Why ping isn't everything, a deep dive on what really matters for game feel

Every gamer knows ping matters. Lower is better. 20ms feels great; 200ms feels bad.

What many gamers don't know: ping is one of several variables that shape "game feel." Two players with the same ping can have wildly different experiences. Two servers with similar pings can feel different. This article unpacks why.

The variables

What contributes to how a game feels over the network:

1. Latency / ping. Round-trip time of a packet.
2. Jitter. Variation in latency.
3. Packet loss. Percentage of packets that don't arrive.
4. Server tick rate. How often the server processes the simulation.
5. Client-side prediction quality. How well the client guesses what happens between server updates.
6. Network conditions of other players. Affects matched gameplay.
7. The game's specific netcode design. Different games handle the same network conditions differently.

A "low ping" with high jitter or high loss is much worse than the same ping with stable network. A "high ping" with smart prediction can feel better than a low-ping server with bad prediction.

Ping is the headline number. The integral of jitter, loss, tick rate, and prediction quality is what you actually feel.

Latency in detail

Ping is round-trip time (RTT): how long a packet takes from your computer to the server and back.

For a typical Minecraft player on a wired connection:

• Your computer to your router: 0-2ms.
• Your router to your ISP's network: 1-5ms.
• Your ISP to the server's ISP: 5-100ms (depending on geography).
• Server processing time: a few ms.

The dominant component is usually geography. Light in fiber covers about 200 km/ms. New York to Los Angeles is ~4000 km, so ~20ms one-way minimum, ~40ms RTT minimum. You can't beat physics.

Other components add to this:

• Wi-Fi: 1-20ms extra.
• ISP congestion: variable, often 5-30ms at peak.
• Server processing on busy server: 1-10ms.

What's "good ping" for what

Game type	<50 ms	50-80 ms	80-150 ms	150-200 ms	>200 ms
Fighting	great	playable	bad	bad	bad
Competitive FPS	great	great	playable	bad	bad
Modern action MP	great	great	playable	bad	bad
Co-op survival	great	great	great	playable	bad
Slow-paced MP	great	great	great	great	playable
Turn-based	great	great	great	great	great

These ranges are rough but useful. They explain why "good enough ping" varies.

Jitter: the silent killer

If ping is 50ms steady, the network is "fast." If ping varies between 30ms and 200ms, the network is "jittery" even though average ping looks similar.

Jitter wrecks game feel because:

• Client prediction can't extrapolate cleanly when updates arrive irregularly.
• You see rubber-banding: objects snapping to new positions as updates arrive late.
• Server reconciliation gets confused: which "current" position is real?

Common causes of jitter:

• Wi-Fi. The #1 cause for home users. Especially bad with weak signal or interference.
• Congested ISP last mile. Especially during peak hours.
• Buffer bloat. Excessive queueing in some network device. Causes latency spikes under heavy use.
• Power-saving on phones / laptops. Idle phases create variable response times.

A typical Wi-Fi connection might have jitter of 5-30ms. Wired ethernet is typically under 5ms. The difference is huge in game feel.

Packet loss

A small percentage of packets are normally lost. Even 0.1% loss is common. Game design typically tolerates this.

When loss gets above 1-2%, problems appear:

• Position updates skip; movement feels jerky.
• Hits register inconsistently.
• Players appear to teleport.
• Voice chat becomes unintelligible.

At 5%+ loss, most games become essentially unplayable.

Causes of loss:

• Wi-Fi interference.
• Overloaded network equipment.
• Faulty cables.
• Congested ISP paths.
• Old or failing modems.

Tick rate

The server's tick rate is how often it processes the simulation and sends updates. Common values:

Game	Tick rate	Tick interval
Minecraft Java	20 Hz	50 ms
Counter-Strike 2 / CS:GO competitive	64 to 128 Hz	7.8 to 15.6 ms
Valorant	128 Hz	7.8 ms
Most modern action games	30 to 60 Hz	16 to 33 ms

Higher tick rate = more frequent updates = smoother feel for player actions and reactions.

But: higher tick rate also means more bandwidth. 128-tick servers send roughly 2x the data of 64-tick. For modern broadband, this is fine. For cellular or limited bandwidth, it can hurt.

A 20-tick game on a stable network can feel better than a 64-tick game on jittery network. Tick rate is one variable among several.

Client-side prediction

Most modern games use client-side prediction: the client guesses what the server will do, applies it locally, and corrects when the server's authoritative update arrives.

When done well, the player sees instant response to inputs even at 100ms ping. When done badly, the player sees stuttering or rubber-banding.

Quality of prediction depends on:

• The game's design (how forgiving / forgivable is prediction).
• How predictable the player's actions are (running in a straight line: easy; complex maneuvers: hard).
• Update frequency from server.
• Quality of server reconciliation.

Some games (Overwatch, modern Source engine games) have excellent prediction. Others struggle. This is why two games with similar advertised ping can feel different.

Lag compensation

Server-side mechanism: the server "rewinds" the game state to account for player latency when checking actions.

Example: Player A shoots Player B at a moment when both are at certain positions. By the time the shot packet reaches the server, both players have moved. Lag compensation rewinds to where they were at the moment of the shot to determine if it hit.

Done well, lag compensation lets players with different pings have a fair-feeling experience. Done badly, it produces the "I shot them and somehow died" experience where each player's view of who hit whom is different.

This is one of the hardest problems in game netcode. Modern games handle it well; older or less-polished games can be janky.

How the other players' networks affect you

In multiplayer, the slowest player's network affects everyone:

In peer-to-peer games: the host's network is critical. Other players are limited by the host's connection.

In dedicated-server games: the server's view of each player is rewound based on each player's ping. A player with high ping or packet loss can cause inconsistent action resolution that affects you ("I killed him but he killed me from across the map a second later").

This is why some games separate matchmaking by ping or region. Mixing players with vastly different network quality degrades everyone's experience.

Geographic latency: physics

The fundamental floor on ping is physics. Light in fiber:

• 200km/ms in glass.
• ~2/3 the speed of light in vacuum.

Some real-world floors:

These are theoretical minimums. Real latency is higher. Game servers in the wrong region can never feel right. No game can be played from Australia to a New York server at 30ms ping. The packets can't physically arrive that fast. This is why game services have regional servers.

Cellular and mobile

Cellular networks add specific challenges:

• Higher base latency (3G: 100-200ms, 4G: 30-80ms, 5G: 10-50ms depending on deployment).
• Higher jitter than wired.
• Variable signal strength.
• Cell handoffs cause brief outages.

Mobile gaming has different netcode patterns than PC/console gaming partly because of these constraints. Mobile games often tolerate higher latency and design around predictable cellular behaviors.

Server proximity and hosting

For game hosting:

Choosing server location matters more than choosing server hardware (within reason). A fast server in the wrong region beats a slower server in the right region.

For small communities: pick one server location near most players.

For larger communities: consider multiple regional servers, with players choosing or being routed appropriately.

For global communities: accept that some players will have worse experience. Or split the community by region.

Anycast for game traffic isn't usually viable (stateful, long sessions). Use unicast servers in specific regions instead.

Bandwidth vs latency

Often confused. They're different:

• Bandwidth: how much data can flow per unit time. Measured in Mbps.
• Latency: how long a single packet takes. Measured in ms.

Most games use very little bandwidth (50 KB/s per player typical). Bandwidth is rarely the issue. Latency almost always is.

Marketing "1 Gbps internet" implies a great gaming connection. It doesn't necessarily, if latency is poor.

How to actually measure

For server admins:

Real measurement of game feel involves combining:

• ping or mtr from typical player locations.
• Real player feedback ("does the server feel responsive?").
• Game-specific metrics (TPS for Minecraft, etc.).

For players:

ping -c 100 to the server: get min/avg/max/loss. The max-vs-min tells you jitter. The loss tells you reliability. The avg tells you latency.

mtr -c 100 to the server: see the path and per-hop reliability.

In-game ping displays are useful but often less reliable than OS tools.

The integrated game-feel equation

Combining everything, the experience of "this game feels good" requires:

• Stable, low-ish latency (under 80ms is the goal for action games).
• Low jitter (max ping no more than 2x min).
• Minimal packet loss (under 1%).
• Server with adequate tick rate.
• Good prediction and lag compensation.
• Other players with reasonable networks.

Failing on any one of these degrades the experience.

For hosting decisions

If you're picking hosting for a game server:

Location is dominant. A great host far from your players is worse than an OK host near them.

Network quality matters. Hosts with extensive peering deliver better latency.

Hardware matters for tick rate. Slow CPUs = low TPS = bad feel. See our "CPU cores vs clock speed" article.

Don't over-pay for bandwidth. Game servers don't need much. Unless you're hosting downloads.

Test before committing. A friend playing on the server for 30 minutes will reveal more than any marketing page.

Conclusion

Ping is one of several variables that shape game feel. Latency, jitter, packet loss, tick rate, prediction quality, server location, and your opponents' networks all matter.

A server with low ping that feels bad is suffering from one of: jitter, loss, low tick rate, or bad lag compensation.

A server with moderate ping that feels great has: stable latency, good prediction, capable hardware, and a fair network path.

When evaluating a game host for latency-sensitive games, look beyond the headline ping. Test for stability. Ask about peering. Verify tick rate. The "feel" is the integral of many small factors.

Coming up

Next: NAT and game hosting, the specific challenges of hosting servers and connecting to peer-to-peer games when home networks use NAT.