How These Projections Work

A brief explainer about what goes into each projection.

My projections are built using the last three seasons of each player’s career on a per-minute basis, weighted by recency using a multi-variate regression, regressed to the mean based on sample size, and age adjusted. To account for the sudden increase in scoring two seasons ago, point projections were also adjusted to last season’s goals-per-game level.

Those per-minute projections for each stat are then applied to a time-on-ice projection based on each player’s ice-time last season, adjusted for their likely place on the depth chart this season. That means that a player moving from the second line to the first line would see a bump in ice-time equal to the difference between a first and second line player. The adjustments were based on expected depth charts that were derived from conversations with a beat writer from each team.

This year I’ve introduced future projections, up until 2025 in the dynasty package. How this is accomplished is through an age curve applied to time-on-ice, goals, assists and shots in each subsequent season. Keep in mind that the farther into the future a projection goes, the larger the margin of error is.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s