Best analytical players 2022 draft

[Silkytentacles]

Was curious to who were the best performers in these categories were for the upcoming draft.

Click to read more...

 

[Michael Farkas]

What categories?

 

[Silkytentacles]

Just curious in general for players who drive good analytical data under any type of categories associated with analytics

 

[Antiillafire]

Adam Sykora has really good analytics offensively and defensively from guys in Slovak league. Smaht scouting projects him as a top 9 forward, and a second round of early third.

 

[FlyguyOX]

Hard to get that kind of info on undrafted prospects unfortunately

 

[Michael Farkas]

Just curious in general for players who drive good analytical data under any type of categories associated with analytics

Analytics isn't a category. Nor are categories "associated" with analytics. Teams and proprietary sources count different things in different ways depending on what they find valuable.

Things like Corsi became public domain largely because NHL clubs had moved past it already. But when it became public, a lot of fans jumped to attention (same thing happened with plus/minus when it was an advanced stat, same thing with save pct. when it became an "advanced" stat...now, they hold much less appeal, of course).

This is ever-evolving. The real good ones are the folks who can combine everything. Proper talent evaluation as a scout, knowledge of how the game is played and what the expectation of a player to do within team tactics (coaching), and the ability to apply useful numerical inferences to these events (analytics).

If anyone is incomplete in any of those areas, they'll get a worse result. Just because someone puts something up as a statistic, doesn't mean it isn't biased. They're interpreting an event as the event that it's being labeled as and giving it a value in their evaluation process. So, you have to be careful about where you look and what you're looking at.

Good example, when I was at the draft in 2018, I got to chatting with a third-party analytics company who was down there (I won't say who) and they had NHL clients (they wouldn't say who haha) and they told me a little bit about their process (which I really didn't care for because it didn't have enough eyes on it) they showed me what the output of their model was...pretty common names at the top, fair enough, nothing earth shattering... but there was a player who was, if I recall, somewhere between 5 and 8 on their list overall in the draft...he ended up being a selected in the 3rd round and right now he's playing in the Cheese...

So, again, just because it's a number, doesn't make it right or objective or unbiased...similar to the eye test...my interpretation of what's valuable might not be yours..."buyer" beware if all of three boxes aren't checked off properly.

 

[Silkytentacles]

Analytics isn't a category. Nor are categories "associated" with analytics. Teams and proprietary sources count different things in different ways depending on what they find valuable.

Things like Corsi became public domain largely because NHL clubs had moved past it already. But when it became public, a lot of fans jumped to attention (same thing happened with plus/minus when it was an advanced stat, same thing with save pct. when it became an "advanced" stat...now, they hold much less appeal, of course).

This is ever-evolving. The real good ones are the folks who can combine everything. Proper talent evaluation as a scout, knowledge of how the game is played and what the expectation of a player to do within team tactics (coaching), and the ability to apply useful numerical inferences to these events (analytics).

If anyone is incomplete in any of those areas, they'll get a worse result. Just because someone puts something up as a statistic, doesn't mean it isn't biased. They're interpreting an event as the event that it's being labeled as and giving it a value in their evaluation process. So, you have to be careful about where you look and what you're looking at.

Good example, when I was at the draft in 2018, I got to chatting with a third-party analytics company who was down there (I won't say who) and they had NHL clients (they wouldn't say who haha) and they told me a little bit about their process (which I really didn't care for because it didn't have enough eyes on it) they showed me what the output of their model was...pretty common names at the top, fair enough, nothing earth shattering... but there was a player who was, if I recall, somewhere between 5 and 8 on their list overall in the draft...he ended up being a selected in the 3rd round and right now he's playing in the Cheese...

So, again, just because it's a number, doesn't make it right or objective or unbiased...similar to the eye test...my interpretation of what's valuable might not be yours..."buyer" beware if all of three boxes aren't checked off properly.

Thanks for that definitely some interesting stuff there. Always fascinating seeing how nhl teams operate. Im always taking things with a grain of salt and try to blend it in with other things when it comes to scouting. I made this thread because I was mostly looking for more prospect info. Like I hear that Lambert is one of the best transitional players moving the puck up ice and getting offensive zone entries based on data tracked by scouching

 

[Guttersniped]

Analytics isn't a category. Nor are categories "associated" with analytics. Teams and proprietary sources count different things in different ways depending on what they find valuable.

Things like Corsi became public domain largely because NHL clubs had moved past it already. But when it became public, a lot of fans jumped to attention (same thing happened with plus/minus when it was an advanced stat, same thing with save pct. when it became an "advanced" stat...now, they hold much less appeal, of course).

This is ever-evolving. The real good ones are the folks who can combine everything. Proper talent evaluation as a scout, knowledge of how the game is played and what the expectation of a player to do within team tactics (coaching), and the ability to apply useful numerical inferences to these events (analytics).

If anyone is incomplete in any of those areas, they'll get a worse result. Just because someone puts something up as a statistic, doesn't mean it isn't biased. They're interpreting an event as the event that it's being labeled as and giving it a value in their evaluation process. So, you have to be careful about where you look and what you're looking at.

Good example, when I was at the draft in 2018, I got to chatting with a third-party analytics company who was down there (I won't say who) and they had NHL clients (they wouldn't say who haha) and they told me a little bit about their process (which I really didn't care for because it didn't have enough eyes on it) they showed me what the output of their model was...pretty common names at the top, fair enough, nothing earth shattering... but there was a player who was, if I recall, somewhere between 5 and 8 on their list overall in the draft...he ended up being a selected in the 3rd round and right now he's playing in the Cheese...

So, again, just because it's a number, doesn't make it right or objective or unbiased...similar to the eye test...my interpretation of what's valuable might not be yours..."buyer" beware if all of three boxes aren't checked off properly.

What do you mean by “playing in the Cheese”?

 

[Michael Farkas]

Like I hear that Lambert is one of the best transitional players moving the puck up ice and getting offensive zone entries

This is a really good base. It's much more specific, there's a plausible answer(s) to it. And again, you open up another layer. I'm going to walk away from the thread after this, as it's not really designed for me...but "getting offensive zone entries" is a good one..."transitional [play]" is a good one. But what's next? Even that needs to be built on. What do these things represent? What do they lead to? There are real applications to these pieces of information, and they need to be connected to others.

To the above post, Cheese = ECHL

 

[DoingItCoolKiwi]

Saw these stat cards on top US prospects from earlier this year. Is this what youre looking for?

 

[emptyNedder]

My own model tries to identify undervalued D prospects (basically D ranked outside the top 50).

The prospects who appear significantly undervalued are: Otto Salin, Jake Furlong, Hudson Thornton, Adam Engstrom, and Niks Fenenko.

A few prospects are ranked in the 40-60 range, but model out as late 1st rounders: Tomas Hamara and Artyom Duda.