"Expected" Assists?

[Filthy Dangles]

Based on the quality of chances or passes a player generates? I've seen certain metrics like 'Slot pass completions' before, I think that might be a good proxy and then you can build from that with other inputs. They can be fickle, and things like secondary assists which are probably a little more random exist. And truly quantifying the quality of a pass is probably pretty hard to do.

But has anyone attempted it or laid out a model for it?

Click to read more...

 

[morehockeystats]

The pass data is not publicly available.
My expected assists data is based on a mixture of previous assist data (Elo) and linemate goal-scoring quality (incl. pp)

 

[Mickey Marner]

Corey Sznajder tracks shot assist and you can download the data if you subscribe to his patreon.

Corey Sznajder is creating public database for hockey microstats. | Patreon

But I don't think you could do anything with the data beyond making a version of shooting percentage, except for passing instead. i.e., First assists ÷ first shot assists.

Worth mentioning the data is manually tracked, so the shot numbers recorded are usually different than the official NHL boxscore numbers.

 

[J bo Jeans]

Im sure Mcdavid would be the league leader or close to it. The amount of times I've seen his line mates botch prime scoring chances he generated is unbelievable.

 

[HockeyFan2022]

Check out Mitch Brown's twitter. He tracks xA1 (expected primary assists) for draft eligible prospects on a regular basis. Great stuff.