I tracked it several different ways (with all sorts of nifty graphiness). Over time, I tracked spending and savings. I used spending to determine the approximate amount I needed and savings to determine how fast I was headed there. So... I plotted "time to retirement"... computed several ways:

* using data from the last 365 days, ugly brute force calculation where

$goal = $expense365d / 0.04; # 4% swr

$savings = ($income365d - $expense365d) / 12; # savings per month

$months = <very ugly brute force calculation where I count months, add in average savings and interest>

* same as above, but using 5year data

* same as above, but using 30 day data

* same as above, but using 90 day data

* automated run of firecalc using data from the last 365 days... to determine when retirement would be possible

(

http://www.firecalc.com/)

* same as above with data from last 5 years

Here is an example output graph... This is 5 year data, so my brute force ugly method and the firecalc method with 5 year data. The big spike 2011-2012 was me thinking I was smart and "optimizing" the calculations to factor out the costs associated with us building a house. I thought I accounted for it in a special way... and ended up factoring in the costs twice. The awful thing was it took me a year to notice I had done it.