Does anyone miss hockey analysis before analytics?

[MardyBum]

If you don't like them, don't use them. You're not going to accept someone elses "eye test" over your own anyways, so it doesn't make a difference whether they exist or not to you. No one is forcing you to do anything different, you just don't like change.

It's not "analytics" or "advanced stats" fault your arguments suck.

 

[Hisch13r]

God no. That shit was dumb as f***

 

[SnowblindNYR]

I also don't love some analytics. I think PDO is a dumb stat which makes a lot more sense as two separate stats. But for some reason the analytics community needs one stat to tell everything.

Here's an example of why I think PDO is dumb:

Team A: Save % 0.950 and shooting percentage 5%. PDO = 100. Good

Team B: Save % 0.920 and shooting percentage 8%. PDO = 100. Good

Team C: Save % 0.950 and shooting percentage 8%. PDO = 103. Lucky

Team D: Save % 0.920 and shooting percentage 5%. PDO = 97. Unlucky

Save % and shooting percentage are independent. So all four of those scenarios should be just as likely. However, the way we view the teams is completely differently.

Edit: It's not really true that all 4 scenarios are just as likely considering the save % and shooting percentages aren't all 50%. That said I doubt anyone quoting PDO ever measures how likely the scenarios are. And I'd bet if you would you'd find that the top 2 scenarios aren't significantly more likely than the bottom 2.

 

[SnowblindNYR]

If you don't like them, don't use them. You're not going to accept someone elses "eye test" over your own anyways, so it doesn't make a difference whether they exist or not to you. No one is forcing you to do anything different, you just don't like change.

It's not "analytics" or "advanced stats" fault your arguments suck.

I don't have an issue with analytics per se. In fact I like them. I'm a math guy and have a blog where I crunch number myself. I do miss simpler days where everything wasn't overanalyzed based on PDO. That admittedly is an emotional argument that was the basis of this thread.

The rational argument is that I also think there's something to be said about the people using analytics. Anyone that can sort PDO and tell that some team's number is higher than previous years thinks he's a smart analytics guy. There's a reason that in a real world analytics are used by people that actually have a concept of data. They usually have some quant degree. On HF boards it's pretty much any 15 year old in his mom's basement.

 

[AlainVigneaultsGum]

 

[Adam Fox Time]

Analysis based on analytics means the analysis is actually based on some truth.

Before you had idiots like Mike Milbury who had analyst jobs simply because they had a history in the league.

 

[SnowblindNYR]

Analysis based on analytics means the analysis is actually based on some truth.

Before you had idiots like Mike Milbury who had analyst jobs simply because they had a history in the league.

Would you prefer an idiot like Milbury trying to do math instead?

 

[powerbomb]

It’s not even worth having a discussion with someone who has learned the miracle of mathematics. They’re so convinced they now possess a crystal ball that they’ll warp any argument around any random number. It’s like a guy filibustering by reading the dictionary, and then nodding along with his buddies about how good at debate he is.

 

[I Hate Blake Coleman]

Stats in general are not indicative of future performance, which a lot of people miss.

I stopped trying to argue once the fancy stats took over. I don't give enough of a crap to learn much more. The old heads who said the eye test was best got laughed at but I trust what I can see now more than I do some spread sheet.

 

[Martin Skoula]

I also don't love some analytics. I think PDO is a dumb stat which makes a lot more sense as two separate stats. But for some reason the analytics community needs one stat to tell everything.

Here's an example of why I think PDO is dumb:

Team A: Save % 0.950 and shooting percentage 5%. PDO = 100. Good

Team B: Save % 0.920 and shooting percentage 8%. PDO = 100. Good

Team C: Save % 0.950 and shooting percentage 8%. PDO = 103. Lucky

Team D: Save % 0.920 and shooting percentage 5%. PDO = 97. Unlucky

Save % and shooting percentage are independent. So all four of those scenarios should be just as likely. However, the way we view the teams is completely differently.

Edit: It's not really true that all 4 scenarios are just as likely considering the save % and shooting percentages aren't all 50%. That said I doubt anyone quoting PDO ever measures how likely the scenarios are. And I'd bet if you would you'd find that the top 2 scenarios aren't significantly more likely than the bottom 2.

You need more extreme numbers to get into luck territory, 97-103 is the normal range. If a team with 15% shooting % runs into a team with .950 goaltending, one of them is necessarily going to have their numbers reduce closer to the average. An unstoppable force can’t regularly run into an immovable object without one of the two breaking down.

 

[Bounces R Way]

Hockey is just a sequence of 1s and 0s and it is up to us the viewers to do the math.

 

[Three On Zero]

Once you realize that advanced stats are highly subjective and don’t work for hockey, your life becomes much better

 

[Sidney the Kidney]

The problem with analytics are the people who use them, not the analytics themselves. Just like any sort of statistical data set.

This first sentence basically sums up my issue with them. It's not the analytics themselves, or whether or not they're useful. It's with a lot of the folks who use them incorrectly to "prove" things that these analytics don't even "prove".

 

[Super Hans]

View attachment 767385

Whatever did happen to Pierre McGuire?

 

[wmupreds]

It's easy to use them wrong, but I sure as hell am not going to pretend my "eye test" where I may see a given team 4 or 5 times a year is better than using data in an educated way

 

[Mattilaus]

Trying to add context to a teams performance by using statistics is a really silly thing to be upset about. Why do people want LESS information? I want as much information as I can get. Not just stats either. Give me numbers, give me game tape, give me professional scouting reports and post game interviews, give me all of it. It baffles me why some people want less information. If you don't want stats then just don't pay attention to them. No reason to come complain about them on the internet.

 

[Pablo El Perro]

Modern analytics empowers too many people to ignore the subtle nuances that make hockey the greatest game on earth.

It’s a game full of emotion, momentum, physicality, and flat out luck sometimes. Gotta closely watch the games to appreciate what is truly going on.

Analytics are just statistics. Unfortunately, not many people get the value and with it limitations of statistics, bad education in the Americas, and how they apply to the game itself.

 

[Pablo El Perro]

I also don't love some analytics. I think PDO is a dumb stat which makes a lot more sense as two separate stats. But for some reason the analytics community needs one stat to tell everything.

Here's an example of why I think PDO is dumb:

Team A: Save % 0.950 and shooting percentage 5%. PDO = 100. Good

Team B: Save % 0.920 and shooting percentage 8%. PDO = 100. Good

Team C: Save % 0.950 and shooting percentage 8%. PDO = 103. Lucky

Team D: Save % 0.920 and shooting percentage 5%. PDO = 97. Unlucky

Save % and shooting percentage are independent. So all four of those scenarios should be just as likely. However, the way we view the teams is completely differently.

Edit: It's not really true that all 4 scenarios are just as likely considering the save % and shooting percentages aren't all 50%. That said I doubt anyone quoting PDO ever measures how likely the scenarios are. And I'd bet if you would you'd find that the top 2 scenarios aren't significantly more likely than the bottom 2.

Regarding PDO, that's a bad application. Most good teams, on the year, are above 100. When people say it's unsustainable, they should be looking at teams with teams above 105 or so. Having said that, some teams sustain it.

 

[SnowblindNYR]

Regarding PDO, that's a bad application. Most good teams, on the year, are above 100. When people say it's unsustainable, they should be looking at teams with teams above 105 or so. Having said that, some teams sustain it.

Is 105 going to move the needle that much over 103?

 

[PaulD]

Yep. All this XY%FU garbage means nothing to me. l watch hockey with my eyes, not a stat sheet.

What he said !

 

[Pablo El Perro]

Is 105 going to move the needle that much over 103?

Over a large enough sample size, it will. But it's just a statistic. Which, in itself, is only known after the fact and can't be applied to a singular element of a sample, in other words, a team. A team with a 105 PDO, we can say it's likely they regress. But it doesn't mean they will in the least. It's just saying over a large enough sample of teams, most will. Unfortunately, most stats guys don't know how statistics apply to what they measure.

 

[PaulD]

I know this much

when on teams wins 6-1 in 3 periods

and another teams wins 3-2 in 3 periods + 3 on 3 OT + 6 penalty shots

THEY SHOULD NOT REWARDED THE SAME

 

[Pablo El Perro]

Over a large enough sample size, it will. But it's just a statistic. Which, in itself, is only known after the fact and can't be applied to a singular element of a sample, in other words, a team. A team with a 105 PDO, we can say it's likely they regress. But it doesn't mean they will in the least. It's just saying over a large enough sample of teams, most will. Unfortunately, most stats guys don't know how statistics apply to what they measure.

So I guess my take is analytics are helpful for explaining this or that after the fact, but they aren't the game itself.

For example,the Blackhawks PDO was better than the Panthers today, the Hawks still lost in regulation.

 

[ORRFForever]

YES!

Because whatever argument you make, there's a stat to refute it.

Post Auston Matthews disappears in the playoffs and 100 Toronto fans will attack you with pages of (obscure) stats that "prove" Matthews should win the Conn Smythe after losing in the first round.

Trust your eyes!

 

[PainForShane]

So... this morning the Edmonton Oilers fired their coach after the team started 3-9-1 (probably they decided to fire him after the team started 2-9-1, they won their game last night).

In the meantime, as of today, here is how the Oilers measure up with regard to expected goals for percentage at 5v5:

Yes, that 57.82% expected goals for percentage is good for best in the league.

I'd say actual results are much more important than the hypotheticals implied by fancy stats. How about you