What is cedhtools?

cedhtools is an analytics platform designed specifically for competitive Commander (cEDH) players in Magic: The Gathering. Our application aggregates tournament data from Topdeck.gg, processes it, and presents it in an easily digestible format to help you make data-driven decisions for deck building and metagame analysis.

Whether you're trying to optimize your existing deck, explore new commander options, or stay ahead of the metagame trends, cedhtools provides you with the analytics and insights you need to gain a competitive edge.

How to Interpret the Data

Our data is organized around commanders and their associated cards. Here's what the key metrics mean:

• Inclusion %: The percentage of decks in our database that include a specific card for a given commander.
• Impact on Win Rate: The average win rate of a specific commander playing a specific card, minus that commander's average win rate overall.
• Confidence: A measure of how reliable the win rate impact data is, based on statistical analysis.

Understanding Confidence Scores

Our confidence score (0-100) is calculated using three key statistical measures:

• Sample Size (0-40 points): Based on power analysis for binary outcomes (win/loss). We use a sigmoid function to smoothly scale the score based on the number of games played with the card. A target of 100 games is used as a conservative estimate for statistical significance.
• Statistical Significance (0-30 points): Uses chi-square test with continuity correction for large samples, and Fisher's Exact Test for small samples (n < 5 in any cell). The score is based on the p-value:
  • p < 0.001: Very strong evidence (30 points)
  • p < 0.01: Strong evidence (20 points)
  • p < 0.05: Moderate evidence (10 points)
• Effect Size (0-30 points): Uses Cohen's h to measure the magnitude of the win rate difference, with 95% confidence intervals. The score considers both the effect size and the precision of the estimate:
  • |h| ≥ 0.8: Large effect (30 points)
  • |h| ≥ 0.5: Medium effect (20 points)
  • |h| ≥ 0.2: Small effect (10 points)
  The score is adjusted based on the width of the confidence interval - wider intervals reduce the score.

The final confidence score (0-100) is color-coded:

• Green (≥70%): High confidence in the win rate impact, based on large sample size, strong statistical significance, and precise effect size estimates
• Yellow (40-69%): Moderate confidence, with some uncertainty in one or more components
• Red (<40%): Low confidence, typically due to small sample size, weak statistical significance, or wide confidence intervals in the effect size

This scoring system helps you identify which card recommendations are based on solid statistical evidence versus those that should be taken with more caution. The confidence score is particularly useful for:

• Sample Size Assessment: Understanding if we have enough games to make reliable conclusions
• Statistical Reliability: Knowing how likely the observed differences are due to chance
• Effect Precision: Understanding how precise our estimate of the card's impact is

Data Interpretation Considerations

When using cedhtools for your analysis, please keep in mind:

• Sample Size Matters: Data from commanders or cards with fewer appearances in tournaments may be less statistically significant. We indicate sample sizes wherever possible, such as the number of entries for a specific commander, or the number of decks that include a specific card.
• Meta Context: The effectiveness of cards and commanders can change dramatically as the metagame evolves. Always consider when the data was collected and what the meta looked like at that time.
• Correlation vs. Causation: A high win rate for a card doesn't necessarily mean it's causing those wins. Consider the broader context of the deck and its strategy.
• Playgroup Variance: Tournament data reflects a competitive environment that may differ from your local playgroup's meta.
• Data Freshness: The cedhtools ETL pipeline updates data daily, but there may be a lag between tournament results being published and appearing in our database.
• Deck Registration Accuracy: cedhtools data relies on accurate deck registration from tournament organizers and participants.
• Decklist Updates: Due to how we collect data from Moxfield, it is entirely possible that the decklists we collect from a tournament are not the same as the decklists that were actually played, due to decklist updates by their authors. This means that the data we collect may not be 100% accurate. This tool is intended to be used as a guide, not a definitive source.

Data Sources

cedhtools currently aggregates data from Topdeck.gg tournaments. We're continuously working to expand our data sources to provide an even more comprehensive view of the CEDH landscape.

Our ETL (Extract, Transform, Load) pipeline processes this data daily to ensure you have access to the most current information available.

Data is collected from September 24, 2024 onwards.

Feedback and Suggestions

We're constantly improving cedhtools based on user feedback. If you have suggestions, questions about our data, or need help interpreting specific metrics, please reach out through our contact form or join our community Discord.