Player DM'ed retrospective summary

[jeffreywescott]

@shereefb commented on Mon Aug 01 2016

After every player on a project has completed the retrospective, the player should receive a DM in echo with the following information:

---

Depends on RocketChat/Rocket.Chat#358
Depends on RocketChat/Rocket.Chat#365

---

Retrospective results for #happy-tiger:

Feedback from your team:

• Jeffrey, I like working with you because you shower every day and I'd like working with you even more if you chewed with your mouth closed.
• Jeffrey, I like working with you because you have an interesting way of deconstructing problems, and I'd like working with you even more if you took more risks.
• ...

Stats earned from this project:

name	score
learning support	70%
culture support	84%

Contribution to the project:

Your contribution to the project: 42%
Team size: 3
Expected contribution for team size: 33.33%
Relative contribution: +8.67%

averaged	self-assessed	team-assessed
42%	40%	45%

Hours contributed to this project:

player	num hours
@deadlyicon (Jared Grippe)	5
@jeffreywescott (Jeffrey Wescott)	42
@tannerwelsh (Tanner Welsh)	34

[jeffreywescott]

Assigning @prattsj ...

[tannerwelsh]

Updated description with a few changes; good enough for now. Thanks @jeffreywescott

[shereefb]

@prattsj the expected contribution should be based on total hours divided by players, not just total players. For example if shereef claimed 50 hours, and jeffrey claimed 30, and sj claimed 20, then shereef's expected contribution is 50%

[heyheyjp]

@shereefb: got it - thanks.

[shereefb]

side note, I'm glad someone finally spoke up about jeffrey's chewing habits. #truthovercomfort

[heyheyjp]

@jeffreywescott, @tannerwelsh, @shereefb: I'm really, uncomfortably confused about the Relative Contribution stat and could use some clarification on a number of points:

When looking at the stats spreadsheet, I was surprised to find that:

• the absolute RC values were so small
• some of them were negative

...because in the Statistics section of the design guide, RC is described as the following:

The percentage of work contributed by a player to a goal.

Determined in the retrospective process by calculating the average of each player on a team's estimation of contribution percentages.

Formula: sum of all contribution estimates for player / team size

For example, if on a 3-player team Player A rates their own contribution at 20%, and the other two players rate Player A's contribution at 40% and 45%, respectively, Player A's RC is 35%.

This clearly isn't what's represented by the figures in the RC columns in the spreadsheet:

So my first couple of questions are:

• Are these really - as they seem to be - two different things?
• If so, why are we calling them both "relative contribution"?

Regardless of what the answers to those questions are...I had to go over this section in the spec here a few times to feel like I really got what it's saying. I don't know if confusion around RC was specifically something that came up as a strong tension with the learners Friday, but my gut sense is that they're gonna (continue to) be confused, too:

Contribution to the project:

averaged	self-assessed	team-assessed
42%	40%	45%

Relative contribution: +8.67%
(Higher than expected contribution of 33.33%)

What can we do to make this crystal clear?

[heyheyjp]

@jeffreywescott: Rocket.Chat doesn't support all of the Markdown features we get in GitHub-flavored markdown, including fancy table rendering. So, we can't easily deliver on something that looks like:

name	score
learning support	70%
culture support	84%

We'll need to come up with an alternative way to style the stats in the DMs.

Here's one suggestion, falling back to something similar to what we spec'd in #111:

Learning Support   [ 70%]
Culture Support    [ 84%]

Open Rocket.Chat issue:
RocketChat/feature-requests#642

[tannerwelsh]

I'm really, uncomfortably confused about the Relative Contribution stat and could use some clarification on a number of points:

Thanks for bringing this up @prattsj - it's a semantic inconsistency that's been plaguing us for weeks. I've been trying to clarify but have not succeeded in eliminating all inconsistencies.

Here's how I re-worded the stat just now:

Your contribution to the project: 42%
Team size: 3
Expected contribution for team size: 33.33%
Relative contribution: +8.67%

The two numbers we're working with are contribution (which I think we should call absolute contribution) and relative contribution. Contribution is just the % that a player contributes to a project (average of all team's inputs, including the player themselves). Then there is this hidden constant number - the expected contribution. This is how much a player is expected to contribute given the size of the team.

Relative contribution is the difference between contribution and expected contribution, which is why it can be either positive or negative.

[heyheyjp]

@tannerwelsh: thanks - this is an improvement. Can I poke at this a little bit more? Now that it's clear that what you're calling relative contribution is the delta between what a player's estimated contribution is and what their expected contribution is...I'd like to say that it feels like "relative contribution" might not be the best name for this value. It is not immediately obvious what this value is relative to; I can only infer that by looking at it in comparison to other numbers.

Second, I wanted to make a couple of small suggestions for formatting this section:

• Put team size at the top/group the "relative contribution" items
• If we include team size, we should also include total hours since that will play into the figures, too
• Be explicit about "Your contribution" being your estimated and/or averaged contribution

Team size: 3
Team hours: N
Your hours: M

Your estimated contribution to the project: 42%
Your expected contribution: 33.33%
Relative contribution: +8.67%

[shereefb]

Aha! Now I see the confusion. I've been using relative contribution to mean a players contribution relative to another players contribution (the definition that @prattsj) pasted. Not relative to what a player "should" have contributed.

@tannerwelsh I agree with @prattsj, what you're calling relative contribution should be called something else. Contribution delta / expectation delta or something.

[jeffreywescott]

We used to say RC and RCD.