NokiMo
Excel LADZ
Excel LADZ

patreon


Excel LADZ - Weekly Update: Monday 10 July

G’day lads, I hope you enjoyed the MLB Model released onto Excel LADZ. The model is inherently basic, and only involves two core steps:

This second step may be complicated to grasp for some. Why not use a regular probability distribution; e.g. the Poisson or Binomial Distribution?

Well, unlike many other sports where scoring occurs primarily at a single target, such as a goal or basket, baseball involves players progressing through a series of bases to score runs. This makes it more difficult to record a score in a single innings. However, when a team does score, they are more likely (compared to the Poisson Distribution) to record a high number of runs. Consider the chart below:

The above example was for a team expected to score 4.59 runs in a match. It is evident the regular Poisson Distribution drastically underestimates the probability of scoring a low (0) and a high (3+) number of runs in a match, compared to the more accurate probabilities presented by the specialised formula.

Ultimately, when building any model is it critically important to choose an appropriate distribution. Otherwise, the model loses its accuracy.

Reminder: For the MLB Model to update with new stats each day, you must ‘Refresh All’. This icon is within the ‘Data’ tab of Excel, and must be clicked in order for the Power Query to refresh its connection to the internet.

Short Term Goals

Medium Term Goals

If you have any questions/suggestions on video ideas, please put them in the comments lads! You can either comment on this post, or ask on the Excel LADZ Discord Server.

Thanks,

Excel LADZ.

Comments

Appreciate the update lad.

Samuel Waring


Related Creators