We all know what poker is. A game of cards where a player uses secrecy and deception in an attempt to turn a game of luck into a game of skill. In this game, players are pitted against each other to see who can get the best hand of cards with the winner generally taking home a nice cash prize while the rest of the table walks away with less than they had when they sat down. Though I have never been one to put money on the line, I have played my fair share of poker through the years wagering things like M&M’s or peppermints, and I have learned to appreciate the mixture of skill and luck that it takes to walk away victorious. 

Planning Poker is an Agile tool for teams that has just enough in common with the traditional card game that it was bestowed the same name. For starters, Planning Poker is played with a deck of cards, secondly it relies on secrecy from each player. That’s where the similarities however stop. Whereas traditional poker is mono a mono, or each man against the other, Planning Poker is a team sport, with each member bringing valuable insight and perspective to the mix that the other members may not possess. Below I’m going to explain how Planning Poker works, it’s advantages over traditional estimating techniquest, and variations to help you adapt the game to your environment. Let’s dive in!

How Planning Poker Works

There are six steps to a round of planning poker. A game is made up of any number of rounds. Generally, enough rounds to estimate all the work in the next sprint. The steps are:

  1. The Presentation
  2. The Estimation
  3. The Reveal
  4. The Discussion
  5. The Second Round
  6. The Truce

The Presentation

In the presentation stage, the facilitator (often referred to as the Scrum Master) presents a User Story which has been prepared ahead of time. Note that depending on how technical the project is, these User Stories are often created in conjunction with the team in a backlog refinement meeting (also known as backlog grooming). After the User Story is read outloud, the facilitator spends a few minutes elaborating on any critical nuances and the team has a chance to ask questions for clarification. The goal is to keep this discussion short while still giving the team time to gain the clarity they need before estimating its complexity. 

The Estimation

When the team is ready, the facilitator asks for everyone to select a card and place it face down on the table. When forced to place an estimation on a User Story, there are often additional questions that arise. Clarification at this point should be encouraged as long as it does not drag on too much. As long as the cards have not been revealed, team members are able to pull their card and replace it based on the clarification.

The Reveal

When everyone on the team has a card on the table, the facilitator asks for confirmation that everyone is ready, then tells them to flip their cards in unison (often on the count of three). When all cards are revealed the facilitator verbally identifies the number with the most votes, or in the case that there is no clear majority, the middle number.

The Discussion

After all cards have been revealed and the majority vote has been defined, the facilitator asks every player who’s card does not match the majority number to explain the logic behind their estimation one at a time. This can be uncomfortable for some members, but it must be emphasized that this discussion phase is where the magic of Planning Poker takes place. With a team full of professionals from different backgrounds each player’s estimation is based on their experience. A skilled facilitator can coax out the logic behind a player’s estimate so that the team can gain the benefit of everyone’s collective experience. This gives everyone on the team a voice and leads to better estimation. This is how you keep from letting the Type-A leaders of the team dominate the conversation, and instead help surface the valuable experience of the more analytical quiet types. This also helps junior members of the team learn to think like their more experienced counterparts over time. 

The Second Round

Once the discussion has teased out any unknown clarification or relevant experiences, the team puts their cards back in their hands and again selects their estimate to place face down on the table. When everyone is ready the facilitator again tells them to flip their cards in unison. Though this can occur a third or fourth time if needed, two rounds is generally all that’s required. 

The Truce

Once the second round is complete, the team must settle on a singular final estimate that everyone can agree on. This is often easy, but can sometimes require negotiations and futher discussion. The important thing to note is that it is the team that settles on the final number, not the facilitator. 

The Advantages of Planning Poker

Team Building

One of the greatest benefits of planning poker is that it can be fun. Teams often look for icebreaker games, or team building events to build team camaraderie and morale. How great would it be to have a built-in team building activity every sprint?

Everyone Gets Heard

Whenever you get a team together, leaders emerge. Whether this is due to their seniority, or their outspoken nature, the leaders are often given credence to make the decisions and their vote seems to count for more in traditional estimation meetings. With Planning Poker this is not the case, everyone gets an equal voice and regardless of their seniority, their thought process gets teased out.

Skill Calibration

Along with the valuable opinions, Planning Poker teases out the inexperienced opinions and quickly spreads knowledge and experience across the team. This is why it is critical that everyone on a team participate in estimating all stories. Regardless of whether or not one member has the domain knowledge or is planning to work on that specific story. When you require everyone to participate, the game naturally starts to calibrate each player on all skills relevant to the team. Before long you will see that individuals from one discipline can accurately estimate the effort it takes for work in another discipline. This leads to better team collaboration and cross discipline empathy. 

Better Estimates

The traditional way of estimating work is to give a team lead a spreadsheet with all the tasks for a project and ask them to work with the team to get the estimates nailed down. What inevitably happens is that the leader will estimate everything and then ask the team for conformation. This technique relies on the experience of a single team member and does not help the less experienced members learn how to give accurate estimates. Another technique used by Agile teams is the makeshift Planning Poker exercise. In lue of having the right deck of cards or an online Planning Poker tool, teams will ask for everyone to shout out the number, or put it in a chat box in a video chat. Though this is functional, it still allows the less outspoken or junior members of a team to listen for the team leaders estimate and then default their answer to align. This undercuts the key benefits of Planning Poker by promoting the experience and opinions of a singular person or personality type inside the team. In contrast, following the approach Planning Poker as laid out in this article, leads to better collaboration and more accurate estimates.

Statistically Accurate Predictions

The goal of estimation is to provide an accurate forecast of time or effort for any given chunk of work. The problem is that as humans the planning fallacy seems to be unavoidable.

The planning fallacy is a phenomenon in which predictions about how much time will be needed to complete a future task display an optimism bias and underestimate the time needed. – Wikipedia

This is by far one of the greatest advantages of Agile Planning Poker. Instead of hours, Planning Poker usually uses Story Points, T-shirt sizes, or other comparative items that can be relatively compared to similar items. So instead of saying that one task will take 8 hours and another 16, we are able to say that one story or task is 3 points, or a small t-shirt while another 8 points, or a medium t-shirt. After a few iterations or sprints we are able to look back and create statistical models showing that 3 story points on average will take team A 10 hours whereas 8 story points will take Team A an average of 22 hours. By using statistical trends we are able to generate a much more accurate model for predicting the completion of future work while avoiding the planning fallacy phenomenon.

Variations

Depending on the use case and the organization constraints, there are a variety of measurements that can be used in a Planning Poker game to leverage the benefits without requiring a monumental shift in the units of measurement expected in the outcome. Here are some of the options:

Story Points – Fibonacci

The Fibonacci sequence, also known as the golden ratio, is the standard unit of measurement used in Agile. It is generated by adding the current number and previous number together to generate the next number in the sequence. These numbers are 1, 2, 3, 5, 8, 13, 21…etc. and are often appended with the numbers 100 to represent “Too large” and the sign of infinity to represent “Unestimatable”.

These numbers seem to work well because it eliminates a lot of the discussion around whether a User Story should be 8 or 9 story points. In reality it’s easy to say that a User Story is close to twice the size of another, while it is much more difficult to say that a story is one eighth larger than another. 

Story Points – 1-10

While fibonacci has its advantages, some teams prefer the specificity that sequential numbers provide, example: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10. For instance what should you do using the Fibonacci sequence if a User story is larger than another estimated as a 5 but smaller than another estimated at an 8? For highly analytical teams sequential story points may be more comfortable. The pitfall to avoid however is allowing teams to directly correlate sequential numbers to hours. For Story Points to be effective this mental model must be decoupled. 

T Shirt Sizes

T-shirt sizes are another great option when estimating work. This is often used at higher levels of abstraction such as the planning entire projects or inishitives, but can also be used to estimate User Stories and has the advantage of helping teams break free from correlating estimates with hours.

Hours

Though I believe that Story Points are a better estimation measurement because of their ability to be statistically accurate vs literally accurate, (Read more in our article on Story Points), many organizations are not prepared to make this transition and require upfront hour estimates for investment planning or resource allocation. In these cases it is completely possible to use Planning Poker to estimate actual hours while still leveraging the majority of benefits listed above. To facilitate this, using a sequentially numbered deck of cards can be used. 

Days

Though less popular than other techniques because of its lack of specificity, some teams estimated by days. Especially when planning out larger chunks of work, this estimation technique allows for a little more margin than hours, while still providing a specific time frame estimate out of the gate. The deck for this type of estimation is sequential and is often capped at 5, with larger chunks of work being broken down for better precision.

Confidence Votes

At the team level, rating by Story Points or T-shirt sizes makes sence, and gives a voice to every team member. In scaled versions of Agile, such as SAFe, Nexus, or LeSS, many teams are combined to accomplish collective goals for large organizations. In these frameworks there is the need to let every voice be heard at the Program or Organization level as well as the team level. This is often facilitated by a confidence vote, where all members, from all teams are asked to rate their confidence that the larger collective can complete all of their goals in the given time frame. This is generally rated on a scale from 1-5 with anyone giving a 1-3 rating being given the chance to speak up and share their concerns in front of the larger group.

Go-no-Go

In technology companies, or other project based organizations, many teams work in parallel to build and then release a product to one or many customers. In these instances, planning happens upfront, but the need for a final confidence vote can sometimes help all teams gain a voice in whether or not a project is ready for final launch. This type of exercise is generally binary with either yes or no being the only options. 

Yes / No voting

Every organization that functions with some level of democratic process has times where they need to vote on a course of action. Planning Poker can be used to facilitate these votes by offering card decks with yes or no options. Much like project planning, the organization can leverage the benefit of hearing everyone’s vote without the innate peer pressure from voting aloud. 

Tools

Planning Poker is a powerful technique, but requires a modicum of preparation by the facilitator. User stories must be prepared and special card decks must be acquired. To reduce these barriers and facilitate collaboration amongst remote teams, online planning poker tools can be used.