Shot based stats and how to account for home arena scoring standards

[Freudian]

We know that for the real-time stats in the NHL there are wildly differencing standards for awarding hits/givaways/takeaways/blocks from arena to arena. That makes them more or less useless for anything important.

With shot attempt based stats becoming more and more popular, what is done to account for home/away scoring bias when it comes to shots/blocks/missed shots? If the data isn't good coming in, the data on the other end won't be good.

Just looking at the missed shot statistics the teams that are 1st-10th in missed shots in the league have 53.6% of their missed shots at home. The teams 11th-20th have 51% of their missed shots at home. The bottom ten teams in the league have 50.1% of the missed shots at home.

Is the same pattern evident for blocks and shots? Is one of the key factors for having good corsi/fenwick numbers having a good statistics staff at home?

Missed shot statistics:

Click to read more...

 

[Steerpike]

Check this out.

 

[Steerpike]

OnDiss% is (Home SF/FF)-(Away SF/FF)
OnADiss% is same but for shots against.

As goalies might influence their SA/FA numbers and backups mostly play away games its best to see scorekeeper differences by looking at OnDiss% though.

A quick conclusion might be that some goalies get overrated while other's get underrated. With Fleury probably the most overrated and Beriner and quick among the most underrated.

Because different scorekeepers seem to disagree on what's a shot and what's a missed shot it makes more sense to look at Fenwick shooting and save percentages:

This more clearly shows Rask/Bob are a little overrated while Quick and Bernier are underrated.


While you might think, "why not just look at goalie away stats?" This totally cripples the sample size, especially due to the fact that backups mostly play on the road.

 

[Freudian]

If we expand to all corsi events and compare the difference between home and overall corsi, the trend where the top corsi teams have on average a bigger difference at home than teams in general continues to emerge.

The difference between home and overall CF% for top ten teams are 1.7%. For the middle ten teams it's 1.4% and for the bottom team teams it's 1.1%.

There seems to be a bigger home corsi bias to the top CF teams. The weakness of the data is illustrated by some of the outliers where Devils are a better corsi team away than at home for example.

As long as collecting these stats are a manual enterprise and there is room for interpretation of when something should be counted as a shot/block/missed shot, all advanced stats derived from it will be weaker than stats based on more robust data.

 

[jwhouk]

Shots are NOT counted equally between home and road. Scorers assistants focus more when it's their own team that is accumulating the stat, not the opposition. It's the reason why hits, giveaways and takeaways are so out of whack - and why the three stars of the game selections are meaningless in some cities.

 

[Hivemind]

bedsheet, what is the header for that yellow column? Can't read white on yellow.

 

[Steerpike]

bedsheet, what is the header for that yellow column? Can't read white on yellow.

Sorry

H(home) SF/FF