Improving Our Estimation Abilities: Human Biases

After we shared estimation thoughts on Twitter, we received a question in the Embedded.fm slack group:

What do you guys do for estimates and how do you improve your estimates? Any resources to help understand estimates more or just helpful things in general?

Following our article on why we estimate, we’ve put together a series of articles where we approach estimation from multiple directions: the understanding of human biases, our practical approach to estimation self-training, practical advice that we’ve learned from experience, and other approaches to estimation.

This article focus on biases that impact our thinking, particularly in regards to making, evaluating, communicating, and changing estimates.

Table of Contents:

  1. Understanding the Human Machinery
  2. The Planning Fallacy
  3. Ambiguity Effect
  4. Anchoring
  5. Framing Effect
  6. Hyperbolic Discounting
  7. Sunk Cost
  8. Not Invented Here
  9. Groupthink
  10. Attentional Bias
  11. Summary
  12. Further Reading

Understanding the Human Machinery

As humans, we are all governed by the machinery that has evolved in our bodies and minds from the very beginning of life on Earth. We are subject to the whims of our emotions and reactions, often without realizing what is happening.

To be rational, we must learn to counter or embrace these forces as appropriate. They will never disappear. We will always wrestle with them.

We’ll review biases that seem to rear their ugly heads whenever we’re making estimates and creating plans. Other biases are relevant for when we try to communicate or update our plans.

For a deeper insight into how biases impact your thinking and behavior, we recommend these resources:

The Planning Fallacy

The most impactful bias is the Planning Fallacy, a phenomenon where an optimism bias is present when making predictions of the time needed to complete a future task, causing us to underestimate the required time. This phenomenon can be seen even when you are aware that previous tasks of a similar nature also took longer to complete than expected.

The Planning Fallacy is internal in nature. We don’t provide optimistic estimates for tasks that others are going to complete. Instead, whenever we are predicting how long it will take another person to complete a task, we tend to show a pessimistic bias and overestimate the time required to complete the task.

A more humorous take on the planning fallacy is Hofstadter’s Law, which states:

It always takes longer than you expect, even when you take into account Hofstadter’s Law.

There are a variety of potential reasons for this, but here are a few that are evident from our experiences:

  • We tend to plan for the happy path, rather than planning for things to go wrong
  • Wishful thinking leads us to giving shorter estimates because we want to complete tasks quicker
  • We make estimates assuming we will put in 8 hours of work a day, which means we forget to account for meetings, meals, interruptions, distractions, sick days, vacations, and last-minute fire drills

In his article Why Software Tasks Always Take Longer Than You Think: A Statistical Model, Erik Bernhardsson proposes that developers are good at estimating the median time to complete a task, but not the mean, which seems to be related to the planning fallacy:

Let’s say you estimate a project to take 1 week. Let’s say there are three equally likely outcomes: either it takes 1⁄2 week, or 1 week, or 2 weeks. The median outcome is actually the same as the estimate: 1 week, but the mean (aka average, aka expected value) is 7⁄6 = 1.17 weeks. The estimate is actually calibrated (unbiased) for the median (which is 1), but not for the mean.

Ambiguity Effect

The Ambiguity Effect causes us to avoid choosing options involving ambiguity, uncertainty, or a lack of information. We tend to select options that we know about or think will have a positive outcome.

We need to clear up ambiguities before evaluating different solutions, because we will be drawn to the option with the least unknowns.

Anchoring

Anchoring is a bias where we will focus primarily on the initial piece of information offered (the anchor). Once the anchor is set, all future discussions, negotiations, arguments, and estimates are discussed in relation to the anchor.

You must be aware of Anchoring when initially creating your estimates, where you are likely to be stuck on the initial approach, timeline, or other information you’re provided with. Additionally, once an estimate has been provided, even if you say it before diving in and thinking deeply about it, future statements will be evaluated based on your initial number. Be careful about throwing out numbers carelessly, you may be setting yourself up for conflict.

Framing Effect

The Framing Effect causes us to evaluate options differently depending on how they are presented. When there is potential for positive impact, such as a financial gain, we tend to avoid risk. Whenever there is potential for a negative gain, such as a loss, we are more likely to take a risk.

The Framing Effect means that context has a significant impact both on how we create estimates and how estimates are received.

Hyperbolic Discounting

Hyperbolic Discounting describes our preference for a reward that arrives sooner over a reward that arrives later. We “discount” the value of the later reward by the delay we have to endure. In simple terms, we prefer the immediate payout to the more difficult route that produces a greater payout, but at a later time. Perhaps this explains why software teams are constantly putting in “hacks” and never doing things “the proper way”.

Sunk Cost

Escalation of Commitment, also known as the Sunk Cost Fallacy, causes us to continue to invest in a course of action in the face of increasingly negative outcomes.

From an estimation perspective, when plans go off the rails, people tend to invest further in the current approach rather than looking for alternate paths. Detach and break that spell if you are able to. Find another solution.

Not Invented Here

Most programmers and engineers are familiar with the “Not Invented Here” syndrome. We avoid using or buying existing products, standards, or ideas created outside of our organization. We prefer to reinvent the wheel internally instead. It doesn’t matter that we are spending more time and money reinventing the wheel than we would pay in fees or royalties.

How does this impact estimates? If we take the long, arduous route of reinventing the wheel, rather than leveraging an existing solution, then, due to the Planning Fallacy, we will have optimistic expectations for completing the task.

Groupthink

Groupthink is a phenomenon related to human social dynamics. We are all drawn by a deep-seated desire for harmony, conformity, and cohesiveness in groups. We want to minimize conflict within the group.

The impact of Groupthink is that when in a group, we tend to avoid raising controversial issues, pushing back on ideas we think are wrong, or providing alternative solutions. We lose the benefits of individual creativity and thinking. With respect to estimation and planning, teams will often pick the “happy path” or choose the estimates that seem most favorable for the team’s goals, regardless of the underlying reality. Dissenters or those with experience who point out potential flaws and missteps are likely to be silenced or ignored.

Attentional Bias

Attentional Bias is the tendency for our perceptions to be affected by our recurring thoughts. In simple terms, if you think about shoes and pay attention to shoes, you are going to notice more people’s shoes out in the world.

From an estimation perspective, the impact of the Attentional Bias is that we are predisposed to identifying problems and risks that we think about regularly, while ignoring other potential issues. Getting an outside perspective can help combat this bias.

Summary

We all intend to be perfectly rational, clear-headed thinkers, when in fact we are stuck with the “machinery” that has evolved since the first inklings of life on Earth.

To a large degree, we cannot escape these biases. They are built into us. We can, however, learn how our minds deceive us. We can adjust our plans and estimates in the face of these biases, as well as understand how others will perceive our plans. We will not always be successful in our efforts, but we will move ourselves closer to reality and rationality.

In the next article, we’ll explore Embedded Artistry’s approach to estimation self-training.

Further Reading

Share Your Thoughts

This site uses Akismet to reduce spam. Learn how your comment data is processed.