Ping typically means the diagnostic command and its reported round-trip time (RTT), while latency is the broader term for packet delay and can mean one-way or round-trip delay depending on context. Most people only ever see the ping number, but one-way latency and jitter matter just as much for gaming, calls, and streaming. Here's how the two relate, how to measure them properly, and what to do with the results.
TL;DR:
- Ping reports round trip time, so it can hide a slow return path; true one way measurement requires synchronized clocks at both ends.
- Competitive gaming generally calls for latency below 30 ms, while most real time uses work well under 50 ms; above 150 ms, lag becomes noticeable.
- Run at least 20 to 30 pings and compare minimum, average, maximum, and deviation; spikes can disrupt calls even when the average stays low.
- Pair a speed test with repeated pings, then use traceroute if results look abnormal, because ICMP may not reflect application traffic.
Table of Contents
- What is ping, and what does it measure?
- What is latency, really?
- Ping vs latency: the key differences summarized
- How ping and latency shape gaming, calls, and streaming
- Tools and methods for measuring ping and latency
- Reading your results and deciding what to do next
- How our tools go beyond a basic ping test
- What actually matters once you understand the numbers
- FAQ
- Sources
What is ping, and what does it measure?
Ping is a command-line utility that sends an ICMP echo request to a destination and waits for an echo reply. The time between sending and receiving that reply is the round-trip time, and that RTT is what most people mean when they say "my ping is 40."
A typical ping run prints a summary with min/avg/max/mdev values alongside packet loss, and each tells you something different, according to the ping(8) Linux manual page:
- Min shows your best-case round trip under current conditions.
- Avg gives the typical experience across the sample.
- Max flags the worst spike, which matters more than the average for anything real-time.
- Mdev (mean deviation) approximates jitter, or how much individual pings wander from the average.
ICMP traffic doesn't always travel the same path or get the same treatment as your actual game or video traffic. Some routers deprioritize or rate-limit ICMP packets, so a clean ping result can still coexist with a choppy video call. Running a single ping rarely tells the full story. Repeated samples over a minute or two reveal whether delay is stable or spiking, which is the detail a one-off test hides.
What is latency, really?
Latency is the general concept of delay as a packet moves across a network, and it comes in two flavors: one-way delay (source to destination only) and round-trip delay (there and back). Ping measures the second kind by default; the first kind is harder to pin down and far less commonly tested.
Measuring true one-way delay requires tightly synchronized clocks at both ends, since you're comparing a timestamp from one machine against a timestamp from another, as RFC 7679 explains. That synchronization problem is exactly why most everyday tools default to round-trip measurement instead: it only needs one clock.
Several things add up to create the latency you experience:
- Propagation delay: how long the signal takes to physically travel the distance, bounded by the speed of light in cable or fiber.
- Transmission and serialization delay: the time to push bits onto the link.
- Queuing delay: time spent waiting in router or switch buffers during congestion.
- Processing delay: time servers and network equipment spend handling the packet.
Jitter (variation in delay between packets) and packet loss ride alongside latency and often matter more for how a connection actually feels.
Ping vs latency: the key differences summarized
The short version: ping is a tool and a round-trip number; latency is the broader concept that number is trying to describe. Here's where that distinction actually changes how you troubleshoot:
- Ping is a command, latency is a metric. You run ping; latency is what it (and other tools) attempt to quantify.
- Round-trip vs one-way. Ping reports RTT by default; true one-way latency needs synchronized clocks and specialized tools, per RFC 7679, so it's rarely measured outside enterprise monitoring.
- Different jobs. Ping is built for quick reachability and rough RTT checks; dedicated latency measurement supports capacity planning and quality-of-service work.
- Asymmetry hides in RTT. Because round-trip measurement blends the forward and return paths, a fast outbound route can mask a slow return route, according to RFC 2681.
Pro Tip: If your ping looks fine but calls still stutter, suspect an asymmetric route or jitter rather than assuming the latency number is lying to you.
How ping and latency shape gaming, calls, and streaming
Numbers only matter in context. Competitive gaming generally wants round-trip latency under 30 ms, most real-time activities are comfortable under 50 ms, and anything consistently above 150 ms starts producing noticeable lag in calls and interactive apps, according to TP-Link's guidance on good ping and latency.
A connection averaging 40 ms can still feel terrible if jitter swings 30 ms in either direction or if packet loss is clipping audio and dropping frames. Averages flatten out the spikes that your brain actually notices.
A few quick reference points:
- Latency under 30 ms is excellent for competitive, twitch-reflex gaming.
- Latency between about 30 ms and 50 ms is solid for most games and video calls.
- Latency between about 50 ms and 150 ms is noticeable but often workable; ITU-T Recommendation G.114 treats one-way delay below roughly 150 ms as largely transparent for most interactive use.
- Above about 150 ms of latency, expect visible delay in calls and gameplay, with 400 ms as an upper planning limit for exceptional cases.
For gaming specifically, our guide on diagnosing high ping in games walks through isolating the cause in under half an hour.
Tools and methods for measuring ping and latency
Ping, traceroute, and browser-based speed tests each answer a different question, and pairing them gives a much clearer picture than any one alone, as TechTarget notes in its explainer on ping.
- Ping confirms reachability and gives a quick RTT baseline with packet loss.
- Traceroute shows the hop-by-hop path, which helps you spot exactly where delay or loss is introduced.
- Speed tests add throughput, jitter, and often loss in one pass, closer to what your applications actually experience.
- One-way delay tools exist for specialized monitoring, but they depend on synchronized clocks, per RFC 7679, which is why they're uncommon outside dedicated infrastructure.
A practical sequence: run a speed test first, then ping your target multiple times to check for consistency, then traceroute if something looks off, then drill into per-hop or one-way diagnostics only if the problem is still unclear. Traders chasing microsecond-level execution windows have written about this kind of layered diagnosis too; the approach in this guide to reducing trading latency is more extreme than most people need, but the sampling discipline translates well to everyday troubleshooting.
Pro Tip: Run at least 20 to 30 pings before judging a connection. A single ping tells you almost nothing about stability.
Reading your results and deciding what to do next
The average matters less than the spread. A low average with a high max or high mdev means intermittent spikes, which is often more disruptive to a call or match than a slightly higher but steady number.
- Check min/avg/max/stddev together. A tight range across all four means a stable connection; a wide gap between avg and max means something is causing periodic spikes.
- Try quick fixes first. Switch to a wired connection, reboot the router, test at a different time of day, or try an alternate game server or video call region.
- Collect evidence before contacting support. Save traceroute output, repeated ping logs, and a history of speed-test results so your ISP has something concrete to work from instead of a one-time complaint.
For gaming specifically, the fixes in hitting under 20ms ping for gaming cover the router and ISP-side adjustments most people miss.
How our tools go beyond a basic ping test
A ping command tells you RTT and packet loss, but it won't tell you what streaming resolution your connection can actually sustain or whether your router is adding delay under load. Our speed test measures latency, jitter, packet loss, bufferbloat, and DNS resolution time in a single browser-based run, on Wi-Fi or cellular, from phone, tablet, or laptop.
A few things our tests surface that a plain ping check misses:
- Bufferbloat, which spikes latency specifically when your connection is under load, the exact moment a call or match needs stability most.
- Maximum supported livestream or broadcast resolution, calculated automatically from your tested speeds.
- Test history, so you can compare today's results against last week's instead of guessing whether a problem is new.
Broadcasters and streamers dealing with upload-side instability can check our livestream monitor for ongoing upload tracking during a live session.
What actually matters once you understand the numbers
The obsession with chasing the lowest possible ping number is a bit misplaced. A stable 45 ms connection with minimal jitter will feel better in a video call or a match than a connection that averages 20 ms but spikes to 150 ms every few seconds. Stability is the underrated variable, and most casual troubleshooting advice skips it entirely in favor of chasing a single lower average.
The other gap is direction. Because ping blends your outbound and return paths into one number, a lot of people fix the wrong thing: they upgrade bandwidth when the real issue is a congested hop three routers away, or they blame their own setup when the problem sits on the other end of the call. Bandwidth and latency are genuinely different problems, and more of one rarely fixes the other.
If you only take one thing from this: look at your max and your jitter before you look at your average. That's where the real experience lives.
— Peter D
FAQ
Is 100 ping a 1 second delay?
No. Ping is measured in milliseconds, so 100 ping means 100 milliseconds, which is one-tenth of a second. That's noticeable in fast-paced games but far from a full second of delay.
Is 200 ms latency bad?
For most real-time uses, yes, delays above 150 ms are on the high side. ITU-T guidance treats delays beyond roughly 150 ms as noticeable, and competitive gaming or smooth video calls generally need much lower numbers.
What is good ping and latency?
Under 30 ms is considered excellent for competitive gaming, and under 50 ms works well for most everyday real-time use, according to TP-Link's guidance. Jitter and packet loss matter too, since a low average can still feel rough if either one is high.
Is 27 ms latency good?
Yes, a latency under 30 ms is considered excellent for competitive gaming, per TP-Link's guidance. As long as jitter stays low and packet loss is minimal, those conditions support smooth real-time performance across gaming, calls, and streaming.
Sources
- TP-Link blog: What Are Good Latency & Ping Speeds?
- RFC 7679 - A One-Way Delay Metric for IPPM
- RFC 2681 - A Round-trip Delay Metric for IPPM
- ping(8) Linux manual page