How the Sabres’ 2022 NHL Draft strategy was boosted by Sam Ventura and the analytics department

[LGR4GM]

https://theathletic.com/3475314/2022/08/11/sabres-analytics-sam-ventura-nhl-draft/?source=emp_shared_article

Quote

When Ventura started, analytics for the draft boiled down to taking the numbers of how a player scored in his own league and trying to project that out to how he would score in the NHL.

“Those ideas are a decade old now so it’s not state-of-the-art anymore,” Ventura said. “Now we have play-by-play data and can dig into the details of what’s happening on the ice in each moment from leagues all around the world. We can get almost an NHL-level analysis of players but with players playing in their own league. It has changed the game for us not only for what we’re able to say about players but the confidence with which we can say what we say. That’s the big difference from when I started. When I started it was, ‘We think player A is better than player B but we’re not so sure.’ Now we can form those opinions with a much higher degree of confidence.” 

This is a fascinating read and something if you are curious about the draft you should definitely take the time to go through. 

[LGR4GM]

They are looking at primary points (goals and 1st assists) and even strength (p1/ev). 

They are looking scoring relative to the team you are own (that's how you get Seth Jarvis). 

They are looking at age compared to scoring. 

They are looking at league strength as a variable (which might be why the Q isn't a place they seem to draft out of a ton recently). 

They are looking at playmaking which is probably some combination of passing stats (hd passes, passes leading to hd chances, passes that result in zone entries, and also controlled entries etc...) 

Basically it sounds like they spent most of 2021 building the models they want for drafting going forward and now they can focus on tweaking those models and updating them. Outside of the Leinonen pick, I can see how this was used. That Leinonen pick though is still really odd. 

[dudacek]

4 minutes ago, LGR4GM said:

They are looking at primary points (goals and 1st assists) and even strength (p1/ev). 

They are looking scoring relative to the team you are own (that's how you get Seth Jarvis). 

They are looking at age compared to scoring. 

They are looking at league strength as a variable (which might be why the Q isn't a place they seem to draft out of a ton recently). 

They are looking at playmaking which is probably some combination of passing stats (hd passes, passes leading to hd chances, passes that result in zone entries, and also controlled entries etc...) 

Basically it sounds like they spent most of 2021 building the models they want for drafting going forward and now they can focus on tweaking those models and updating them. Outside of the Leinonen pick, I can see how this was used. That Leinonen pick though is still really odd. 

I think Leinonen must have been more of an organizational decision, not a scouting decision.

And what I mean by that was that Adams felt it was a necessity to add a good prospect to the goaltending stable this year and made a calculated decision that using pick #41 on "the best goalie available" even if he wasn't the best player available was the best way to do that.

[GASabresIUFAN]

It all sounds great for scoring forwards, but what about preventing goals?  

[Doohicksie]

1 hour ago, LGR4GM said:

That Leinonen pick though is still really odd. 

I don't think so.  I think he was the goalie they wanted and determined that he likely wouldn't be there when their next pick came around.

[Doohicksie]

4 minutes ago, GASabresIUFAN said:

It all sounds great for scoring forwards, but what about preventing goals?  

I'm guessing they are looking at that one of two ways:

• Draft for point production, understanding that some players who fail to produce at higher levels can be salvaged as defensive role players.  I'm always surprised to hear that people of the John Scott/Brad May/Rob Ray mold were often top scorers at lower levels.
• There is a second set of metrics that they have for defensive players and don't talk about it much/people don't ask about it much/it's just not as sexy.  If it exists this set of metrics is probably applied to later round picks (no higher than 3rd round).

Back to goalies:  I wonder if they have goalie metrics that set Leinonen uniquely apart from the rest of the goalies in this year's class and simply didn't want to settle for second best. 

Given where Kevyn's record on goaltending right now, he may have decided to spend high-ish picks over the next draft or two toward making sure there are great goalies in the pipeline for either future Sabres use or to get a blue chip skater at the deadline during a playoff run (like the Panthers did with Reino).  Get a bunch solid goalie prospects, harvest the best for use by the team, use the rest (which may still be very good) as trade bait.

[LGR4GM]

58 minutes ago, GASabresIUFAN said:

It all sounds great for scoring forwards, but what about preventing goals?  

Considering Ventura didn't speak about all they look at, I would assume they look at entry denial, shot charts, breakouts to gauge some of the defensive metrics. 

[pi2000]

2 hours ago, LGR4GM said:

They are looking at primary points (goals and 1st assists) and even strength (p1/ev). 

They are looking scoring relative to the team you are own (that's how you get Seth Jarvis). 

They are looking at age compared to scoring. 

They are looking at league strength as a variable (which might be why the Q isn't a place they seem to draft out of a ton recently). 

They are looking at playmaking which is probably some combination of passing stats (hd passes, passes leading to hd chances, passes that result in zone entries, and also controlled entries etc...) 

Basically it sounds like they spent most of 2021 building the models they want for drafting going forward and now they can focus on tweaking those models and updating them. Outside of the Leinonen pick, I can see how this was used. That Leinonen pick though is still really odd. 

So basically TRpm 🙂

[GASabresIUFAN]

7 minutes ago, LGR4GM said:

Considering Ventura didn't speak about all they look at, I would assume they look at entry denial, shot charts, breakouts to gauge some of the defensive metrics. 

I would hope so, but I haven’t seen that reflected in who they are drafting, especially early in the draft.  Getting Novikov and Komarov seem like intelligent late rd picks, but I’d have liked to see at least one defensively responsible D taken in the top 3 rounds in the last 2 drafts.

[dudacek]

I’ve never considered Owen Power irresponsible defensively.

 

[LGR4GM]

29 minutes ago, GASabresIUFAN said:

I would hope so, but I haven’t seen that reflected in who they are drafting, especially early in the draft.  Getting Novikov and Komarov seem like intelligent late rd picks, but I’d have liked to see at least one defensively responsible D taken in the top 3 rounds in the last 2 drafts.

Owen Power

 

Not sure how you can see Komarov and Novikov as good picks who are defensively responsible and then be mad they were not in the top 3 rounds with the next sentence. Who cares where they get them as long as they get them. 

Edited August 11, 2022 by LGR4GM

[GASabresIUFAN]

4 minutes ago, LGR4GM said:

Owen Power

 

Not sure how you can see Komarov and Novikov as good picks who are defensively responsible and then be mad they were not in the top 3 rounds with the next sentence. Who cares where they get them as long as they get them. 

The history of late round picks of making the NHL is not great to say the least.  I’m happy they are prospects but I would have liked to seen a commitment to a better defensive prospect earlier in the draft that has a higher chance of contributing.  Yes I accept that we have a young D group with 3 really solid kids, but that’s only 50% of the group and I don’t really see any internal help for them for years to come, especially when Johnson bolts for allegedly greener pastures.

[PerreaultForever]

That's great, as long as they combine all that with a look at the player's character, his attitude, his compete level. 

Do they have stats that differentiate how they play in an easy game or regular game vs. say when they are losing and need extra effort? You know what I mean, the kind of thing actual scouts see. You need to combine that with the metrics for the true measure of the man to be. 

[LGR4GM]

6 hours ago, GASabresIUFAN said:

The history of late round picks of making the NHL is not great to say the least.  I’m happy they are prospects but I would have liked to seen a commitment to a better defensive prospect earlier in the draft that has a higher chance of contributing.  Yes I accept that we have a young D group with 3 really solid kids, but that’s only 50% of the group and I don’t really see any internal help for them for years to come, especially when Johnson bolts for allegedly greener pastures.

Okay, we disagree.

6 hours ago, PerreaultForever said:

That's great, as long as they combine all that with a look at the player's character, his attitude, his compete level. 

Do they have stats that differentiate how they play in an easy game or regular game vs. say when they are losing and need extra effort? You know what I mean, the kind of thing actual scouts see. You need to combine that with the metrics for the true measure of the man to be. 

Nope. Ventura is clearly too stupid to look at score effects on player stats. You should email him about it. 

 

[MBD]

10 hours ago, GASabresIUFAN said:

It all sounds great for scoring forwards, but what about preventing goals?  

You do the opposite.

[Brawndo]

12 hours ago, LGR4GM said:

https://theathletic.com/3475314/2022/08/11/sabres-analytics-sam-ventura-nhl-draft/?source=emp_shared_article

This is a fascinating read and something if you are curious about the draft you should definitely take the time to go through. 

[PerreaultForever]

4 hours ago, LGR4GM said:

Okay, we disagree.

Nope. Ventura is clearly too stupid to look at score effects on player stats. You should email him about it. 

 

Just can't help yourself can you? 

[LGR4GM]

7 hours ago, PerreaultForever said:

Just can't help yourself can you? 

I feel like the article covered what you are referring to and I have heard Forton and Adams discuss similar things. 

I get the impression some of you didn't read the article based on comments. 

[LGR4GM]

10 hours ago, Brawndo said:

You ever think about how crappy his lawyer was? Dude gets jumped in a bar parking lot by 3 guys (who clearly want to do awful things to his wife) and it was considered unreasonable force to toast one of them. Also now I want to watch Con-Air

[LGR4GM]

54 minutes ago, LGR4GM said:

I feel like the article covered what you are referring to and I have heard Forton and Adams discuss similar things. 

I get the impression some of you didn't read the article based on comments. 

Okay, because I want to answer this and not be flippant. Scouching looks at something called defense zone transitions which is the opposite of offensive zone transition. So instead of did you carrying the puck in with control or pass it in with control, did you stop that from happening. Jiri Kulich AND Matt Savoie are both right around 10th in the dataset with 13.61% and 13.64% DZT% or defensive zone transition percentage meaning that on 13.6% of entries, these are the guys being involved in play disruption which is pretty good for a forward. Only about 32% of the time did Kulich allow the entry which is really solid for a kid playing in a men's league (or any forward). 

Basically what you can do is invert zone entries. If you prevent entries v. if you create entries. Since we entries are tracked you can also track the inverse and teams are becoming more cognizant of players that stop entries from occurring. Owen Power is actually really good at stopping entries because he takes up a ton of room, is hard to turn around, and he has a very good active stick and gap. 

What you want is a forward who is involved in the D zone entries (aka he is the player defending) and who doesn't allow those entries as much as possible (Kulich denied about 68% of the entries he was involved with). 

I do believe Karmanos would include such things in his statistical profile and I do believe the scouts would double check to see how those breakups look in real time. Stats and Eyes combining. Both Kulich, Savoie, and Östlund are involved defensively so we can only hope that part of their games continue to grow. 

[dudacek]

9 minutes ago, LGR4GM said:

. Owen Power is actually really good at stopping entries because he takes up a ton of room, is hard to turn around, and he has a very good active stick and gap. . 

This is what I always thought gets missed in all those “size” debates that break out from time to time.

Range is a thing, and reach can be a very effective tool in creating that.

[LGR4GM]

22 minutes ago, dudacek said:

This is what I always thought gets missed in all those “size” debates that break out from time to time.

Range is a thing, and reach can be a very effective tool in creating that.

Yes but the other end of that is drafting Maverick Lamereoux in the first because he's big. Size can't make up for everything, just like lack of it can't be made up for with skill. 

[PerreaultForever]

8 hours ago, LGR4GM said:

I feel like the article covered what you are referring to and I have heard Forton and Adams discuss similar things. 

I get the impression some of you didn't read the article based on comments. 

The Athletic doesn't let you read the links. How many times has this come up, like a million? We don't all subscribe to it. Feel free to post the relevant quotes.

7 hours ago, LGR4GM said:

Yes but the other end of that is drafting Maverick Lamereoux in the first because he's big. Size can't make up for everything, just like lack of it can't be made up for with skill. 

I remember people saying Chara wouldn't be any good. So there's that. 

[LGR4GM]

5 hours ago, PerreaultForever said:

The Athletic doesn't let you read the links. How many times has this come up, like a million? We don't all subscribe to it. Feel free to post the relevant quotes.

I remember people saying Chara wouldn't be any good. So there's that. 

Only a fool thinks every player with a passing resemblance to an exception is also an exception. 

[PerreaultForever]

40 minutes ago, LGR4GM said:

Only a fool thinks every player with a passing resemblance to an exception is also an exception. 

and only a fool will keep arguing even when wrong. over and out.