The Uncovered Truth About PDO and SH%

[BKarchitect]

I'm glad we've moved from discussing how regression works on relatively small numbers to discussing if dictionaries are lying about what words mean.

Well, I will say this poster has me reconsidering all those time growing up that my parents and teachers told me I was underachieving.

“Guys, you know…what I achieve is equal to my ability. Stop throwing your biased expectations around and pressuring me to achieve more! I can only achieve exactly what I am capable of achieving!”

 

[Dekes For Days]

No. it does not. Overachieving is not a word

Of course overachieving is a word, and it has an established meaning in the hockey world.

 

[ItWasJustified]

Of course overachieving is a word, and it has an established meaning in the hockey world.

No. You can't achieve more than your current ability.

 

[ijuka]

And how do you achieve luck? I give you a hint: You don't achieve luck by standing still on the ice.

I mean, ice hockey is recognized to be one of the most luck-based sports. It's not basketball.

If you don't believe that that's the case, then I can't help you.

 

[Ol' Jase]

No. You can't achieve more than your current ability.

That would indicate performance is a static metric, which, of course, is not the case.

 

[Dekes For Days]

No. You can't achieve more than your current ability.

I'm not sure where you got these ideas, but they're wrong. Teams can get outcomes for stretches that underachieve and overachieve what they are capable of sustainably doing. If you have a 100 point team, that team got to 100 points through stretches both above and below 100 points. Their "ability" is represented by their average, not their unsustainable peak.

 

[The90]

Yes it is.

No, I'm not and yes it does when it comes to humans doing something.

According to who? Who is expecting? You, me, everybody, somebody, anybody?

Flipping a coin is not, and will never be a game based on innate and trained skills like hockey so the comparison is absurd.

Big stats guy, huh?

 

[AnInjuredJasonZucker]

How do you achieve more than your ability/potential?

Come on, man. I gave you a link and everything. The least you could do is read it before asking a question that will be rendered irrelevant by said link.

 

[Bank Shot]

nonsense. if you watched them roll over anaheim and arizona recently you'd have to admit there's more to it than "they were all just on a heater, hur de hur". what a sadly weak necro-chirp. what i will say is that the canucks were on a score first heater and got used to playing with the lead and are now really struggling to play disciplined hockey from behind.

the larger explanation is that good teams have adjusted to the canucks system and exposed their defence, plus multiple canucks players have gone very cold and not just come off a heater. secondary scoring especially fell apart after the allstar break with joshua getting hurt and blueger and suter falling off the planet and lindholm coming in and sucking the air out of the building. pettersson has been playing like ass. plus demko has gone down and desmith is letting in deflating goals.

they still dominate ozone possession, generate chances and walk over lesser teams, but good teams have worked out the canucks weaknesses and the canucks have not adjusted and are now trying to learn how to play from behind, which has been painful and eratic with a ton of blown coverage.

You're 100% wrong. The Canucks were never going to continue to shoot 13% at 5v5 because that is something that teams don't do.

My argument is backed by decades of data on shooting percentage.

Your argument is backed by narrative......

 

[WarriorofTime]

I mean, ice hockey is recognized to be one of the most luck-based sports. It's not basketball.

Basketball has tons of variance. The three point shot is a +EV but ultra high variance action. Take that with two teams over the course of a game.. and well its not uncommon to say whichever team shoots better that particular night often wins.

 

[iamitter]

Come on, man. I gave you a link and everything. The least you could do is read it before asking a question that will be rendered irrelevant by said link.

It's fine, he clearly has some traits in common with neutron stars.

 

[Atoyot]

No. You can't achieve more than your current ability.

Is it possible for a team to win a game that they deserved to lose?

 

[GirardSpinorama]

PDO is obviously correlated to team success. However, outlier PDOs arent sustainable, so when teams "regress" in PDO they tend to also regress in the standings.

 

[ijuka]

Basketball has tons of variance. The three point shot is a +EV but ultra high variance action. Take that with two teams over the course of a game.. and well its not uncommon to say whichever team shoots better that particular night often wins.

Sportsbooks for NHL have had 0.654 log loss over the past ~5200 games, and 61.9% correct predictions.

Sportsbooks for NBA have had 0.612 log loss over the past ~5500 games, and 66.9% correct predictions.

So no, NBA does not have tons of variance, at least in comparison to hockey.

 

[WarriorofTime]

Sportsbooks for NHL have had 0.654 log loss over the past ~5200 games, and 61.9% correct predictions.

Sportsbooks for NBA have had 0.612 log loss over the past ~5500 games, and 66.9% correct predictions.

So no, NBA does not have tons of variance, at least in comparison to hockey.

That’s just because they have a wider spread in talent amongst teams than a depth based sport like hockey. It’d be like if you played the best five man lineup in hockey each like seventy five percent of the game. But the games themselves are still high variance even if less high variance than nhl. So nice try there.

 

[BurnabyJoe7]

Yes it is.

No, I'm not and yes it does when it comes to humans doing something.

According to who? Who is expecting? You, me, everybody, somebody, anybody?

Flipping a coin is not, and will never be a game based on innate and trained skills like hockey so the comparison is absurd.

No. You can't achieve more than your current ability.

Are you honestly trying to suggest there is no luck in sports? If so that is a ridiculous claim. The word has nothing to do with ability.

I'm glad we've moved from discussing how regression works on relatively small numbers to discussing if dictionaries are lying about what words mean.

Lmao. Discussing stats with someone that rejects a word's definition straight from the dictionary is definitely ambitious, and the discussion has gone as expected.

 

[daver]

Lmao. Discussing stats with someone that rejects a word's definition straight from the dictionary is definitely ambitious, and the discussion has gone as expected.

I think the discussion has actually overachieved.

 

[The90]

 

[JaegerDice]

I for one am shocked that a team that is 17th in the league in xGF/60 5v5 and 22nd in the league in SF/50 5v5 is no longer scoring goals at a top 5 rate.

SHOCKED.

 

[ItWasJustified]

That would indicate performance is a static metric, which, of course, is not the case.

No, it would not.

I'm not sure where you got these ideas, but they're wrong. Teams can get outcomes for stretches that underachieve and overachieve what they are capable of sustainably doing. If you have a 100 point team, that team got to 100 points through stretches both above and below 100 points. Their "ability" is represented by their average, not their unsustainable peak.

Nope. Your ability is your individual and team peak.

Come on, man. I gave you a link and everything. The least you could do is read it before asking a question that will be rendered irrelevant by said link.

Your dictionary link didn't render my argument irrelevant. You can't measure expectations.

Is it possible for a team to win a game that they deserved to lose?

Yes. Because there's always the factor of an individual mistake or an individual not being good enough/having enough ability.

 

[Ol' Jase]

No, it would not.

Nope. Your ability is your individual and team peak.

Your dictionary link didn't render my argument irrelevant. You can't measure expectations.

Yes. Because there's always the factor of an individual mistake or an individual not being good enough/having enough ability.

You are making absolutely no sense whatsoever.

 

[ijuka]

Yes. Because there's always the factor of an individual mistake or an individual not being good enough/having enough ability.

So you don't believe in lucky bounces?

 

[Romang67]

Measuring expectations isn't exactly a novel concept. We've been doing it for hundreds of years.

 

[dgibb10]

So you don't believe in lucky bounces?

Not even just lucky bounces.

Goaltending performance is incredibly volatile.

You can get lucky just in that variance there. For example, Vancouver had all 3 of their matchups vs Edmonton in the beginning of the year when those goalies couldn't stop a beach ball.

Skinner has since gotten his shit together and Campbell has been sent down and replaced by a much better goalie.

That's luck right there.

You may catch a team with the flu going through the dressing room, or battling injuries, or on a cold streak. That's luck right there.

 

[ijuka]

Measuring expectations isn't exactly a novel concept. We've been doing it for hundreds of years.

Some would say that the brain's main function is predicting the future, aka expectations. If you couldn't have expectations then you'd be too afraid of leaving your house.