Have a Big Decision to Make? Learn How to Create a Decision Tree to Find the Best Outcome

Decisions, decisions. From what to have for breakfast to whether to make a major career change, our lives are filled with choices big and small.

For the everyday choices, going with your instinct or doing a quick pro/con list is often sufficient. But what about those complex, high-stakes decisions that can have major consequences? The ones that keep you up at night as you try to work through all the angles and possibilities?

That‘s where decision trees come in. Developed in the 1960s by American computer scientist John Walker, decision trees provide a structured, analytical approach to decision making. By breaking down a complex choice into a clear diagram of options and outcomes, decision trees help you see the big picture, weigh tradeoffs, and zero in on the best path forward.

In this comprehensive guide, we‘ll walk through what decision trees are, why they work, and how to create your own step by step. Armed with this versatile tool, you‘ll be able to approach future decisions with confidence and clarity. Let‘s dig in!

What Exactly is a Decision Tree?

A decision tree is a map of the possible outcomes of a series of related choices. It starts with a single node, representing the initial decision to be made. This branches out into each of the possible options or paths. Those paths then lead to further nodes, representing the results or sub-decisions triggered by the earlier choice.

The tree continues branching out layer by layer until a final outcome is reached at the end of each path. The result is a sprawling diagram outlining every potential chain of events that could follow from your initial choice.

Example decision tree for deciding where to eat

A simple decision tree for deciding where to eat. By Mcld [CC BY-SA 4.0], from Wikimedia Commons

The structure of a decision tree comprises:

  • Root node: The starting decision, placed at the top or beginning of the tree
  • Decision nodes: Subsequent decision points or forks in the road
  • Chance nodes: Points where the outcome is uncertain, represented by a circle and probabilities
  • End nodes: The final outcomes at the end of each path
  • Branches: The arrows connecting nodes and representing the flow from one decision or outcome to another

While the concept of mapping out choices in a tree diagram may seem basic, decision trees have a surprising amount of analytical horsepower under the hood. That‘s thanks to two key features:

  1. The use of probability estimates to express the likelihood of chance events, and
  2. The inclusion of quantified outcomes, such as costs and revenue, at the end nodes

By attaching concrete numbers to each path, decision trees allow you to mathematically evaluate and compare different paths. And when faced with uncertainty, inputting probabilities lets you calculate the expected value of a path – essentially, the average outcome you‘d expect over many iterations.

For example, say you‘re deciding whether to launch a new software product. You think it has a 60% chance of becoming a hit, generating $1 million in profits, and a 40% chance of flopping and losing your $400,000 investment. Using decision tree math, you can calculate:

  • The expected value of launching is: (60% $1,000,000) + (40% -$400,000) = $440,000
  • Versus an expected value of not launching of: (100% * $0) = $0

So even though success isn‘t guaranteed and the potential loss is large, the risk is outweighed by the greater likelihood and magnitude of success. Seeing this tradeoff quantified gives you the rationale to confidently proceed.

This is the power of decision trees: Taking the gut-wrenching uncertainty out of difficult choices and grounding them in clear, objective math. By laying out all the considerations and running the numbers, decision trees let you maximize your upside while protecting against catastrophic risk.

How to Make a Decision Tree: Step by Step

Ready to try it yourself? Follow these steps to create a decision tree for your next big decision.

Step 1: Frame the decision

Start by clearly defining the decision at hand and the timeframe it covers. Are you deciding whether to attend college this year? Choosing which job offer to accept? Weighing a major purchase?

Frame the decision as a question and place it at the top of the page or diagram as the root node. For example: "Should I start my own business?"

Step 2: Identify available options

Brainstorm all the paths or options available to you. Draw these branching out from the root node as decision nodes.

For the "start a business" example, your options may be:

  • Bootstrap a startup on the side while keeping your day job
  • Go all-in and quit your job to start the business
  • Pursue funding from investors first, then start the business
  • Don‘t start a business; stay at your current job

Don‘t hold back during this step – write down every approach you can think of, even if some seem outlandish. Prune them back later if needed. The idea is to make sure you‘ve considered every angle.

Step 3: Map out potential outcomes

For each decision node, think through the potential outcomes that could result. What chain of events could unfold? Add chance nodes and branches to represent the possibilities and the probabilities you‘d assign to each.

For instance, if you bootstrapped the business on the side, maybe there‘s a:

  • 10% chance of high success (quitting day job to focus full time)
  • 40% chance of moderate success (keeping day job but growing side income)
  • 50% chance of the business failing

Repeat this process for all the decision nodes. Keep branching the tree out until a meaningful end point is reached for each path.

At this stage, take care not to miss any important outcomes or mistakenly bucket dissimilar results together. Be specific in naming and defining each outcome.

Step 4: Estimate probabilities and values

Once the paths are mapped out, start attaching numbers to the nodes. For chance nodes, estimate the probability of each outcome. These should sum to 100% for each branch.

For end nodes, quantify the value of each outcome, whether in dollars, utility, or other relevant metrics. Include both positive outcomes like revenue and cost savings as well as negative ones like expenses.

This is often the toughest part of the process, requiring research and careful thought to come up with reasonable figures. But it‘s worth the effort – these numbers are what allow the decision tree to weigh and compare paths.

Some tips for probability and value estimates:

  • Ground estimates in data where possible, such as industry benchmarks or historical results
  • Express estimates as ranges where precise figures are unknown
  • Do sensitivity analysis to see how changing estimates affects the outcomes
  • Sanity check estimates with knowledgeable colleagues or experts

Remember, perfection isn‘t the goal. Be as accurate as you reasonably can, but don‘t get paralyzed by uncertainty. As the statistician George Box put it, "All models are wrong, but some are useful".

Step 5: Calculate expected values

Finally, it‘s time to crunch the numbers. To determine the overall value of each path, you‘ll calculate its expected value (EV): the sum of each outcome‘s value multiplied by its probability.

For a simple path with a single outcome, that‘s just:

Expected Value = Outcome Value * Probability

For instance, if bootstrapping had a 10% chance of making $500,000 and a 90% chance of losing your $50,000 investment, the EV would be:

EV = (0.10 * $500,000) + (0.90 * -$50,000) = $5,000

For more complex paths with multiple stages of uncertainty, you‘ll need to multiply through all the probabilities in the sequence to determine the overall likelihood of reaching each end node. This can be easier to see in a table, like this example of deciding whether to drill for oil at different sites:

Decision tree solved to calculate expected values

Example decision tree solved to show expected value calculations. By IkamusumeFan [Public domain], from Wikimedia Commons

By solving for the EV of each path, you can now step back and compare the overall attractiveness of the available options, apples to apples. The path with the highest EV represents the best bang for your buck overall.

However, EV isn‘t always the end of the story. You‘ll still need to factor in qualitative considerations like strategic fit, opportunity cost, and risk tolerance.

An option with a slightly lower EV might still be preferable if it‘s significantly simpler to execute or aligns better with other priorities. Likewise, an high-EV but also high-risk path might be a no-go for a risk-averse decision maker.

Think of EV as an important input into the decision rather than an automatic final answer. The decision tree provides the framework for weighing the tradeoffs; ultimately, human judgment is still needed to take everything into account and make the call.

Advanced Decision Tree Concepts

Basic decision trees are just the tip of the iceberg. As you get more comfortable with the technique, you can explore advanced concepts to handle more nuanced real-world decisions.

Some examples:

  • Sensitivity analysis: Testing how the final EVs change when input probabilities and values are adjusted up or down. This helps identify which estimates are most critical and hedge against uncertainty.
  • Expected opportunity loss (EOL): Calculating the "cost" of making a suboptimal choice to help weigh whether getting more information is worth delaying the decision.
  • Multi-criteria decision trees: Incorporating additional metrics beyond just financial value, such as strategic fit, risk, or social impact. This allows considering more holistic objectives.
  • Influence diagrams: Decision tree-like models that more flexibly show the dependencies between choices, uncertainties, and outcomes over time. Useful for mapping out complex multi-stage decisions.

Of course, entire courses and textbooks are devoted to these methods. The key is to let your use of decision trees evolve along with the complexity of your decisions.

Start simple for low-stakes choices to build your comfort with the core concepts. Then gradually incorporate more advanced techniques as you take on weightier decisions with broader considerations.

Why Decision Trees Beat Going With Your Gut

When faced with a hard choice, our natural instinct is often to just pick a path and run with it. Surely our experience and judgment are sufficient to guide us to the right call – right?

Not so fast. Behavioral economists and cognitive psychologists have identified a host of mental traps and biases that cloud our thinking when it comes to complex decisions. A few of the biggest culprits:

  • Overconfidence bias: Overestimating the likelihood of positive outcomes. e.g. "I‘m sure my startup will succeed."
  • Confirmation bias: Favoring information that confirms preconceptions. e.g. Latching onto data supporting a choice you‘re leaning toward.
  • Availability bias: Relying on readily available information. e.g. Overemphasizing a single vivid anecdote vs. broad statistics.
  • Framing effects: Making different choices based on how options are presented. e.g. Responding differently to a 10% chance of failure vs. a 90% chance of success.

The common thread is that our minds take shortcuts when grappling with complex uncertainties. Rules of thumb and gut calls substitute for the hard work of carefully weighing probabilities and payoffs.

That‘s the beauty of decision trees. They cut through the murky mental fog to shine a bright, analytical light on the situation. Laying out the structure of the decision spurs you to explicitly consider aspects that might otherwise be glossed over. Assigning concrete probabilities and values replaces wishful thinking and handwaving with objective estimates.

Decision tree techniques have consistently been shown to improve real-world decision making. One study found that decision analysis boosted the return on investment of oil and gas projects by 30-40%. Another calculated that decision tree modeling increased the net present value of R&D projects by an average of 32%.

With results like these, it‘s no wonder decision trees have become a staple in industries like management consulting, capital budgeting, and pharmaceuticals. Whenever the stakes are high and the outcomes uncertain, decision trees have a track record of optimizing choices and delivering better results.

Cultivating Your Judgment Through Decision Trees

At their core, decision trees are more than just an analytical tool. They‘re a structured habit for navigating the inevitable tradeoffs and uncertainties of decision making. Regularly using decision trees to break down thorny choices hones your judgment and strategic thinking skills.

Over time, you‘ll find yourself getting better at rapidly identifying decision points, surfacing hidden options, and accurately estimating probabilities. As decision tree thinking becomes second nature, you‘ll be able to parse decisions more clearly even without diagramming every last branch.

In short, practicing with decision trees won‘t just improve the decisions you map out in full – it will fundamentally upgrade your judgment and intuition. You‘ll develop the pattern recognition and mental calluses to parse decisions more skillfully and confidently select the best path forward.

So start integrating this powerful tool into your decision-making process. Create decision trees for upcoming professional and personal choices. Share them with colleagues and collaborate to build a shared framework for team decisions. Teach them to your kids to give them a lasting edge in school and beyond.

Like compound interest, the benefits of sharpening your decision-making accrue over time, paying dividends with every incremental choice. Master decision trees and you won‘t just make radically better decisions – you‘ll set yourself up for a lifetime of better judgment and greater impact.

Decision trees don‘t guarantee perfect outcomes every time. Chance and unforeseen factors always play a role. But by providing a rigorous framework for grappling with tradeoffs and uncertainties, decision trees tilt the odds of success decisively in your favor. They are an indispensable tool not just for making better choices, but for realizing your full potential and achieving your grandest ambitions.

The next time you‘re wrestling with a major decision, don‘t just trust your gut. Take the time to map it out in a decision tree and harness the power of this technique. Your future self – and your future decision-making skills – will thank you.