Jump to content
The Official Site of the Vancouver Canucks
Canucks Community

Benning was right about one thing, advanced stats are useless


terrible.dee

Recommended Posts

JB never said they were useless.   When he said they are like vitamins he meant, they are good for you, but are not going to cure cancer and make you live to 150 years old.

 

They have benefits, but your diet shouldn't revolve around vitamins, they are complements just like advanced stats.

 

You don't need to be Scotty Bowman to know that Letang is better than Hamilton.  However, take a guy that is evenly matched to Hamilton, then compare these two advanced stats, and you can use that to select which player you prefer.

Link to comment
Share on other sites

2 minutes ago, timberz21 said:

JB never said they were useless.   When he said they are like vitamins he meant, they are good for you, but are not going to cure cancer and make you live to 150 years old.

 

They have benefits, but your diet shouldn't revolve around vitamins, they are complements just like advanced stats.

 

You don't need to be Scotty Bowman to know that Letang is better than Hamilton.  However, take a guy that is evenly matched to Hamilton, then compare these two advanced stats, and you can use that to select which player you prefer.

Agreed. that's how advanced stats should be used. however that isn't the way it's been used in the media.

 

Most seem to use it to dump on any trade we've made. 

 

Don't forget we "lost" the Sutter and gudbranson trade because corsi. 

 

It seems to be used similar the +/- stat. you use it when it backs your theory and dump on it when it doesn't support your opinion. I was never a fan of advanced stats and never will be.

Link to comment
Share on other sites

Wow, that was awful and terribly incorrect. Read the guys over at Canucks army. They show that according to the advanced stats the Canucks are about 28th in the league in however they measure Corsi, and are close to that in almost every other advanced stat. The blog sounds like it was written by someone with a personal vendetta against advanced stats...or perhaps just math, and absolutely zero sources cited.

Link to comment
Share on other sites

Advanced stats are pretty good at predicting what a team will do. But for individual players, sometimes how the coach deploys them has more effect than skill or ability.

 

For instance, most of the self-appointed "experts" took a collective dump on the Subban/Weber trade, because their stats told them Subban was better. But the thing is, Weber happened to be paired with one of the best puck-moving D-men in the league, so the whole structure of their breakout went through Josi. Weber played closer to the blueline, basically acting as a wall that would force the team to dump the puck in to Josi, who played further back and would have time and space to make a play. (Newer advanced stats show that Weber was the best defenseman in the league at preventing a zone entry on his side.) The only time Weber would move the puck would be if the play had been defended well, and Josi was under pressure. Since everything else on the ice was structured to get the puck from the left side, making a significant play with the puck was far more difficult for Weber, and so he often took the safe route and dumped it out, to start over.

 

Subban, on the other hand, was the Josi for Montreal. His coach had him playing further back, which hampered his ability to score goals. Most people assumed this was just to compensate for Markov's foot speed, without acknowledging that playing further back gives you more time and space to make your plays.

 

So Subban had ideal circumstances to make his passes, and the whole team was structured in a way to receive passes from him. Weber did not, and often chose not to take risks to maintain possession. Geez, I wonder what the possession stats will indicate? And from this data, the "experts" claimed confidently that they knew better than Team Canada management, and Subban was a better defenseman than Weber.

 

...How's that working out so far, Jost???

 

Link to comment
Share on other sites

37 minutes ago, D-Money said:

Advanced stats are pretty good at predicting what a team will do. But for individual players, sometimes how the coach deploys them has more effect than skill or ability.

 

For instance, most of the self-appointed "experts" took a collective dump on the Subban/Weber trade, because their stats told them Subban was better. But the thing is, Weber happened to be paired with one of the best puck-moving D-men in the league, so the whole structure of their breakout went through Josi. Weber played closer to the blueline, basically acting as a wall that would force the team to dump the puck in to Josi, who played further back and would have time and space to make a play. (Newer advanced stats show that Weber was the best defenseman in the league at preventing a zone entry on his side.) The only time Weber would move the puck would be if the play had been defended well, and Josi was under pressure. Since everything else on the ice was structured to get the puck from the left side, making a significant play with the puck was far more difficult for Weber, and so he often took the safe route and dumped it out, to start over.

 

Subban, on the other hand, was the Josi for Montreal. His coach had him playing further back, which hampered his ability to score goals. Most people assumed this was just to compensate for Markov's foot speed, without acknowledging that playing further back gives you more time and space to make your plays.

 

So Subban had ideal circumstances to make his passes, and the whole team was structured in a way to receive passes from him. Weber did not, and often chose not to take risks to maintain possession. Geez, I wonder what the possession stats will indicate? And from this data, the "experts" claimed confidently that they knew better than Team Canada management, and Subban was a better defenseman than Weber.

 

...How's that working out so far, Jost???

 

If Subban and Josi both play deeper, then aren't the the ones taking the fore-check body contact to free up pucks for their partner or back checking forward?  I like you system description, but question the players' deployment.  Wouldn't a good coach want Webber taking the body contact, and Josi retrieving loose pucks?

Link to comment
Share on other sites

1 minute ago, Alflives said:

If Subban and Josi both play deeper, then aren't the the ones taking the fore-check body contact to free up pucks for their partner or back checking forward?  I like you system description, but question the players' deployment.  Wouldn't a good coach want Webber taking the body contact, and Josi retrieving loose pucks?

 

A good coach wants his D-men to avoid getting crushed altogether.

 

Subban and Josi are both smooth skaters, who avoid hits. Especially when they are playing further back, so they get to the puck faster than any forechecker can.

 

Link to comment
Share on other sites

That's DEFINITELY his blog. 

I think far from being a trend, we are just starting to properly dive into advanced stats. Like vitamins, there are still different opinions on what the perfect balance is. 

 

48 minutes ago, OptionalPants said:

Wow, that was awful and terribly incorrect. Read the guys over at Canucks army. They show that according to the advanced stats the Canucks are about 28th in the league in however they measure Corsi, and are close to that in almost every other advanced stat. The blog sounds like it was written by someone with a personal vendetta against advanced stats...or perhaps just math, and absolutely zero sources cited.

 

As OptionalPants said, there is a big difference between flat Corsi, and score adjusted Corsi. We have spent a tremendous amount of time trailing in games. We are all familiar with score effects on shots, which is unrealistically boosting our flat Corsi. No one in the analytics community is likely to be tricked by that these days. 

 

We are also finding more and more numbers to track as we dive deeper. High danger shots, percentage controlled entries, entries that result in a shot, multi shot shifts. There is absolutely no reason to avoid using the data. As long as you have a clear enough head to interpret it, and not use it as a crutch to support a preconceived point. 

 

 

Link to comment
Share on other sites

30 minutes ago, D-Money said:

 

A good coach wants his D-men to avoid getting crushed altogether.

 

Subban and Josi are both smooth skaters, who avoid hits. Especially when they are playing further back, so they get to the puck faster than any forechecker can.

 

Actually coaches fully expect players to take hits to free up pucks.  That's why Tanev get hit so often.  It's part of the game to take hits to make plays, isn't it?

Link to comment
Share on other sites

6 minutes ago, Alflives said:

Actually coaches fully expect players to take hits to free up pucks.  That's why Tanev get hit so often.  It's part of the game to take hits to make plays, isn't it?

 

Yes and no alf, you should know nothing is that cut and dry.  YES you take a hit to make a play IF you have to.  But the best D men in the league are also the best at rolling off checks and keeping themselves out of dangerous positions.

Link to comment
Share on other sites

  • b3. locked this topic

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...