Who wins the Calder?

[angus]

Landeskog - he's had the biggest impact as a rookie on a winning team. RNH is going to be an amazing player, but Landeskog has been the more impactful rookie in my eyes.

[doulos]

I like stats that mean something. The stats that mean something are Landeskog's +/- relative to teammates and his goal, assist, and shot totals, as well as his TOI. That's it.

So .. stats that support your opinion are sweet and ones that do not are unsweet. Makes sense. Sweet.

[Big Ev]

Demonstrate.

well as you can see I accumulate 4.2154 rep points per every 1000 posts, thus my Corsi Rep is better than Doulos' who accumulates 2.89114 rep points for every 1000 posts.

[TK12555]

well as you can see I accumulate 4.2154 rep points per every 1000 posts, thus my Corsi Rep is better than Doulos' who accumulates 2.89114 rep points for every 1000 posts.

Actually, corsi rep would probably be the amount of up reps minus down reps every 1000 posts, I guess

[doulos]

Anoher point for Landeskog tonight. Not even fair.

[metaldude26]

well as you can see I accumulate 4.2154 rep points per every 1000 posts, thus my Corsi Rep is better than Doulos' who accumulates 2.89114 rep points for every 1000 posts.

You're right advanced stats are garbage... when misutilized.

It's all about context. Going back to your example of Marchand vs. Datsyuk. What you've shown is not that Marchand is better than Datsyuk but that Marchand has played his role better than Datsyuk has played his role this season.

Similarly, you have posted a lot more than Doulos since rep was instituted therefore you've used numbers to misrepresent your value.

No one is pretending that advanced stats are the be-all end-all but within the correct context they are quite useful. For instance, if we compared Landeskog and Marchand who have similar numbers by your metrics, we can see that Landeskog plays tougher minutes and is also much more effective in that he draws significantly more penalties than Marchand. You could of course come to this conclusion from watching the games but this is supporting data for when regular numbers don't tell the whole story.

[fungchen3]

The problem I see with advanced stats in hockey is that there are far too many intangibles. Hockey, unlike Baseball, is a sport where individual statistics hinge on a player's team. Let's take Gretzky for example - if he didn't have a high-end, intelligent sniper on his wing like Kurri, would he have produced as many points during his days with the Oilers?

There are so many factors that contribute to an NHL player's stats, including:

1. Linemates
2. Chemistry
3. Officiating
4. Emotion
5. Ice conditions
6. Coaching system
7. Special teams
8. Team defense
9. Goaltending
10. Injury

To me, evaluating a Baseball player's worth based on stats is a lot more feasible than trying to evaluate a Hockey player. A batter steps to the plate and 100% controls his own destiny each time he's up to bat. If a centerman in Hockey makes 5 passes that should have resulted in goals, they mean squat if the winger going to the net fails to bury a "gimme" goal. Another winger with better hands or timing could result in that same centerman having 3 extra assists in that game (making the same exact passes).

Look at a defenseman like Erik Karlsson. He's putting up a PPG on offense, but part of that is the fact he's playing for a coach who gives him the green light to take risks. If Karlsson was playing for a different coach, there's a very good chance his production takes a hit by 10-18 points during a season.

Sure, stats can tell some of the story, but in hockey there are too many influences that effect the stats IMO.

[metaldude26]

The problem I see with advanced stats in hockey is that there are far too many intangibles. Hockey, unlike Baseball, is a sport where individual statistics hinge on a player's team. Let's take Gretzky for example - if he didn't have a high-end, intelligent sniper on his wing like Kurri, would he have produced as many points during his days with the Oilers?

There are so many factors that contribute to an NHL player's stats, including:

1. Linemates
2. Chemistry
3. Officiating
4. Emotion
5. Ice conditions
6. Coaching system
7. Special teams
8. Team defense
9. Goaltending
10. Injury

To me, evaluating a Baseball player's worth based on stats is a lot more feasible than trying to evaluate a Hockey player. A batter steps to the plate and 100% controls his own destiny each time he's up to bat. If a centerman in Hockey makes 5 passes that should have resulted in goals, they mean squat if the winger going to the net fails to bury a "gimme" goal. Another winger and that centerman could have 3 extra assists in that game.

Look at a defenseman like Erik Karlsson. He's putting up a PPG on offense, but part of that is the fact he's playing for a coach who gives him the green light to take risks. If Karlsson was playing for a different coach, there's a very good chance his production takes a hit by 10-18 points during a season.

Sure, stats can tell some of the story, but in hockey there are too many influences that effect the stats IMO.

This is why I like advanced stats. Sometimes the run of the mill numbers don't tell the whole story with regard to how effective a player is or can make a weaker player look better. Like I said, it's supplemental data. You use it as part of telling your complete story.

[fungchen3]

This is why I like advanced stats. Sometimes the run of the mill numbers don't tell the whole story with regard to how effective a player is or can make a weaker player look better. Like I said, it's supplemental data. You use it as part of telling your complete story.

I hear ya metal. Here's my thing though (maybe you could answer this for me so I understand better). This is hypothetical, but say we decide to isolate a player like Scott Hartnell. Say during the 1st half of the season, his stats/production, etc. seems to be very powerful. Then, all of a sudden, in early-January, his centerman Claude Giroux breaks a few ribs. Let's say Giroux continues to play through the pain (never misses time) but because of the injury, Giroux is having a difficult time winning battles along the boards and/or making the same passes he was capable of making during the early months of the season. In this case, how would this intangible be reflected in Hartnell's advanced stats? Or would it appear HIS "performance" has dropped, even though he is playing the same exact game?

[metaldude26]

I hear ya metal. Here's my thing though (maybe you could answer this for me so I understand better). This is hypothetical, but say we decide to isolate a player like Scott Hartnell. Say during the 1st half of the season, his stats/production, etc. seems to be very powerful. Then, all of a sudden, in early-January, his centerman Claude Giroux breaks a few ribs. Let's say Giroux continues to play through the pain (never misses time) but because of the injury, Giroux is having a difficult time winning battles along the boards and/or making the same passes he was capable of making during the early months of the season. In this case, how would this intangible be reflected in Hartnell's advanced stats? Or would it appear HIS "performance" has dropped, even though he is playing the same exact game?

This would reflect in his stats but they would reflect in the stats of his linemates as well so the relative contributions would remain the same. But again hypothetically, if that line happens to luck into a few goals despite playing poorly would most people notice the difference? Advanced stats would show you that slow down before the standard numbers could. If however the line did stop producing then this would be a tangible effect.

One of the limitations of advanced stats is separating players from their linemates but within the context of that you can use advanced stats to instead show deficiencies in other numbers, such as Landeskog being so far ahead of his teammates in plus/minus. This is true but it is arguable if he has been better than Ryan O'Reilly. While the two have been linemates for a good portion of the season, O'Reilly has had to play much tougher minutes, which better explains the vast discrepancy in plus/minus than does actual quality of play.

[doulos]

Keep in mind that metrics like these are still quite undeveloped as far as hockey is concerned. Lots of room for improvement but so many great ways that they are being used to analyze players/teams now.

Like anything in fantasy hockey, they should be used in concert with all of the other tools we have, and of course given a critical eye.

I find that a lot of the stat stuff is particularly good for identifying team strengths and weaknesses. My experience with advanced stats has been nothing but good and they helped me win several prizes in playoff pool last year when I read some stats that were being tossed around at a few sites a few weeks before the season was over.

I do think that they can be improved on, but that's no reason to just throw them out completely.