SunPlanner calibration starts after three usable production readings. It compares what the app predicted with what your panels actually produced, then uses that pattern to adjust future estimates. It does not turn on after one reading, and it does not promise one accuracy percentage for every setup.
What does calibration compare?
Calibration compares two numbers for the same day. One is the uncalibrated prediction saved by SunPlanner. The other is the actual solar production you enter at the end of the day.
The difference between those numbers becomes part of the calibration history. SunPlanner uses that history to adjust later forecasts. It is learning the size and direction of the consistent gap. It is not claiming to know which wire, panel condition, or repeated system loss caused it.
Why does calibration wait for three readings?
One day is not enough to show a pattern. A weather miss, an unusual load, incomplete production data, or a simple entry mistake could make that day look different from normal.
SunPlanner keeps the calibration factor at its starting value until there are three usable readings. After the third usable reading, the factor can begin affecting the forecast. Logging more usable days gives the app more history to work with.
What counts as a usable reading?
A usable reading needs a positive prediction and a positive actual-production value for the same day. A day with zero or missing production can remain in your history, but it does not change the calibration factor or move calibration progress forward.
That boundary matters. It keeps a blank entry or a day with no comparable prediction from being treated as evidence about how your panels perform.
Can one strange day take over the forecast?
SunPlanner limits how much one unusual reading can influence the calibration factor. Newer usable readings receive more weight, while ratios that sit far outside the rest of the history are limited before the factor is calculated.
The app can also give more weight to readings from a similar time of year when there is enough useful history. If that seasonal history is too thin, it falls back to the broader calibration history.
Does battery size change calibration?
Battery capacity does not change predicted solar production. A larger battery can store more energy, but it does not make the panels produce more.
Calibration stays focused on the relationship between predicted solar production and the actual solar production you recorded. That keeps the forecast separate from battery sizing and daily load planning.
What does the 25-reading milestone mean?
The app tracks progress toward 25 usable readings as a fuller-history milestone. Calibration still begins after three usable readings and keeps updating as you add more.
Twenty-five readings are not a universal accuracy guarantee. Weather, seasonal coverage, data quality, and the setup itself still matter. The useful signal is whether the calibration history stays consistent as new readings arrive.
What should I do next?
Log complete end-of-day solar production when you have a matching SunPlanner prediction. Keep logging across different weather and times of year. The app will handle the comparison and update the factor when the readings qualify.
You do not need to diagnose every small loss before calibration can help. You need a clean comparison between what SunPlanner predicted and what your panels produced.