Echo team estimation experiment

[lowellrex]

Estimation is an imprecise art and attempting to arrive at a single level of estimated effort elides the uncertainty the team foresaw during the estimation meeting discussion. So the Echo team has decided to embrace the uncertainty by incorporating it into the estimation process. Each developer on the team is given 3 votes to distribute across the various point estimates as they see fit (for example, if I think a ticket is probably a 3 level of effort that may balloon into a 5 I would cast 2 votes for 3 level of effort and 1 vote for 5). After votes are cast the level of effort with the most votes becomes the official estimate and we include brief explanations of why completing the ticket may not end up requiring the estimated level of effort.

Example from #11509

1 | ||
2 | ||||
3 | |||||
5 | |
8 | 

Estimating at 3. Could be 1 because we are simply grouping by a single column in a single table most of the time. Testing could add a significant amount of effort here. The fact that we have to do this for each column means it is probably not totally trivial to do. 5 as a hedge in case there are performance issues that arise that we need to address.

Example from #11442

1 | ||||
2 | ||||||||
3 |
5 |
8 | 

Estimating at 2. Code changes should be straightforward, but manually updating of live data brings this to a 2 level of effort. May be a 1 if the manual live data change is easy.