What will the decline in steps taken become?

Question

\begin{array}{c|c} Date & Value \\ \hline 7/11/2013 & 880 \\ 10/25/2013 & 810 \\ 12/23/2013 & 392 \\ 4/9/2014 & 689 \\ 7/23/2014 & 720 \\ 4/23/2015 & 420 \\ 7/23/2015 & 480 \\ 10/22/2015 & 434 \\ 2/3/2016 & 476 \\ 3/3/2016 & 236 \\ 3/31/2016 & 320 \\ 7/5/2016 & 342 \\ 1/23/2017 & 152 \\ 6/1/2017 & 161 \\ \end{array} Is there a way to predict what the next step counts will be for the rest of the year, every 3 months?

Answer 1

Normalizing the date scale to days since 2013-07-01, and constructing the best linear fit you end up with the equation $$v = 777.93 - 0.4448d$$ with $v$ denoting value and $d$ -- offset in days. The quality of the fit is a decent $R^2 = 76.21\%$, here is the graph from Excel:

Answer 2

I would re-cast your time data as relative to the start (in your case, 7/11/2013), in days. Then plot the data up in Excel in an $xy$ scatterplot, and try to fit curves to it. Make sure you display the equation and the $R^2$ value. If the $R^2$ value is very close to $1$, you have a good fit. Then you can use the fitted curve to predict future values.