So the election happened Tuesade night into Wednesday morning, and yeah, obviously that's the story everybody's been writing about since. I'm not going to rehash who won or what it means for the next four years. Half the internet is already doing that and doing it better than I could. What I actually want to talk about is a much dumber, much more specific thing: Twitter's own numbers from that night, which they posted on their engineering blog yesterday.
The peak minute, right when the networks called the race for Obama around 11pm Eastern, hit 327,452 tweets per minute. That's the number Twitter put out. Just sit with that for a second: not per hour, per minute. I was staring at my phone at the time, refreshing my timeline like an idiot, and I remember it actually lagging for a few seconds right around when CNN made the call. I assumed that was my crappy apartment wifi being my crappy apartment wifi. Turns out it might've genuinely been a traffic spike big enough to strain things, even briefly, at a company that's supposed to be built for exactly this kind of moment.
I've been thinking about this more than the actual result, honestly, because it's the part of the night that's actually new. Presidential elections happen every four years and the format barely changes: anchors, maps, a concession speech, a victory speech. But this was the first one where a huge chunk of the country was narrating it to each other in real time, in public, on a platform that didn't exist during the last one that mattered this much. In 2008 Twitter was three years old and mostly full of people telling you they were eating a sandwich. By Tuesday night it was apparently the infrastructure equivalent of a small country's entire population trying to log into the same website at once.
What gets me is how casually Twitter reported this. It read almost like a weather update: here's the peak, here's how it compares to past records, moving on. No huge victory-lap post, no press conference. Just a paragraph and a chart. I kind of respect that. Compare it to how a company like Facebook handles this stuff, with the giant blue "I Voted" button they were pushing all week that felt a little more like a data-collection exercise dressed up as civic pride. I clicked it because my friends kept posting screenshots, and then immediately regretted it because now I'm sure that gets logged somewhere forever.
(Small tangent, not really related, but I finally got my hands on an iPad Mini this week, picked one up Tuesday afternoon actually, right before all the election stuff kicked off, and I have Opinions. It's a genuinely nice size to hold, way better than the full iPad for reading in bed, but the lack of a Retina display is more noticeable than I expected once you've been staring at a Retina iPhone for a year. $329 for the 16GB WiFi model felt fair walking into the store and felt slightly less fair once I actually used it next to my old iPad 3. Anyway.)
Back to the Twitter thing, because I think it actually matters more than it looks like on the surface. Every one of these moments, the tweet volume during a debate, a Super Bowl blackout, now an election night, is basically free stress-testing for infrastructure that used to only get tested in a lab. Nobody plans a load test for "327,000 tweets in sixty seconds," you just find out live, on the night, in front of everyone, whether your systems hold. They held. Barely lagged for me, and that was probably my own connection more than theirs.
I don't have some grand point to wrap this up with. I just think it's a more interesting number than most of what's going to get written about this week, and I wanted it written down somewhere that isn't a tweet that'll scroll away in ten minutes.