Balance Point treats each weigh-in as a noisy reading of a slower-moving weight. A statistical filter tracks that weight and its weekly rate and carries both across gaps with a widening range. A forecast appears only when enough recent data supports one.
From weigh-ins to a trend
The number on the scale mixes two things: the weight you’re actually carrying, and whatever water and food happen to be on board that morning. Balance Point keeps a running estimate of the first, along with how fast it’s changing, and updates both every time you weigh in.
Each new reading nudges the estimate in proportion to how surprising it is. When the trend is well established, a reading two pounds high moves it only a little; when you’re starting out, or coming back after a break, readings carry more weight because the estimate has less to go on. Looking back over history, the app also uses readings from after each day, which settles the line into its most likely shape.
The filter was built for irregular records. Weighing every morning keeps the range tight, and a sparser record still produces a trend; its range stays wider.
The range around the trend
The shaded range shows how sure the estimate is. Frequent weigh-ins pull it tight. During a gap it widens a little each day, because the weight could have drifted in either direction, and it narrows again once readings return. The weekly rate carries the same kind of range, which is why Today shows something like −0.5 lb/wk ±0.5.
Trend
199.9 lb
Rate
−0.5 lb/wk±0.3
Forecast
195 to 201 lb
Weigh-in
Trend
Range
Plan
Goal
Forecast
Sample data
The plan is a curve you choose
A goal in Balance Point is a campaign: a start and a goal weight, with a shape between them. You can set it by date, and Balance Point works out the pace, or set the pace and let it find the date. The usual shape tapers like real weight change does, starting faster and easing off near the goal; linear, percent-of-bodyweight, and open-ended plans are there too.
Pace is measured as the first week’s change as a share of bodyweight, which is the steepest point of a tapering plan. For someone starting at 240 lb aiming for 175 lb:
Pace
First week, share of bodyweight
First week at 240 lb
Time to 175 lb
Gentle
0.75%
1.8 lb
about 33 months
Steady
1.25%
3.0 lb
about 20 months
Ambitious
1.75%
4.2 lb
about 14 months
Above 1.5% a week, the app says plainly that the plan front-loads the work. Above 2%, it declines the date and offers the nearest one a body can realistically meet. These checks describe the curve; they say nothing about the person drawing it.
The forecast reads only your weigh-ins
The forecast ignores the plan entirely. Once a campaign has at least 112 days of history and 20 weigh-ins, and the trend has visibly started to curve, Balance Point fits a curve whose leveling-off point is free to land anywhere. That freedom matters: a straight line always reaches the goal eventually, while a fitted curve can show that the current trajectory levels off above it.
With less data, the app uses the current rate and its range instead. When recent readings are too sparse for either, the forecast says “Not enough recent data for a date” and draws nothing.
Plan and forecast, side by side
The two lines answer different questions. The plan is when you said you’d arrive; the forecast is when the data says you will at the current pace. Both appear on the chart with their own labels, and the gap between trend and plan reads as “N lb above plan” or “below plan”.
What Balance Point never stores
Trends and forecasts, ranges included, are calculated each time you look and are never saved as weigh-ins. The only values that reach Apple Health are ones you type. When you export your history, the file holds your real readings, so a gap in the record stays a gap.
Very little. A single reading shifts the trend according to how far it sits from the line and how settled the trend already is, so one high morning after a salty dinner moves it by a fraction of the jump.
Why can the trend for last month change after I weigh in today?
+
Looking back, Balance Point uses readings from both sides of each day, so later weigh-ins sharpen the estimate for earlier ones. The value it showed at the time stays a fair estimate of what was known then.
What does the ± next to the weekly rate mean?
+
It’s the range on the rate. A reading of −0.5 lb/wk ±0.5 means the data fits anything from about −1 to 0 lb a week, so the trend is falling or flat, and the speed is still uncertain.
Balance Point is in development
It’s coming to iPhone and iPad. Leave your email on the home page and Rival Bear will write when it’s on the App Store.
