesn’t publish which ground station serves a given home.
• How the game counts. Some games show the network round trip, others add server processing time. Two games on the same line can show different numbers (our reading).

That’s why the fair comparison is your in-game ping on Starlink against your in-game ping on the line you have now, on the same server, at the same time of day.

Good ping by game type (rule of thumb)

No standards body publishes “good ping” by genre. These bands are our rule of thumb from how the games are built: how fast the action is, and whether the server smooths over delay. They are not Starlink figures.

Set against that table, Starlink’s published median lands in the first column for every genre, and its p99 line is never worse than the second. The in-game number for a nearby server should usually stay in those two columns. Where it fails is the right-hand column, which is about steadiness rather than the average. Title-level notes are on which games play well.

What do the averages hide about spikes?

A median of 25.7 ms says nothing about the one second your screen froze. Here is what Starlink does and doesn’t say about those seconds.

• Satellite switching. A Starlink dish hops between satellites many times a minute, and Starlink says most of those hops go unnoticed. Some will land mid-match anyway.
• No published interval. Starlink does not publish a fixed satellite handoff or network reconfiguration interval (such as 'every 15 seconds'); the 15 seconds in its own reports is how often routers report latency samples. Treat any fixed handoff timer you read about as unconfirmed.
• No jitter figure. Starlink does not publish a jitter or packet-loss figure for its US service.
• Obstructions. A tree, roof or pole that keeps blocking part of the dish’s view brings short outages, lost packets and lower speeds, which gamers feel as lag spikes and disconnects.
• Weather. hard rain, a snowstorm or hail can make the link stutter for a moment.
• What the app records. Starlink checks uptime 10 times a second on every dish and shows any outage of 100 ms or longer in the app. After a rough patch, the app can also warn about elevated packet loss over the last 15 minutes; Starlink blames weather, a temporary network issue or obstructions.

Spikes are the reason a speed test can read fine while a match stutters. The lag guide takes them in order: wire it, clear the sky, then look at the router and the server region.

How do evenings and plan priority change ping?

Starlink treats 6 pm to 11 pm local time as peak hours. Its 2024 latency report used a narrower window: it focused on peak hours of 6-9 PM local time and averaged samples over 15 seconds. Either way, peak overlaps with weeknight gaming in every US time zone, and the June 2025 median is a peak-hour figure, which makes it the realistic one to plan around.

Plan priority matters at those hours. Starlink says Residential Max outranks the 100 and 200 Mbps Residential tiers, so homes on those two are likelier to slow down on a congested network, and that Residential Max sits at the top of the Residential priority order. Starlink publishes no separate ping figure by plan, so how much that changes your evening ping is not something anyone can quote. Inside the house, Starlink says the queue manager on its Wi-Fi router keeps one person’s big download from raising another person’s game latency.

Reading your own numbers

Once you have Starlink, the app’s Statistics show speed, uptime, latency, outages and alerts, viewable remotely. To compare fairly:

1. Wire the console or PC, or at least test once wired, so Wi-Fi isn’t in the result.
2. Note the game’s ping on the same server region at two times: mid-morning and around 9 pm local.
3. Check the app for outages and alerts after a bad match. A cluster of short outages points at the sky, not the network.
4. Compare with the same game on your current line, if you still have it.

The route a packet takes explains each hop if a number puzzles you.