Guest post: Planning Poker – The Power of Two

To plan an Agile project most teams play Planning Poker to estimate the size of user stories. Tools are available like a deck of cards, paper or nowadays even Apple/Android apps. Simplicity is one of our Agile principles and The Power of Two practice allows you to play Planning Poker without any need for these tools but just using one hand. I’ll share with you the basics and throw in some adaptations that happened when I used The Power of Two with my teams.

The Basics
The basics for The Power of Two are straightforward and all you need are team members with five fingers on one hand. The number of points for the size of a user story represent the number of fingers shown by a team member using the following sequence with each successor doubling in size:

• No fingers by showing your fist => 0 points
• One finger => 1 points
• Two fingers => 2 points
• Three fingers => 4 points
• Four fingers => 8 points
• Five fingers or full hand (two to the power of five) – off the charts!
• Flat hand – pass

Finally, except for showing estimates by using number of fingers according to this sequence, the Power of Two just complies with the rules for playing Planning Poker.

The Power of Two really stands out for me because of it’s proven simplicity in multiple ways:

• The use of only one hand for showing estimates which makes it fast and easy;
• The estimates covering a wider range (1-8) than the same Fibonacci sequence (1-5) which creates clearer separation of estimates;
• The team soon taking over control from the facilitator for when a new estimate is due. Actually it feels like the team gets in a mode simular to when you’re playing Rock-Paper-Scissors (RPS);
• The flexibility to play Planning Poker anywhere, anytime with anyone because you don’t require any tools.

When playing Planning Poker using The Power of Two at least consider the following lessons learned:

• When playing The Power of Two for the first time with a team write down the sequence on a whiteboard for easy reference and to avoid mistakes;
• Initially relate one story point as one ideal developer day because this conveniently translates to user story sizes that fit within one iteration;
• At least try the Power of Two sequence, instead of the Fibonacci, because I never had the need, or a request from a team member, to convert back to a Fibonacci sequence.

Adaptations
I have used the Power of Two with my teams by always starting with the basics as just described. During Planning Poker I have often noticed the following adaptations from by my team members:

• Using half an index finger to indicate a small story of 0.5 point;
• Taking the average for the higher end estimates as consensus. For example if we only have estimates of 4 and 8 points remaining, go for 6 points and avoid wasting time;
• Set aside the larger user stories initially (five fingers or full hand) because they normally won’t fit in one iteration. They need to be decomposed and re-estimated or re-estimated as an epic using a higher sequence of points that’s consistent with the lower sequence. For example a higher sequence with 20 points for a user story is still ten times as big as the 2 points from the lower sequence. And don’t forget: for the higher sequence the Power of Two can be used again (for example: 10-20-40-80)!

Finally, enjoy using The Power of Two when playing Planning Poker and let us know your experiences!

For information about Planning Poker: http://en.wikipedia.org/wiki/Planning_poker