Story / Cycle Time correlation

Are we any good at estimating? 

Most teams estimate the requirements they work on.. and many believe that it's a good indicator of the time it takes to do the work.. but is it?

Teams use this visual in retrospectives to identify outliers and reflect on how to improve their estimating. 

Alternatively, they may decide, like quite a few teams, that it's not worth recording the actual estimates, even if the act of estimating is in itself valuable. Whilst other teams have done away with estimating all together. 

 

 

Do estimates predict how long a piece of work will take? 

The chart shows size (1) along the X-axis and lead/cycle time in days (2) on the Y-axis.

Users select the states they want to include in the cycle time calculation (3) 

The correlation coefficient is shown (4) - the higher the number, towards 1.0 the higher the correlation. 0.37 is one of the stronger correlations that we have seen. Which indicates how poor most projects are at estimating. !

Bugs and stories are shown in different colours - with a count of the number of unestimated stories (5) that have NOT been included in the calculations.