Reckon

Product Introduction

1. Definition: Reckon is a specialized iOS and iPadOS application, technically categorized as a decision journal and probabilistic calibration tracker. It functions as a personal data logging tool designed to capture the meta-cognitive aspects of decision-making.
2. Core Value Proposition: It exists to combat hindsight bias and provide empirical feedback on judgment quality. Its primary value is enabling users to measure and improve their decision-making calibration—the alignment between their stated confidence in predictions and the actual frequency of correct outcomes—through structured, private logging.

Main Features

1. Structured Decision Logging: The app provides a templated entry system for recording predictions before an outcome is known. Users input a specific forecast, a line of reasoning, and a numerical confidence level (e.g., 70%). This creates a timestamped, falsifiable record that prevents memory revision.
2. Dynamic Confidence Check-ins: As a decision unfolds, Reckon prompts for brief updates. Users can adjust their confidence percentage up or down and tag new information as positive or negative. This feature creates a historical audit trail of the decision's evolution, capturing the reasoning process over time.
3. Resolution & Calibration Analytics: Upon outcome resolution, users log what actually happened and their satisfaction. After accumulating data (typically a few dozen decisions), the app's reliability diagram algorithm processes the dataset. It visually plots confidence levels against outcome frequency, identifying systematic overconfidence or underconfidence across different confidence bands.
4. Offline-First, Privacy-Centric Architecture: The application employs a local-first data model. All records are stored directly on the user's device. Synchronization across iPhone and iPad is handled exclusively via the user's own iCloud account, utilizing Apple's CloudKit framework. There is no backend server, account system, or external data upload, ensuring complete data ownership and privacy.

Problems Solved

1. Pain Point: Hindsight bias and uncalibrated judgment. People often misremember their original predictions after learning an outcome, believing they "knew it all along." Furthermore, without tracking, individuals cannot objectively assess if their internal sense of confidence (e.g., "I'm 80% sure") corresponds to an 80% accuracy rate in reality.
2. Target Audience: The primary user personas are knowledge workers, investors, project managers, strategists, and anyone whose effectiveness depends on the quality of their probabilistic judgments. This includes roles like product managers making feature bets, traders assessing market moves, and leaders evaluating strategic initiatives.
3. Use Cases: Essential for reviewing investment theses, evaluating hiring decisions post-mortem, assessing project risk forecasts, improving personal forecasting accuracy, and conducting structured self-experimentation to identify cognitive biases in specific domains (e.g., "Am I overconfident in technical estimates?").

Unique Advantages

1. Differentiation: Unlike generic note-taking apps (e.g., Apple Notes, Notion) which store unstructured text, Reckon is a structured database for meta-cognitive data. It enforces the recording of a numerical confidence estimate and a binary outcome, enabling mathematical analysis that notes apps cannot perform. Unlike subscription-based analytics tools, it is a one-time purchase with no recurring fees.
2. Key Innovation: Its core innovation is the automated generation of a personal calibration curve from real-life decisions. By forcing the separation of prediction, confidence, reasoning, and outcome into discrete, timestamped data points, it transforms subjective judgment into an analyzable dataset. The "check-in" mechanism to track confidence drift is a novel approach to capturing the dynamics of a decision process.

Frequently Asked Questions (FAQ)

1. How does Reckon calculate my calibration? Reckon uses a statistical reliability diagram. It groups your resolved decisions by your stated confidence levels (e.g., all predictions made at 60-70% confidence) and calculates the percentage of those predictions that were correct. It then plots your stated confidence against this actual accuracy rate, visually revealing gaps where you were overconfident (accuracy lower than confidence) or underconfident (accuracy higher than confidence).
2. Is Reckon suitable for tracking daily personal decisions? While possible, Reckon is optimized for significant, uncertain decisions with clear, resolvable outcomes that matter to the user—such as business strategy calls, financial bets, or career moves. Using it for trivial daily choices may lead to data fatigue without meaningful insight into your judgment patterns.
3. Can I export my decision data from Reckon for further analysis? The current implementation prioritizes on-device privacy and simplicity, focusing on its built-in calibration visualization. For advanced users, data export functionality would be a potential future enhancement to allow for custom analysis in spreadsheet or statistical software.
4. Why is a one-time purchase model significant for a decision journal? A one-time purchase aligns with the long-term nature of calibration tracking. Improving judgment is a multi-year endeavor, and a subscription fee could create a disincentive to maintain a lifelong record. The model ensures users can build a continuous, decades-long dataset without recurring cost, increasing the tool's ultimate value and integrity.
5. What happens to my data if I switch to a new iPhone? Your decision journal data syncs via your personal iCloud account. When you set up a new iOS/iPadOS device with the same Apple ID and install Reckon, your data will automatically sync and download from iCloud, preserving your complete history and calibration analysis.