Review Confidence Interval and Power Rankings in sim_runner

[highlyprofitable108]

Description:

There have been observations that the confidence intervals generated by the sim_runner are unusually narrow. This could potentially indicate an issue with the way we're calculating them or with the underlying data. Additionally, the power rankings used as part of the simulation might need a review to ensure they are being utilized correctly.

Tasks:

Review Confidence Interval Calculation:

• Validate the formula used in compute_confidence_interval method. Check if the data being passed to this method has sufficient variance.
• Ensure that the sample size is adequate for the confidence interval calculation.

Examine Power Rankings:

• Review how power rankings are being incorporated into the simulation.
• Validate if the rankings are up-to-date and representative of the current team performances.
• Check if there's any bias introduced by the power rankings.

Monte Carlo Simulation:

• Investigate the monte_carlo_simulation method to ensure that the sampling and subsequent calculations are being done correctly.
• Save and review a subset of the sampled data for any anomalies.

Data Quality:

• Ensure that the data being fed into the simulation is clean, accurate, and free of outliers that might skew the results.

Documentation:

• Update any documentation or comments in the code that might be outdated or misleading.

[highlyprofitable108]

Here's what I'm thinking:

• Finalize the data sent to the user
• Update the way we generate the sim outputs
• Walkthrough the data

[highlyprofitable108]

Done with the updated outputs.

Will review data points later