Hacker News Reader

Comment by jmkni | Hacker News Reader

As somebody from Northern Ireland, I'm really interested in the fact that NI had the biggest drop in traffic as opposed to England/Scotland/Wales after 4PM.

Is this just on that one day or is this normal for NI after this time? Also what is the impliciation being made here?

View on HN

It’s all as compared to the week prior, so the take away would be that folks from NI are slightly more likely to be offline on election day after 4pm than folks from other areas, where they’re more likely to be offline on election day slightly earlier in the day. Idk what more there is to say about that.

The whole piece is more of a “look we’re cloudflare: we process a lot of traffic and can analyze data!” marketing piece than anything.

My guess - US holiday. The time for the drop in traffic chimes with the usual period we have to jump on Trans Atlantic zoom calls. So this day there was a lack of those calls and hence a drop in normal traffic. NI is worse because it’s the hub for UK and even European trans Atlantic traffic.

It’s hard to say as this is not measuring packets but weird sort of CDN proxy effects and I am just guessing :-)

But I suspect it’s not because Putins elections trolls forgot about Belfast for an afternoon :-)

The changes in traffic volume were pretty small (although NI a bit higher). I don't think comparing to only one week earlier tells too much. They should have compared at least 8 previous Thursdays to understand how much variation is normal and how much can really be attributed to something special.

For other measurements where they noted several 100% of increase that much scrutiny seems unecessary.

Might just be a bigger after-work voting rush. I’d assume turnout was higher in NI than elsewhere, in that the result was more in question than in most places.

Reply on news.ycombinator.com