My 41-week spreadsheet shows the loss arrived in three flat stair steps

2026-11-10 1860 views 4 posts

Sunday night I updated the main tab and the 41-week median line did something I’d been half-expecting: it crossed 52.6 lb down, but only after nine flat days. My 7-day average had been parked inside a 1.1 lb band, while the daily scale dropped 2.2 lb over two mornings and then bounced back by 1.4.

In my log that looks less like a smooth line and more like three flat landings with short drops between them. I marked the new low, wrote 'median catching up' in column H, and left the rest alone.

Mildly annoying part: I’d spent two days trying to explain the flat stretch with sodium and sleep notes before admitting the chart already had the pattern. Anyone else tracking weekly medians see these stair-step stretches, or is your trendline smoother than mine?

This is the same pattern from that old “whoosh vs. stall” thread—flat stretches followed by a two-morning drop. I added a monthly waist measurement and a tiny note column for salt/sleep to my spreadsheet, and my flat patches almost always lined up with restaurant week or bad sleep, not actual fat changes.

Are your nine flat days clustered around anything like that, or is it just the scale being dramatic? Either way, I’d trust your 41-week median way more than the daily bounce.

That two-morning 2.2 lb drop then 1.4 bounce is classic scale noise plus a whoosh—water and gut content can swing that much. What I’d add to the spreadsheet: a column for salty/carb-heavy meals and one for hard workouts or soreness, since inflammation can hold water and mask fat loss for days.

In the trials, mean curves look smooth, but individual daily weights are notoriously spiky; your 7-day median is probably doing more work than the daily line. Did the nine flat days line up with travel, salty food, or DOMS?

Weekly waist might also catch change the scale hides.

Systems take: the median crossing right after nine flat days smells like a bucket-boundary artifact—weekly medians will stair-step even when daily data is continuous. Actually I think the 1.1 lb is more like peak-to-trough range, not a ±1.1 band around the mean, so the flat stretch may be tighter than it reads.

I’d plot a 14-day rolling median beside the 7-day average; if the stair stays, it’s trend, if it smooths, it’s sampling.