On Corsi, Evidence, and Proprietary Research

It happened again.

Someone said something against Corsi; maybe even in jest with some slight sarcasm. The masses counter attacked. Then two sides started to fire shots at the other.

It is not the first time, and it won’t be the last time.

In some ways, I understand and agree with the sentiment of letting things just go. Move on. Often we are only preaching to the choir.

Still, it’s an opportunity to teach and use examples of misconceptions, and also warn against potential issues. There are some thirsty to know more.

This is not the first time I’ve written in response to misconceptions, but alas here we are.

Let’s take a look.

Continue reading

Analyzing the Neutral Zone (and beyond!): A Call for Volunteers.

Over here at Hockey-Graphs, I’ve been taking a look at Neutral Zone Tracking and what the results of such tracking can tell us about the modern game of Hockey. My last post, for example, looked at what such tracking told us about play in the offensive, defensive, and neutral zones, and to what extent was performance in each of those zones skill or randomness at the team and individual levels. Before next season begins, I have two more posts planned: one doing further research into repeatability of various neutral zone metrics and another looking at how score effects affect neutral zone play.

The biggest weapon I’ve been using recently to do such research has been Corey Sznajder’s (aka @ShutdownLine) All Three Zones Project (the genesis of which can be found HERE).   Through Corey’s work, we have a nearly complete dataset with Neutral Zone Data for the 2013-14 season, for EVERY team in the league.  The dataset also includes to a lesser extent zone exit and zone entry defense data.  Even more, Corey is actually really close to finally finishing the tracking this summer and we’ll have a complete data set very soon.  This is an incredible resource for new hockey knowledge, and it’s available for anyone , as long as you make a $15 contribution to Corey’s GoFundMe Page, which can be found here.   I think this is an incredibly worthwhile use of funds if you can afford it, and many people have indeed actually already given.

What many people haven’t done is actually taken the data and DONE anything with it.  Again, this data set was released about a year ago now, and yet very few individuals have been doing work with it at all.  This is a bit of a disappointment.  If you include the passing tracking project of Ryan Stimson (Data available here, here, and here), we now have two large datasets filled with information that could help us better understand the game of hockey.

So what I’m asking is this for volunteers to try and actually try and do some work with the data.  There are a lot of questions we can try to answer!

For example, one question someone posed to me on twitter is:

What really is the risk of a D-Man attempting to jump up through the neutral zone and carry the puck-in?  

Now we can’t answer that question fully from the data – a D Man who tries to carry the puck through the NZ and turns the puck over at the red line is making a bad play that our data can’t capture – but we can take a shot by looking at the effects on shots against of carry-ins and carry-in failures by defensemen.  And this is just one question that we can possibly look at.

Please email me at my email address (garik16 AT Gmail) if you’re interested in looking at the data, or leave comments here in the comment section.  Alternatively, if you have other questions you might like answered, please leave them in the comments as well.  Together we can hopefully learn a lot more about the game of hockey.

Hockey Graphs Podcast Episode 20 – The Draftening


Ya, it is a little late and was already released on iTunes, but we thought you want to see it here.

Garret and Rhys return this week to break down the 2015 entry draft by profiling a whole ton of draft eligible player, and look at where they may have gone wrong in their personal draft rankings.

Join us after the jump (or on iTunes!) for a detailed look at 2015’s top prospects.

Continue reading