Extrapolation Calculator
Enter your known points and the x you want to predict. The calculator draws a straight line through your data and extends it.
Enter your known data, one point per line as x, y (at least 2 points), then the x you want to predict.
y(800) = 85.2
Line: y = −0.018x + 99.6 (slope −0.018). 800 is 300 past your last x, which is 100% of your data range beyond it. That is a long way out, so treat the result as a rough guide.
Linear extrapolation formula: y = y1 + (x − x1) × (y2 − y1) / (x2 − x1). It is the same formula as linear interpolation, used with an x that lies outside the two known points. With more than two points, the calculator uses the least squares best fit line instead.
How to use the extrapolation calculator
- Enter each known point on its own line as x, y. Two points give an exact line through them.
- With three or more points, choose “Best fit through all points” to smooth out noise, or “Through the last two points” to follow the most recent trend.
- Enter the x you want to predict. You get y, the line equation, the slope and, for three or more points, R² (how well a straight line fits).
- Read the distance note. The further past your data you go, the less you should trust a straight line.
Worked example: battery capacity after more charge cycles
A phone battery holds 96.0% of its original capacity after 200 charge cycles and 90.6% after 500. Estimate its capacity after 800 cycles.
- Slope = (90.6 − 96.0) / (500 − 200) = −0.018 percentage points per cycle.
- y(800) = 90.6 + (800 − 500) × (−0.018) = 85.2%.
- The line is y = −0.018x + 99.6. 800 is 300 cycles past the last point, the same as the whole data range, so this is a rough guide, not a promise.
With more readings, use them all. For 100, 200, 300, 400 and 500 cycles at 97.9, 96.0, 94.4, 92.3 and 90.6%, the best fit line is y = −0.0183x + 99.73 with R² = 0.9987, and it predicts 85.09% at 800 cycles. Using only the last two readings (92.3 and 90.6) gives 85.5%. The two answers are close because the data is nearly straight.
| Method | Prediction at 800 cycles |
|---|---|
| Line through 200 and 500 cycles | 85.2% |
| Best fit through 5 readings | 85.09% |
| Line through the last two readings | 85.5% |
Real batteries often fade faster later in life, which a straight line cannot see. Our linear extrapolation guide shows a growth example where a straight line fails, and when extrapolation is reasonable.
Extrapolation Calculator: frequently asked questions
What is the difference between interpolation and extrapolation?
Interpolation estimates a value between known points. Extrapolation estimates a value outside them. The formula is the same, but extrapolation is much less reliable because nothing in your data checks the result.
How do you extrapolate a value from two points?
Find the slope (y2 − y1) / (x2 − x1), then add slope × (x − x2) to y2. For example, from (200, 96.0) and (500, 90.6), the value at 800 is 90.6 + 300 × (−0.018) = 85.2.
How far can you safely extrapolate?
There is no fixed limit, but a common rule is to stay within about half of your data range beyond the last point, and only when you have a reason to expect the trend to stay straight. The calculator warns you past that point.
Can I extrapolate in Excel?
Yes. FORECAST.LINEAR(x, known_ys, known_xs) gives the same best fit result as this calculator, and TREND does the same for several x values at once.
Should I use a curve instead of a straight line?
Only if you know the process is curved, such as compound growth or decay. Fitting a high degree polynomial to extrapolate is risky because it can turn sharply just past your data.