Friday, March 16, 2012

Don't play Planning Poker with the Gorilla


Note: For new Gorilla Blogs, head over to The Gorilla Coach
"10 Days" Eric said.

I sighed. "Eric, you're supposed to use the planning cards and put down a value."

As he flipped through his stack of cards I could see him counting to himself. "Okay, fine. I think this is an 89 point task"

Sigh… "Eric, you just counted out ten cards and put that card down."

"Yeah, so?"

I leaned forward and tried to use my calmest voice. "Eric, we're not trying to figure out how many days right now, we're trying to do an abstract estimate on the relative size of each story in the whole project. "

He waved his hand, "What's the point? If we're painting a house and you ask me to estimate how long it will take to paint the garage door I can put down a 2 or just tell you two hours. You're asking me to use an abstract number when I already know how long it will take."

I sat back in my chair perplexed. The problem was that I could see his point. We were tackling each task one at a time and no matter what I did, I couldn't break the team out of thinking about how many days something would take. And I knew this was a slippery slope that led back to padding, swags and a whole slew of inefficient planning practices.   

A long meeting later I dragged myself into my office. Only to find both my chair and the guest chair occupied. One of the occupants I expected to be there, a post meeting debrief on this last meeting was much to much for my personal gorilla to miss. Seeing the black swan flipping playing cards onto my desk was quite another thing.

Flip, flip, flip "144!" Hogarth said, triumphantly.  Looking at the black feathered swan he grinned. "Hah, that beats your 21."

"Hogarth!?" You'd think I would get tired of yelling his name. Or maybe I would just get used to his ever present level of restrained chaos. You'd think... "What are you doing with my Planning Poker decks?"

"Oh, hey there," Hogarth turned his cheerful smile my way. "Wanda and I were just playing war."

Tossing her last card, a 2, down Wanda slid off the chair. "I have to get back to the server room anyway, one of the HVAC's is about to short out because a gopher climbed on the roof and ate through the power cable."

Trying not to think of electrocuted gophers and overheating servers, I ignored the departing black swan and turned my attention back on Hogarth. "Why are you playing war with my Fibonacci cards?"

"Oh, that. Well I figured you might want them and then we just kind of got bored waiting for you to be done not estimating the release."

I blinked at him. "Hogarth, I know you were eavesdropping on the meeting. You know perfectly well the team thinks planning poker is useless." I tossed myself into the vacated guest chair with a sigh. "And I'm not sure I don't disagree with them."

"Since when have you believed there's only been one way to use a tool?"

Blink, blink… Why did he always have to be right?


Story Point Estimating
This is one of the foundational principles of Scrum software development (and Agile  in general). Done well, story estimating can create an incredible level of predictability and transparency. A high performing team can look at a set of requirements and provide a high confidence assessment of when they will done.

But this isn't about the value or use of estimating. There is plenty written on this and I'm not here to rehash it. I believe in estimating, just as I believe in the principles of agile and good management.

What I'm hear to say is I think Planning Poker  is the wrong way….

What? Come on, I'd be a lousy gorilla talker if I wasn't willing to tackle tough subjects head on. I do absolutely think story point estimating is the right thing. I just think planning poker is not the best way to do it. 

Naturally this begs the question of "why?" I have two reason for my concerns on the use of planning  poker:
  1. Each user story (feature, tasks, item) is evaluated on its own. When you do planning poker you are just looking at how long it will take to paint the garage door. You are hard pressed to look at it in relation to putting a new latch on the back gate. I think the mind is hard pressed not to assign a physical value to the estimate. "I think this will take two hours, that's about a 3."
  2. At the end of the day, the planning poker process occurs in an individuals head.  Yes, there is a discussion among the team, but only after each person has drawn a line in the sand. Once the human mind draws a line, it finds it hard to move that line. We've invested ourselves in that line.

So what the heck do you do?

Well if you've been following my recent agile related blogs, you'll know I'm a very big fan of something called the "Team Estimation Game" developed by Steve Bockman. I personally like to call it "Team Planning Solitaire." The style of the game reminds me of classic Klondike solitaire and the interaction of the team makes it anything but a solo activity (what can I say, I like irony). I will also note that while I learned this from Steve Bockman and Agile Learning Labs , I have seen some similar exercises in the last few months. Whether they are parallel development or evolutions of Bockman's game, I don't know.

I detailed how this game worked in "How much is that Gorilla in the window." At the highest level, the
Process starts with the story cards in a pile. Each person takes a turn laying out a new card or moving an existing card on the table. At the end, you have a line of cards in rank order. Only then do you break out the planning poker cards. Your estimates end up being based on the relation to the other stories, not to how long one story will take.

Why is the Team Estimating Game so good?
  1. Stories are estimated in relation to one another and not in a vacuum. "Is painting the garage door more or less effort than replacing the latch on the back gate? All right, is painting more or less effort than rehanging the front door."
  2. It's a team activity. Nothing happens in anyone's head. It is all out on the table and very straight forward. You aren't sitting there arguing that the Database re-architecture will take a week or two weeks. You are just trying to decide if it is more or less effort that localizing the user interface to Japanese.

At the end of the day I think Team Estimating is much more effective than planning poker. Leave the poker cards for a nice game of Klondike or Texas Hold 'Em.

Who's ante is it?

Joel Bancroft-Connors
The Gorilla Talker
Want me to talk to your gorilla? mailto:jbancroftconnors@gmail.com
You can follow me on twitter, @JBC_PMP