This is part three in a series on estimating. Part one was "Don't Estimate Software…
I had a fun experience last week that I want to share with you guys. It was my first week with an awesome new client. The plan was to start with two days of general agile training, followed by a day of release planning workshops, and then iteration planning. Once the team starts sprinting, I’ll come back for a day a week to do a little coaching, and to help keep everything on track. All in all, a pretty standard engagement.
This team has a solid product vision, and a pretty good understanding of their features and needs. The features and needs were pretty easily translated into high-level, value driven epics. We spent some time breaking down the first set of features into finer grained user stories. Because the team was new to agile, and really had no idea of their capacity, we took the extra time to break their user stories down into tasks.
Okay, so you guys know the drill. User stories get estimated in points, tasks get estimated in hours. We decided to use the Fibonacci series for the user stories, but the team wanted to use a binary sequence for estimating task hours. I’m okay with either… and since it’s really not my call… we went with binary. The only problem was that we didn’t have enough binary planning poker cards so that the entire team could get involved. We had to get creative.
The ideas came quick, but what we ended up with was a version of the old rock-paper-scissors game. Someone realized that each of the allowable estimate values, could be represented as a single digit, using a power of two…
1 = 0
2 = 1
4 = 2
8 = 3
16 = 4
32 = 5
…and, we could each vote with one hand by throwing the exponent, rather than the entire estimate value. Only, in a room full of software developers would someone come up with that. It was elegant… it was geeky… we could do it without supplies… so we went for it.
Since there was no way to hide your estimate using fingers (say, by putting the card face down on the table) the developer leading our session introduced the rock-paper-scissors idea. But… rather than saying rock-paper-scissors-shoot… he went with Ro-Sham-Bo-shoot. I had never heard that version, but once again… it was funny… it was geeky… and the team loved it… so it stuck. Each time, after they talked about a task, someone would look around and ask… ‘rochambeau’ ?
Once everyone agreed… they played their hand… literally.
It was an awesome… fun… emergent experience, and we came up with a new approach that I am sure I’ll want to use with other teams. The estimating went really fast, faster I think than messing around with physical cards. Either way, it was a great team building experience. It was a creative and effective way of playing planning poker. I’m curious if anyone has every played planning poker this way? Is it documented somewhere?
If you happen to be interested… here is a link to the Wikipedia entry for Rochambeau.
The Urban Dictionary had a much more colorful explanation… this one made me laugh out loud.
Give this approach a try and have some fun with it!