How I built the Guild Wars 2 event timer

The event timer is a clock. It reads the fixed UTC schedule every Guild Wars 2 event runs on, converts it to your local time, and counts down, with no login, no API key, and no call to the game at all. This is the build log: the order I did things in, and why each call went the way it did.

Why the timer came first

The event timer was one of the first pages I built for Guild Wars 2, and that wasn't an accident. Two things made it the obvious place to start.

The demand is enormous. “gw2 event timer” and “guild wars 2 event timer” pull somewhere around twelve thousand searches a month between them, the biggest single term in the whole GW2 keyword sweep. People want to know when the next world boss is up, they want it constantly, and they want it now.

And the pages already serving that demand leave an opening. Whatever they're like as tools, the page-level SEO is thin: not much structure, barely any explanation, nothing that reads as answer-first to a crawler or an AI. A term with that much volume where the incumbents haven't done the on-page work is the rare gap a new site can actually walk into. Big demand plus soft optimisation is the combination worth building early, and building properly.

The core bet: it's just a clock

World bosses spawn on a fixed daily rotation tied to UTC, not to your server. Every megaserver instance shares that one clock, so a timer never has to read the game, it compares now to a published table. That's the whole trick.

Which means the entire thing runs in the browser. No backend, no account, no waiting on anyone's server. It loads instantly and works on a phone with nothing signed in, because there's nothing to sign in to.

The catch: the official Guild Wars 2 API has no live event endpoint. I checked early, expecting to need it. Turned out I didn't. A fixed schedule doesn't need a live feed, it needs the right numbers. So the whole question became getting the numbers right.

Getting the numbers right

One rule, set before I wrote a line of the schedule: never type a spawn time from memory. A timer that's wrong is worse than no timer: someone misses the boss, blames the page, and never comes back.

The first pass leaned on the wiki's World boss page for core Tyria and Heart of Thorns. Good enough to start, not good enough to trust at scale. So I went to the better source: the wiki's own Widget:Event timer/data.json, version 5.1, the same data the in-game /wiki et panel runs on. If it's right in the game, it's right here.

That file stores each map as a repeating pattern of phases, plus a partial stretch that plays from midnight. It isn't a list of times, it's the recipe for one. So I rebuilt every map's timeline from it in code, pulled the exact spawn minute for each event, and ran the whole schedule through a script to confirm it lined up before any of it went near the page.

That check earned its keep on day one. It caught two of my own early mistakes: Dragon's Stand was sitting on the wrong offset, and the Verdant Brink night bosses were off by nearly two hours. Both were wrong in the hand-built version. Both are right now, because the script disagreed with me and the script was holding the real data.

What I built that no tool asked for

Most of the timer exists because it's good, not because anything flagged it. SurferSEO never saw any of this. It's product, not optimisation.

Two timers in one

A toggle up top: World bosses, Meta events, or All. Pick one and you see only that kind, grouped by release. Leave it on All and everything sits in a single list with whatever's live floating to the top. Same engine, three ways to read it.

Sound alerts

Tap the bell on any event and the timer plays a heads-up two minutes out and again at spawn, and flashes the browser tab so you catch it from another window. Your picks are saved on your device, so they hold between visits. Built on the Web Audio API, the sounds are generated in the browser, so there's no audio file to host and nothing extra to load.

Everything, not just the classics

I didn't stop at core and Heart of Thorns. The timer carries every fixed-schedule event I could verify: Path of Fire, the Living World seasons, the Icebrood Saga, End of Dragons and the Soo-Won fight, Secrets of the Obscure, Janthir Wilds, Visions of Eternity, and the Eye of the North public bosses. Roughly forty events across ten releases, all on the one clock.

The full world-boss table

Late on, I noticed only six bosses had a write-up while the timer tracked forty. That reads as arbitrary. Why these six? So I added a table covering all thirteen core bosses at a glance, with the write-ups kept as spotlights on top. Surfer happened to suggest a table around the same time. I'd have built it anyway; the reason was the inconsistency, not the score.

And the things I left out on purpose

Living World Season 3, Drizzlewood Coast, the festival events, anything that runs continuously or only appears for a few weeks a year stays off the timer. A clock that counts down to something that's always on, or gone for ten months, is a clock that lies. The wiki leaves them off for the same reason, and I followed it.

Then I ran SurferSEO

Surfer scores a page against the pages already ranking for a keyword. Mine is guild wars 2 event timer (and the short gw2 event timer), the headline term at roughly twelve thousand searches a month, the one people actually type.

I treated the content score as a coverage map, not a quota. The useful question it answers is “what do the ranking pages cover that I don't?”, never “how many more times can I say the keyword.” Density chasing is how a page ends up sounding like a page about saying its own name.

So every cluster it surfaced got a real home, or it got dropped. The boss names became a spawn-times section with each boss as its own heading, useful on its own merits, and it carries great jungle wurm, frozen maw, claw of jormag and the rest without anyone forcing them. The trademark terms became an attribution footer, which a fan tool should carry anyway. The alert words became an alerts FAQ, because I'd actually built alerts to write about.

What I didn't do is stuff. Surfer also scrapes the chrome off competitors' pages: bar, mouse, cancel, icons, default, the wording of cookie banners and settings menus, not content. Forcing those into prose reads as noise to a person and as noise to Google. The score doesn't need every term lit to rank, and the page reads better without them.

Surfer's suggested opener: “This page provides a live GW2 event timer covering world bosses and map metas across all of Tyria.”

What I shipped: “A live Guild Wars 2 event timer covering world bosses and map metas across all of Tyria…”

Same keyword, same first sentence, without the “This page provides” throat-clearing. The tool wanted the phrase; it didn't need the preamble.

The AI-readability checklist

Surfer also runs an “AI readability” pass. Mine came back zero of thirteen. I read all thirteen and split them down the middle: the ones that made the page better, and the ones that would have made it worse.

The six I kept:

• Full schedule table. Scannable, and it closed the coverage gap I'd already spotted myself.
• Define the terms. One line on what a world boss and a meta event are, and the 2–3h cycle, for the player who doesn't know yet.
• Note the 00:00 UTC reset. True, and it answers a question people actually have.
• An audience line. Who it's for, why it matters, new player chasing a first Tequatl, veteran running a boss train.
• Name the alternatives. A nod to /wiki et and community trackers. I already linked the panel; the nod costs nothing.
• H3 subheads in the explainer. Breaks a wall of text into things you can skim.

The four I dropped:

• “Move the explanation above the timer.” No. The tool is the point, you don't bury it under three paragraphs of preamble. Crawlers can scroll; the player checking when Tequatl's up shouldn't have to.
• “Add 'Now, let's look at…' transitions.” Dated filler. Against the voice guide on sight.
• “Rewrite the first sentence to 'This page provides…'.” Clunkier, and the keyword is already in the opener.
• “Bulletise the FAQ boss list.” The table already gives the scannable version; the prose answer suits the voice.

And two it flagged that were already there:

• “Add an H1.” There is one, the page title. Surfer just isn't seeing it.
• “Make each FAQ question an H3.” They already are.

How I weighed it

When two of these pull against each other, and they do, the order is fixed. Higher up the list wins.

1. Accuracy. A wrong time kills trust faster than a missing one. Everything else is downstream of the numbers being right.
2. The player using the tool. The timer stays at the top of the page. The reading is for after they've got their countdown.
3. The voice. No robotic filler buys a green bar. If it doesn't sound like Crit Pick, it doesn't ship.
4. Keyword coverage. Real, and only where a term has somewhere honest to sit.
5. The checklist score. Last, and optional. A number to glance at, not a target to hit.

Surfer sits at the bottom of that list on purpose. It's a sharp map of what the competition covers, and I used it as one. It doesn't get a vote on the parts that matter most.

Still open

The page isn't finished.

The core world bosses have a table and write-ups now. The forty-odd expansion and meta events still only live in the widget, not in prose. A second table, expansion and meta times, is the obvious next step, and it'd bring the written coverage level with the tool.

The bigger play is splitting the page in two: a dedicated world boss timer and a meta event timer, sharing the one engine. “gw2 world boss timer” and “gw2 meta event timer” are their own searches, three pages for three terms instead of one page for one. I held off because the single page with a toggle is the better experience to use. The split is a ranking decision, not a product one, and I'll make it when the ranking case is worth the extra surface.

living document, grows with the site.