GDT: 2023 NHL Entry Draft Rounds 2-7 - 10AM ET, ESPN, ESPN+, SN

Status
Not open for further replies.

William H Bonney

Registered User
Feb 27, 2002
25,498
7,927
Colorado
Advanced analytics in the NHL are mostly trash, anyways.

Yep. The average fan doesn't understand how hard it is to generate quality structured data at scale. It's hard with apps that where you determine the structures that inform your data; NHL advanced stats are light-years behind that. The data quality is poor and highly subjective.
 

OilersFanatics505

Registered User
Aug 11, 2008
5,525
5,241
Yep. The average fan doesn't understand how hard it is to generate quality structured data at scale. It's hard with apps that where you determine the structures that inform your data; NHL advanced stats are light-years behind that. The data quality is poor and highly subjective.
hockey is also a very fluid game. Harder to not be subjective unlike baseball that is very rigid and structured to create those data points.
 

ChiHawks10

Registered User
Jul 7, 2009
28,771
22,704
Chicago 'Burbs
Yep. The average fan doesn't understand how hard it is to generate quality structured data at scale. It's hard with apps that where you determine the structures that inform your data; NHL advanced stats are light-years behind that. The data quality is poor and highly subjective.

Yep. Hockey is just so much more random than any of the other major sports, so it will always be hard to get any good data, or build any really good data models because of it. It's just not a sport that lends itself to structured data and analytics.
 

Bounces R Way

Registered User
Nov 18, 2013
37,048
59,651
Weegartown
Unfortunately true. The data quality in advanced stats is so poor and the analytical conclusions are usually worse.

Agree there. Very unreliable. There's a couple really bad ones. WAR and GSVA I find to be a complete waste of time. xG models also need to come a long way. That doesn't necessarily mean these stats hold no value but as a primary method of analysis they're simply not there.

In my mind until they microchip the pucks what's the point.
 

spfan

Registered User
May 4, 2009
1,863
541
I hope Colorado takes a swing with that pick from Montreal
Well they did. Although not a great fit. Avs don't need more d-men. Makar, Byram, gonna be hard for Gulyayev to really get the role he needs.

I like the player, just not the fit.
 

William H Bonney

Registered User
Feb 27, 2002
25,498
7,927
Colorado
Agree there. Very unreliable. There's a couple really bad ones. WAR and GSVA I find to be a complete waste of time. xG models also need to come a long way. That doesn't necessarily mean these stats hold no value but as a primary method of analysis they're simply not there.

In my mind until they microchip the pucks what's the point.

Same. I have no qualms with folks pursuing advancements in the NHL's data approach, whether the logging or analyses. It's just the dishonesty about where we're at in the progression. We have access to a ton of new stats and most of them are trash. The quality of new stats will improve but it's baby steps and yet people like Chayka pretend like they've cracked a code with their shit algorithms.
 

spfan

Registered User
May 4, 2009
1,863
541
The draft ends how the draft was overall. Low upside, certainty over high upside swings.
 
Status
Not open for further replies.

Ad

Upcoming events

Ad

Ad