Algorithms to Live By

1. Optimal Stopping: When to Commit in Life's Decisions

"The 37% Rule defines a simple series of steps—what computer scientists call an "algorithm"—for solving these problems."

The 37% Rule. This mathematical principle suggests that when faced with a series of options (like potential partners or job candidates) and needing to choose the best one, you should examine 37% of the options to establish a baseline, then choose the next option that's better than all those you've seen. This rule applies to various life decisions:

• Partner selection: If you plan to date between ages 18 and 40, the optimal time to start seriously considering a long-term partner is around age 26.
• House hunting: Look at 37% of available houses before making an offer.
• Parking: In a linear parking lot, start looking for a spot after passing 37% of the lot.

The rule balances the risks of stopping too early (missing out on better options) and stopping too late (having already passed the best option). While it doesn't guarantee success, it optimizes your chances of making the best choice given limited information.

2. Explore vs. Exploit: Balancing New Experiences and Favorites

"Exploration is gathering information, and exploitation is using the information you have to get a known good result."

The explore/exploit tradeoff. This concept addresses the tension between trying new things (exploring) and sticking with what we know works (exploiting). It's relevant in various aspects of life:

• Dining out: Try new restaurants or return to favorites?
• Career choices: Pursue a new field or deepen expertise in current one?
• Travel: Visit new destinations or return to beloved spots?

The optimal balance changes over time:

• Early in life: Focus more on exploration to gather information and experiences
• Mid-life: Gradually shift towards exploitation as you identify preferences
• Later in life: Lean more heavily into exploitation, enjoying known pleasures

Strategies for balancing:

• Upper Confidence Bound algorithm: Choose options with the highest potential upside
• Gittins index: Assign values to options based on their potential and uncertainty
• A/B testing in business: Systematically try new approaches while maintaining successful ones

3. Sorting: The Power and Limitations of Organizing Information

"Scale hurts."

Efficiency decreases with size. Sorting becomes exponentially more difficult as the amount of data increases. This principle applies to both computational and real-world sorting tasks:

• Digital sorting: Algorithms like Merge Sort and Quick Sort are designed to minimize this scaling problem
• Physical sorting: Organizing a small closet is much easier than a large warehouse

Key insights:

• The best sorting method depends on the specific situation and goals
• Perfect sorting is often unnecessary and can be counterproductive
• Sometimes, leaving things unsorted is more efficient (e.g., messy desk)

Real-world applications:

• Library organization: Balance between perfect order and usability
• Sports rankings: Different tournament structures (e.g., round-robin, brackets) offer trade-offs between accuracy and efficiency
• Personal organization: Recognize when "good enough" sorting is sufficient

4. Caching: Efficient Information Retrieval and Memory Management

"Caching plays a critical role in the architecture of memory, and it underlies everything from the layout of processor chips at the millimeter scale to the geography of the global Internet."

Optimize for frequent access. Caching involves storing frequently-used information in easily accessible locations. This principle applies to both computer systems and human behavior:

Computer caching:

• Multiple levels of cache (L1, L2, L3) in processors
• Content Delivery Networks (CDNs) for faster web access
• Browser caching for quicker page loads

Human caching:

• Keeping frequently used items within easy reach
• Organizing living spaces based on usage patterns
• Mental "caching" of important information

Key strategies:

• Least Recently Used (LRU): Remove the least recently accessed item when space is needed
• Temporal locality: Recently used items are likely to be used again soon
• Spatial locality: Items near recently used items are likely to be used soon

Caching in everyday life:

• Home organization: Keep daily items easily accessible
• Work efficiency: Arrange your desk for quick access to common tools
• Learning: Review important information periodically to keep it "cached" in memory

5. Scheduling: Optimizing Task Completion and Time Management

"If you followed the best possible process, then you've done all you can, and you shouldn't blame yourself if things didn't go your way."

Process over outcome. Effective scheduling is about implementing the best process, not just focusing on results. Key scheduling algorithms and their real-world applications include:

1. Earliest Due Date (EDD):

   • Computer: Minimize maximum lateness
   • Life: Tackle urgent deadlines first

2. Shortest Processing Time (SPT):

   • Computer: Minimize average completion time
   • Life: Knock out quick tasks to build momentum

3. Moore's Algorithm:

   • Computer: Minimize number of late tasks
   • Life: Strategically choose which deadlines to miss when overloaded

Practical scheduling strategies:

• Break large tasks into smaller, manageable chunks
• Use time-boxing techniques (e.g., Pomodoro) to maintain focus
• Build in buffer time for unexpected issues
• Regularly reassess and reprioritize tasks

Avoid common pitfalls:

• Thrashing: Switching tasks too frequently, reducing overall productivity
• Priority inversion: Low-priority tasks blocking high-priority ones
• Over-optimization: Spending more time planning than doing

Remember that perfect scheduling is often impossible due to uncertainty and changing circumstances. The goal is to implement a robust process that adapts to new information and constraints.

6. Bayes's Rule: Making Better Predictions with Limited Information

"Small data is big data in disguise."

Update beliefs with new evidence. Bayes's Rule provides a framework for making rational predictions based on limited information and updating those predictions as new data becomes available. This approach is valuable in various domains:

Applications of Bayesian thinking:

• Medical diagnosis: Updating disease probability based on test results
• Financial forecasting: Adjusting market predictions with new economic data
• Criminal investigations: Reassessing suspect likelihood as evidence emerges

Key Bayesian concepts:

• Prior probability: Initial belief before new evidence
• Likelihood: Probability of the evidence given the hypothesis
• Posterior probability: Updated belief after considering new evidence

Practical Bayesian strategies:

• Maintain an open mind to new information
• Seek out disconfirming evidence to challenge your beliefs
• Regularly update your predictions as new data becomes available
• Recognize the limits of your knowledge and express uncertainty

Bayesian thinking in everyday life:

• Dating: Update your assessment of compatibility as you learn more about a person
• Career decisions: Adjust your plans based on new experiences and feedback
• Consumer choices: Refine product preferences with each purchase and use

7. Overfitting: The Dangers of Overthinking and Overcomplicating

"Err on the side of messiness."

Simplicity often outperforms complexity. Overfitting occurs when a model or decision-making process becomes too complex, fitting noise in the data rather than underlying patterns. This concept applies to both machine learning and human decision-making:

Machine learning overfitting:

• Training data fit perfectly, but poor performance on new data
• Solution: Cross-validation, regularization techniques

Human overfitting:

• Overthinking decisions, leading to analysis paralysis
• Creating overly complex plans that fail in real-world conditions

Strategies to avoid overfitting:

• Embrace uncertainty and imperfect information
• Use simple heuristics for quick decision-making
• Periodically step back to assess the big picture
• Test ideas in small, low-risk scenarios before full implementation

Real-world examples of overfitting:

• Financial markets: Overly complex trading strategies that fail in changing conditions
• Project management: Excessively detailed plans that crumble upon first contact with reality
• Personal relationships: Overanalyzing social interactions, leading to awkwardness

Remember that the goal is to find the right level of complexity for the problem at hand, not to eliminate all uncertainty or imperfection.

8. Relaxation: Simplifying Complex Problems for Better Solutions

"If you can't solve the problem in front of you, solve an easier version of it—and then see if that solution offers you a starting point, or a beacon, in the full-blown problem."

Strategically simplify. Relaxation in computer science involves temporarily removing or simplifying constraints to make a problem more manageable. This approach can be applied to various real-world challenges:

Types of relaxation:

1. Constraint Relaxation: Temporarily ignore certain rules or limitations
2. Continuous Relaxation: Treat discrete choices as continuous variables
3. Lagrangian Relaxation: Convert hard constraints into penalties

Benefits of relaxation:

• Provides a starting point for tackling complex problems
• Offers insights into the structure of the problem
• Can lead to "good enough" solutions when perfect solutions are impractical

Applying relaxation to everyday problems:

• Career planning: Start with an ideal scenario, then adjust for reality
• Budget optimization: Begin with unconstrained spending, then prioritize
• Creative projects: Brainstorm without limitations, then refine ideas

Examples of relaxation in action:

• City planning: Design an ideal city layout, then adapt to geographical constraints
• Product development: Envision a perfect product, then work within technological limitations
• Conflict resolution: Imagine an ideal outcome, then negotiate towards a realistic compromise

Remember that relaxation is a tool for generating ideas and starting points, not a replacement for dealing with real-world constraints. The key is to use the insights gained from the simplified version to inform your approach to the full problem.

9. Randomness: Harnessing Chance for Problem-Solving

"Sometimes the best solution to a problem is to turn to chance rather than trying to fully reason out an answer."

Embrace controlled chaos. Randomness can be a powerful tool for breaking out of local optima and finding innovative solutions. This principle applies in both computational and human problem-solving:

Computational uses of randomness:

• Simulated annealing: Gradually reducing randomness to find good solutions
• Genetic algorithms: Using random mutations to evolve better solutions
• Monte Carlo methods: Random sampling to estimate complex probabilities

Human applications of randomness:

• Creative brainstorming: Random word associations to spark new ideas
• Decision-making: Coin flips to break analysis paralysis
• Learning: Exploring random topics to broaden knowledge

Benefits of incorporating randomness:

• Overcomes cognitive biases and entrenched thinking patterns
• Explores a wider solution space, potentially finding unexpected solutions
• Provides a way to make progress when rational analysis fails

Strategies for harnessing randomness:

• Use random prompts or constraints in creative work
• Periodically introduce random elements into routines to avoid ruts
• Embrace serendipitous encounters and unexpected opportunities

Examples of productive randomness:

• Scientific discoveries: Accidental findings leading to breakthroughs
• Art and music: Experimental techniques producing innovative works
• Career paths: Unexpected detours leading to fulfilling opportunities

While randomness can be powerful, it's important to balance it with structured thinking and to recognize when precision is necessary. The goal is to use randomness as a tool, not to rely on it entirely.

10. Networking: Understanding and Optimizing Human Connections

"The message is simple but profound: if we're willing to accept solutions that are close enough, then even some of the hairiest problems around can be tamed with the right techniques."

Optimize connections, not just content. Networking principles from computer science offer insights into human social networks and communication:

Key networking concepts:

1. Packet switching: Breaking messages into small, manageable pieces
2. Routing protocols: Finding efficient paths for information flow
3. Congestion control: Managing overload in communication channels

Applying networking principles to human interactions:

• Information sharing: Break complex ideas into digestible chunks
• Relationship building: Establish multiple

Last updated: August 11, 2024