What is server time, and why is it different from your clock?
When a concert sale opens "at 8:00 p.m. sharp," whose 8:00 is that? Not your phone's and not your computer's. It is the ticketing site's server clock. The moment the server's clock reads 8:00, the booking button goes live, and any request that arrives earlier gets a "sales haven't started yet" message.
The catch is that the server's clock and your device's clock are never exactly the same. The gap is usually tens to hundreds of milliseconds (a millisecond is 1/1000 of a second), and on a poorly maintained server it can be several seconds. For popular shows the best seats are gone within a second or two, so that small gap can decide the outcome.
Why clocks drift apart
Computers and phones keep time with a tiny quartz crystal oscillator. Temperature and manufacturing differences make it run very slightly fast or slow. Even 0.1 seconds a day adds up to a noticeable gap after a few days.
That is why most devices regularly correct their clocks over the internet against a reference time server, using NTP (Network Time Protocol). But how often they sync, and against which server, differs from device to device and company to company.
- Your phone syncs through your carrier's network or the manufacturer's time server.
- Your PC syncs on the schedule your operating system sets, sometimes only every few days.
- The ticketing server syncs however that company manages its servers.
Three clocks set at different times against different references will naturally disagree. It is the same reason the seconds tick over at slightly different moments if you put your phone and PC side by side.
The second variable: network delay
Clock differences are not the whole story. Time also passes between the moment you click and the moment your request reaches the server. That is called latency, and the full time for a request to go out and a response to come back is the round-trip time (RTT).
On a wired connection within Korea, a round trip to a Korean server usually takes a few tens of milliseconds. On mobile data or crowded Wi-Fi it is longer and less steady. As a rule of thumb, your request reaches the server in about half the round-trip time.
In short, to click right on the hour you need to know two things: ① how far ahead of or behind your clock the server's clock is, and ② how long your request takes to reach the server. A server time site measures both and shows you what time it is by the server.
How server time is measured
When you visit a website, the server sends a few pieces of information in the response headers along with the page. One of them, Date, holds the current time as the server sees it. Measuring server time is, at its core, reading that value.
But the Date value only goes down to the second. It says "20:00:03" without telling you whether that is 3.1 or 3.9 seconds. Taken at face value it can be off by up to a full second.
Catching the moment the second changes
The way around this limit is to ask many times. By recording precisely when each request was sent and when each response came back, and comparing that with the second the server reports, you get a window in which the server's clock must lie. Sending requests at slightly different moments narrows the overlap of those windows until the moment the server's second ticks over is pinned down to within tens of milliseconds.
Chochim sends ten such requests to each ticketing site, takes the middle of the overlapping window as the server's clock, and shows half the window's width as the margin of error (±). The smaller the margin, the more reliable the reading.
Correcting for your own device too
The measurement happens on Chochim's server, so we also measure the gap between your device and our server. When you open the page, a few short signals go back and forth, and we correct your device's clock using the fastest round trip. That is why the time on screen follows the ticketing server's clock, not your device's.
What to keep in mind when reading the numbers
- If the margin is large, measure again. Your connection may have hiccupped. Wired beats Wi-Fi, and Wi-Fi usually beats mobile data.
- Check again just before the sale. Server clocks keep drifting, and admins sometimes reset them. Re-measure about ten minutes before opening.
- Some sites can't be measured accurately. Some put security or caching servers in front, which may show a different clock from the real booking server. Treat those readings as a reference only.
- On sites with a queue, timing is not the only factor. Queue numbers are assigned in order of arrival at the server, so entering once at exactly the right time matters most.
So when should you click?
Your request needs to arrive at the server's top of the hour, so in theory you would click at "the hour minus the one-way delay." But arriving too early earns you a "not open yet" message and a retry, which costs more time. That is why Chochim recommends aiming for the server's top of the hour but clicking a touch late, 0.1 to 0.3 seconds. The full strategy is in Concert ticketing, start to finish.