The headline shows the likely number; the range is one tap away#
Since September 24, 2026, the main screens show plain numbers. A meal tile reads "280 kcal", not "≈ 280 kcal", and the Today totals, Plan lanes, Library rows and History drop the approximately sign too. The small badges that used to sit on meal cards, "Photo estimate", "Label-based", "Mixed evidence", "Estimated", are gone as well. What stayed is the range. Next to the calories on a meal tile, a small ±20 chip summarizes the range's width. To see the actual low and high estimates for calories and protein, tap the chip on the log receipt or open the meal's details.
So nothing about how the estimates are made changed with this release. Where the uncertainty is displayed did, and the reasoning was about scanning: the Today screen is often checked for one number, and we wanted the marks on it to carry information. The headline carries the one number you act on; the detail carries how sure that number is. This post explains the move and the research on showing uncertainty that it rests on. Why calorie counts come with a range in the first place is covered in why calorie counts are ranges.
What moved where#
| Before September 24 | Now |
|---|---|
| "≈ 280 kcal" on meal cards and tiles | "280 kcal" with a ±20 chip |
| "≈" on Today totals, Plan lanes, Library, History | Plain numbers |
| Evidence badges on every meal card ("Photo estimate", "Label-based", "Mixed evidence") | No badge; a Missing detail marker only on meals that lean on an assumed amount |
| Range line and ±% chip in meal details | Unchanged |
| Receipt total | "488 kcal ±50"; tap the ±50 to see the range for calories and protein |
Two things did not move at all. Meal details still show the range line and the ±% chip, next to a Source section describing what the estimate was based on, such as a photo, a label or your description. And the weekly report still draws each day as low, likely and high, because a week is exactly where uncertainty adds up. The evening layout these numbers sit in is described in the Finish Today redesign.
Why the approximately sign went#
The repeated ≈ sign marked estimates as approximate, but did not distinguish narrower ranges from wider ones: a scanned label and a guess about a party plate both said "≈". The redesign aims to make that distinction easier to scan. The evidence badges had a related problem. A badge that says "Estimated" on every meal describes nearly every meal, so it carried little information while adding reading work on the busiest screen.
The replacement is a pair of signals that vary. The ±N chip is small when an estimate is tight and larger when it is loose, and a larger number after ± indicates a broader range. The Missing detail marker appears only when a specific amount was assumed, which makes it rare enough to notice. Both keep the uncertainty visible without asking you to decode a symbol on every line.
A mark that appears on every number cannot tell narrow estimates from wide ones.
Does showing a range undermine trust in the number?#
This was the real risk of keeping ranges at all, and the research on it is more reassuring than intuition suggests.
A team at Cambridge ran five experiments, including a preregistered replication with a national sample and a field experiment on the BBC News website, with 5,780 participants in total1. People read short news-style texts about figures such as unemployment or tiger numbers, with no uncertainty (just a point estimate), with a numeric range added to the point estimate, or with a verbal caveat. People did perceive more uncertainty when it was shown, which is the point. But trust in the number and in its source dropped only slightly, and mostly when the uncertainty was expressed in words rather than as a numeric range. In one experiment, a numeric range and a "+/−" format raised perceived uncertainty by about the same amount (d = 0.50 each), while an explicit verbal statement raised it more (d = 0.68).
These experiments support displaying numeric uncertainty. They did not test moving ranges behind a tap or removing approximation signs. The same group's broader review cautions that effects vary between individuals and formats, and that the literature on communicating this kind of uncertainty is still limited2. So the claim is modest: showing a numeric range next to a number does not appear to cost much trust, and wording it vaguely costs more.
Uncertainty improved decisions in weather experiments#
A range is not only a disclaimer. In a set of experiments on weather forecasts, participants playing a road-maintenance manager decided whether to pay to salt roads against possible ice3. Those given uncertainty estimates alongside the forecast made better decisions overall: they took precautions when they were warranted and held back when they were not, more often than those given a single number. When forecasts became less accurate, people with only single-number forecasts became reluctant to act, and uncertainty information blunted that effect. Uncertainty estimates also increased trust in the forecast, and combining them with simple advice worked best of all.
The same idea could help when interpreting a food log. With illustrative figures: if today reads 1,650 of 1,900, the headline leaves 250 kcal. If the displayed total spans 1,500 to 1,800 kcal, the gap to 1,900 spans 100 to 400 kcal, and whether a dessert fits depends on where in that spread your day really sits. That is why the range sits one tap from the number rather than being removed. A broad review of risk communication ends on advice that fits here: acknowledge expert guidance, be clear about objectives, and work closely with the intended audience4. For a food log the objective changes with the moment. On the main screen, it is the likely value. When you are deciding something close, it is the spread.
A range here is a plausible spread for the estimate, not a statistical confidence interval, and it does not make the log exact. Its job is to keep an unweighed plate from looking as certain as a weighed one. How large everyday counting error actually is, and why it does not simply add up meal by meal, is covered in the realistic margin of error for calorie counting. And a range can be wrong too: a meal you misdescribed can fall outside it.
FAQ#
Why did the approximately sign disappear from my calories?#
It appeared on every number, so it no longer told you which estimates were loose. Since September 24, 2026 the headline shows the likely value, and a ±N chip beside it summarizes how wide its range is.
Where can I see the calorie range for a meal?#
Tap the ±N chip on the log receipt, or open the meal's details, where the range line and ±% chip are shown.
Is the ± number a confidence interval?#
No. The chip summarizes the width of the estimate's plausible range; tap it to see the actual low and high estimates. It is not a formal statistical interval.
Does showing ranges make people trust the numbers less?#
Only slightly, in large experiments. A numeric range beside a number cost little trust, while vague verbal caveats cost more.
Sources#
- van der Bles AM, van der Linden S, Freeman ALJ, Spiegelhalter DJ. The effects of communicating uncertainty on public trust in facts and numbers. Proc Natl Acad Sci U S A. 2020;117(14):7672-7683.
- van der Bles AM, van der Linden S, Freeman ALJ, et al. Communicating uncertainty about facts, numbers and science. R Soc Open Sci. 2019;6(5):181870.
- Joslyn SL, LeClerc JE. Uncertainty forecasts improve weather-related decisions and attenuate the effects of forecast error. J Exp Psychol Appl. 2012;18(1):126-140.
- Spiegelhalter D. Risk and uncertainty communication. Annu Rev Stat Appl. 2017.
Source: BurnWeek — "Where the calorie ranges live in BurnWeek now", https://burnweek.fit/blog/plain-numbers-ranges-on-tap/. Licensed CC BY 4.0: free to quote or reuse with a link to this page.
