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

[Jim Bob]

How Sam Ventura pushed the analytics team to help secure the Sabres' 2022 Draft haul

With so many picks in the 2022 NHL Draft, the Sabres turned their focus to amateur scouting with major influence from analytics.

theathletic.com

When the Sabres hired Sam Ventura as their vice president of hockey strategy and research in July of 2021, he had just a few weeks to acclimate himself to the organization before the NHL Draft. He came supplied with opinions on the prospects in the draft but hadn’t yet built out his staff, his statistical systems or the relationships with the Sabres’ front office. He played a role in that draft class, but it was nothing compared to the plans he had for what the team could do in the future.

Heading into 2022, Ventura had a lot of ideas about how to revamp Buffalo’s use of analytics in the draft process. He hired Domenic Galamini Jr. as a data scientist and Matt Barlowe as a data engineer. With that team in place, Ventura got to work building data systems with the draft in mind. He and the organization knew that they would have plenty of draft capital to work with, so Ventura and his staff made that their priority.

“We were able to direct something like 90 percent of our efforts toward projects that were directly applicable to the draft,” Ventura said. “We knew going into the year with the draft capital that we had, obviously the three first-round picks, and then 11 picks total by the end of the year, we knew draft research was going to be where we needed to spend most of our time for most of the season.

“I pushed those guys pretty hard throughout the year. We were able to build draft analysis tools that I would say match or exceeded anything I’ve ever been a part of.”

That whole article is a must read for Sabres fans like us here. I also think it is a great read for any hockey fan that is into an analytical approach to team building.

If this is the type of coverage we get from Fairburn moving forward, I am really excited.

Click to read more...

 

[RefsIdeas]

Very good read! Kind of makes statements made by NHLe people (ie. Bader) even more laughable when they'll criticize a Sabres draft pick. The data that Ventura has available to him is miles better than what Bader would have.

 

[Jim Bob]

Very good read! Kind of makes statements made by NHLe people (ie. Bader) even more laughable when they'll criticize a Sabres draft pick. The data that Ventura has available to him is miles better than what Bader would have.

What Bader does and what Ventura does are two completely different things.

Ventura not only has wider access to a variety of stats to help build models, he is also working for a team with a scouting staff.

I think the most interesting thing is how both Ventura and his staff and Forton and his staff are working well together. Too often it can devolve into scouting vs analytics and not scouting + analytics. Forton's post-draft press conference was one glimpse into how well both groups are working well together. This article was another glimpse into it.

I wonder how much Ventura and crew will be able to influence the pro scouting department to help gather intel to identify trade and FA targets.

 

[RefsIdeas]

What Bader does and what Ventura does are two completely different things.

Ventura not only has wider access to a variety of stats to help build models, he is also working for a team with a scouting staff.

I think the most interesting thing is how both Ventura and his staff and Forton and his staff are working well together. Too often it can devolve into scouting vs analytics and not scouting + analytics. Forton's post-draft press conference was one glimpse into how well both groups are working well together. This article was another glimpse into it.

I wonder how much Ventura and crew will be able to influence the pro scouting department to help gather intel to identify trade and FA targets.

Oh, I know. Bader will present his information as if a pick was "good" or "bad" though - seemingly from an analytical perspective. But it's so much deeper than the numbers he has access to.

 

[Chainshot]

What Bader does and what Ventura does are two completely different things.

Ventura not only has wider access to a variety of stats to help build models, he is also working for a team with a scouting staff.

I think the most interesting thing is how both Ventura and his staff and Forton and his staff are working well together. Too often it can devolve into scouting vs analytics and not scouting + analytics. Forton's post-draft press conference was one glimpse into how well both groups are working well together. This article was another glimpse into it.

I wonder how much Ventura and crew will be able to influence the pro scouting department to help gather intel to identify trade and FA targets.

Sounds like they have solid integration which is truly refreshing with the way this team has fumbled its way around for so long. Forton has expressed that they can use the data to direct who they should be getting more views on and then also when the data and the eyeballs don't match, digging deeper. It isn't just one or the other.

As for your last point, it sounds like that is what is next.

Oh, I know. Bader will present his information as if a pick was "good" or "bad" though - seemingly from an analytical perspective. But it's so much deeper than the numbers he has access to.

Ventura hinted at the model for NHLe for prospects is outdated.

 

[Jim Bob]

Oh, I know. Bader will present his information as if a pick was "good" or "bad" though - seemingly from an analytical perspective. But it's so much deeper than the numbers he has access to.

That is Bader's job as a content creator. His job is to form an opinion and share it.

Just like the wide variety of independent draft analysts, Bader has his approach to predicting who will work out long term. Of course he will use his method to form opinions on prospects.

 

[Jim Bob]

Sounds like they have solid integration which is truly refreshing with the way this team has fumbled its way around for so long. Forton has expressed that they can use the data to direct who they should be getting more views on and then also when the data and the eyeballs don't match, digging deeper. It isn't just one or the other.

The other cool thing to me was how Ventura and crew used second half data to help identify guys that pop mid-season that were not on the team's radar previously.

With this being the first draft where Ventura has had a big part in it, it will be interesting to watch if there are new trends that start to show with their draft strategy over the next handful of drafts.

 

[Chainshot]

I sense a theme today -

Inside look at Buffalo Sabres

NHL.com is providing in-depth roster, prospect and fantasy analysis for each of its 32 teams from Aug. 8-Sept. 8. Today, the Buffalo Sabres.

www.nhl.com

 

[Jim Bob]

I sense a theme today -

Inside look at Buffalo Sabres

NHL.com is providing in-depth roster, prospect and fantasy analysis for each of its 32 teams from Aug. 8-Sept. 8. Today, the Buffalo Sabres.

www.nhl.com

"There is no rush. There will be growing pains."

Dahlin with the tagline for the 2022-23 Sabres season.

 

[Archie Lee]

Not that it was ever really in doubt, but the article pretty much definitively confirms that there was never any intention to acquire veterans players this off-season through trade or higher-profile free agency. The focus of the analytics team was on building out tools for the draft. Now that is done and I get the sense the focus will switch somewhat to building out pro-scouting analytics tools.

 

[Dingo44]

I sense a theme today -

Inside look at Buffalo Sabres

NHL.com is providing in-depth roster, prospect and fantasy analysis for each of its 32 teams from Aug. 8-Sept. 8. Today, the Buffalo Sabres.

www.nhl.com

I just want to say that Donny looks SHARP here - love the color combo (including the glasses):

 

[debaser66]

I sense a theme today -

Inside look at Buffalo Sabres

NHL.com is providing in-depth roster, prospect and fantasy analysis for each of its 32 teams from Aug. 8-Sept. 8. Today, the Buffalo Sabres.

www.nhl.com

Everybody projects for less points than they had last season. thats quite a strange take, especially for Dahlin (50 points).
Except for Power were I would be happy if he would reach 44pts.

 

[Fjordy]

Everybody projects for less points than they had last season. thats quite a strange take, especially for Dahlin (50 points).
Except for Power were I would be happy if he would reach 44pts.

44 pts?

I think for Dahlin last season everything turned out to be as unsuccessful as possible, so that he scored 53 points. Our PP didn't work very well, Olofsson couldn't shoot and our attack this season should look more interesting anyway. Dahlin also roamed from Butcher to Pysyk and Jokiharju playing with them in pairs.

I am absolutely sure that Dahlin will score 60+ points next season (of course, if there is no long-term injury).

 

[jc17]

Ventura hinted at the model for NHLe for prospects is outdated.

I would say it sounded like he was saying they can get more granular now. But interestingly, later in the article they mention the work of Wu and Nandakumar as influences, which was summarized by age, league scoring, and point shares, so not much different than NHLE.

The 2nd half production comment also stood out to me. I remember seeing someone a few years ago say they looked for that in prospects, and anecdotally it did seem that guys who had big 2nd halves, did well later on. A notable prospect that had big 2nd half production in his DY was Quinn. But I imagine its a tricky balance, where by splitting the season in 2, you might have an insight to development, but you also lose half the sample size.

 

[debaser66]

44 pts?

I think for Dahlin last season everything turned out to be as unsuccessful as possible, so that he scored 53 points. Our PP didn't work very well, Olofsson couldn't shoot and our attack this season should look more interesting anyway. Dahlin also roamed from Butcher to Pysyk and Jokiharju playing with them in pairs.

I am absolutely sure that Dahlin will score 60+ points next season (of course, if there is no long-term injury).

I am not sure what those projections should be based on, they seem completely off base or random.
I also think Dahlin will have 60pts+

 

[Doug Prishpreed]

I am not sure what those projections should be based on, they seem completely off base or random.
I also think Dahlin will have 60pts+

They're likely assuming Dahlin will score less because Power will take some of his PP time.

 

[SnuggaRUDE]

Going from NHLe to point shares does obsolete NHLe. Point shares include defensive contribution.

The take away shouldn't be to crap on what people are doing with public data, it really shows how much they are capable of doing. Public obvious success of some analytics nearly shamed the teams into doing it for themselves.

Baseball had to be brought kicking and screaming into data analysis, and it's far easier to model. Attitudes changed quite abruptly. Until they did data was looked down upon as nerds in their nerderies doing nerdy nerd nerdings. (not to say it wasn't also that)

 

[Ace]

The problem with Bader is that when he criticizes a Buffalo pick he’s literally doing it from the relatively uninformed chair. We have smarter people than him building better systems than him with more information than him. And the GM uses it. If he wants to tell you how like…San Jose doesn’t get it or Philly or something…alright I’ll listen. If he wants to argue his more limited information is smarter than Sam Ventura…he’s wrong. Arrogantly so.

 

[SnuggaRUDE]

The problem with Bader is that when he criticizes a Buffalo pick he’s literally doing it from the relatively uninformed chair. We have smarter people than him building better systems than him with more information than him. And the GM uses it. If he wants to tell you how like…San Jose doesn’t get it or Philly or something…alright I’ll listen. If he wants to argue his more limited information is smarter than Sam Ventura…he’s wrong. Arrogantly so.

How could we possibly know who's right? Buffalo's Ventura informed picks have no body of work to evaluate.

 

[Ace]

How could we possibly know who's right? Buffalo's Ventura informed picks have no body of work to evaluate.

If you knew ten words in French and were speaking to a fluent French speaker and corrected their words…that’s what this is.

We know who is more informed. We know it. It’s not debatable.

 

[SnuggaRUDE]

If you knew ten words in French and were speaking to a fluent French speaker and corrected their words…that’s what this is.

We know who is more informed. We know it. It’s not debatable.

If that French speaker says one of those 10 words wrong the bulk of their knowledge doesn't make them right.

We know who has more data, we don't yet know whose models are better. Even if we suspect it's Ventura.

Your prognostication is reasonable, but you're presenting it as metaphysical truth; and beyond that as a reason to condemn someone who doesn't agree.

 

[Jim Bob]

If that French speaker says one of those 10 words wrong the bulk of their knowledge doesn't make them right.

We know who has more data, we don't yet know whose models are better. Even if we suspect it's Ventura.

Your prognostication is reasonable, but you're presenting it as metaphysical truth; and beyond that as a reason to condemn someone who doesn't agree.

Projecting who’s projection is more accurate before the results are in is something, huh.

I want Ventura and the Sabres picks to be better than what Bader did. I know I like Power more than Brandt Clarke. But, I don’t see the need to call someone arrogant for having an opinion.

But, I guess that is because I am arrogant enough to question their pick of a goalie in the 2nd round.

 

[SnuggaRUDE]

Projecting who’s projection is more accurate before the results are in is something, huh.

I want Ventura and the Sabres picks to be better than what Bader did. I know I like Power more than Brandt Clarke. But, I don’t see the need to call someone arrogant for having an opinion.

But, I guess that is because I am arrogant enough to question their pick of a goalie in the 2nd round.

Right? Even projecting projections before results would somewhat reasonable. CONCLUDING projections were more accurate before they're tested is asinine.

 

[dasaybz]

This was a great read. I wonder what those models entail? Keeping it close to the vest is interesting, don't want to give away company secrets. I guess we'll see if all this data driven stuff equates down the road in a couple years. It's refreshing there are a ton of smart guys in the front office now and they seem to all be on the same page. Kev definitely deserves credit for that.

 

[Diaspora]

Right? Even projecting projections before results would somewhat reasonable. CONCLUDING projections were more accurate before they're tested is asinine.

Can I run this by my attorney before I agree?