The Diversity Score: when your YouTube watch history collapses into 3 channels, something is wrong.

What collapse looks like from inside

The strange thing about a doomscroll spiral is that it doesn't feel like one in the early stages. It feels like you're watching good content from channels you like. The problem is not that the content is bad, it's that you're watching it in a way that's increasingly narrow, increasingly reactive, and increasingly driven by the algorithm rather than by your own choices.

By the time you notice something is off, you've often been in the pattern for days or weeks.

The Diversity Score is built to catch that earlier.

How the score works

Diversity is calculated from your watch history across a sliding window, 7, 30, or 90 days depending on which view you're looking at. It measures the breadth of channels, topics, and video formats in your watching, weighted toward recent viewing. A high Diversity Score means you're watching across a range of channels and topics. A low score means your viewing has concentrated.

What makes it a warning signal is the direction of change, not the absolute value. Some people naturally watch fewer channels than others. That's fine. What matters is when your Diversity Score starts dropping steadily over consecutive days, that pattern is the signal.

My 90-day data

Over a recent 90-day period, my Diversity Score looked like this: the first 30 days averaged 74, comfortable across a range of channels and topics. Week 5 and 6 were stable. Then, in roughly week 7, the trend shifted. The 30-day average started moving down: 68, 61, 57, 49.

By the time the score hit 49, I was aware something had changed in my viewing, but I hadn't put words to it. Looking at the underlying data made it legible: 73% of my watch time in that period was concentrated across 4 channels. Three of them were channels I actively chose to watch. The fourth was a channel that had gradually increased in my history through algorithmic recommendation, I'd clicked it a few times, the algorithm had learned, and it was now appearing in 40% of my browsing sessions.

Watch time had not spiked dramatically. Total hours per week were roughly similar to my normal baseline. The diversity collapse was happening silently, within roughly the same total time budget.

Why diversity is the early signal

When a doomscroll spiral starts, the first thing that changes is not how much you watch. It is what you watch. The algorithm identifies the channels and formats that are generating your engagement and leans into them. Diversity contracts. You're still watching for your normal number of hours, but those hours are increasingly shaped by algorithmic narrowing rather than your own choices.

Total watch time typically increases later, as the narrowing deepens and the content environment becomes more effective at keeping you engaged. By that point, the spiral is well established.

The Diversity Score catches the contraction phase. Because it operates on a rolling window, a downward trend over 5-7 days shows up clearly before the watch time numbers move significantly.

The channel concentration metric

Alongside the Diversity Score, Gazenest surfaces channel concentration: what percentage of your watch time in a given period came from your top 3 channels. In healthy weeks, my top-3 concentration is usually around 35-40%. The weeks where I'd call my YouTube use intentional, it's often lower.

When my top-3 concentration hits 65% or above, I know something has shifted. The content might be good. The channels might be ones I genuinely value. But 65% concentration almost always means I've been following the algorithm rather than choosing, because the algorithm, once it identifies what works, will fill as much of your time as you'll let it with those channels.

Seeing the number changes the dynamic. It's easy to tell yourself you're choosing channels you like. It's harder to maintain that story when the data shows 65% concentration in 3 channels over 14 days.

What I did when I saw the trend

When my Diversity Score hit 49 and I understood what had happened, I did two things. I looked at the channel that had grown algorithmically and made a deliberate choice about it, I decided I actually did value it, so I subscribed officially, which moved it from recommendation territory to something I'd acknowledged choosing. And I spent a week using only my subscription feed as a starting point for YouTube sessions, bypassing the homepage.

The Diversity Score recovered over the following two weeks: 57, 64, 71. Not a dramatic intervention. Just a small structural change that interrupted the narrowing.

The measure that matters before the one you're watching

Most screen-time advice focuses on total watch time as the primary metric. Reduce your hours, set a daily limit, take regular breaks. These are not bad suggestions, but they're operating on the outcome rather than the process.

By the time watch time has increased noticeably, the underlying pattern driving that increase is often well established. Diversity is the leading indicator. It tells you what's happening to the structure of your viewing before the total hours reflect it.

I check my Diversity Score more often than I check my total watch time, for this reason. It's the number that gives me enough lead time to change something before the spiral is already underway.

The Diversity Score is available on Gazenest Plus, with 90-day trend view.