Here's a time series from one of the advocates of #Noestimates showing the number of stickies processed on a weekly basis. This chart is used to conjecture that decisions can be made using emperical data from the past. Trouble starts when there is no mention of the wide variance of this data and the credibility of the forecast of the future made from this data.
When you place that time series data in an array Numbers and point R at it with these commands
NETS=ts(NE.Numbers) - convert the raw numbers to a time series
NETSARIMA=arima(NETS, c=order(0,1,1)) - make an ARIMA object
NEFORECAST = forecast(NETSARIMA) - make a forecast using that
plot(NEFORECAST) - plot it
You get a nice plot from R-Studio's tool...
Yes, R and its ARIMA function says FORECAST, and yes a FORECAST is an ESTIMATE of the possible range of values in the future using data from the past.
But here's the Showstopper
All project work operates in the presence of uncertainty. This uncertainty creates the random outcomes as shown in the first diagram, that are used in the second diagram to development the Autoregressive Integrated Moving Average, that can then be used to reveal something about the future.
In this example, You may want to reconsider making decisions with other peoples money when the 80% confidence in a possible outcome is itself a - 56% to +68% variance? really. Flipping coins gets a better probability of an outcome inside all the possible outcomes that happened in the past. The time series is essentially a random series with very low confidence of being anywhere near the mean.
And the Final Question
If the advocates of #Noestimates are producing data that clearly shows there are random processes in the software projects they work, and these random processes are created by the underlying uncertainties of the project work (both reducible and irreducible uncertainties), then why are they stating that estimates are waste and should not be used when it is crystal clear these random processes and the resulting data cannot be removed?
It seems to me those #NoEstimates advocates have willfully ignored their own data, showing the underlying random processes, while willfully ignoring the fundamental process of making decisions in the presence of the uncertainties that create these random outcomes require estimates.
