Toggle sidebar

OHL Team adoption of Analytics?

sfan

Registered User
Jun 26, 2013
573
0
Ottawa
What teams have adopted advanced hockey analytics as part of their player development and scouting?

The Soo were famously early adopters. @Zac_Urback of Mississauga is public. Who else?

The only other indicator I have is occasionally seeing evidence of scratched players high in the stands doing time-on-ice data capture (tablet app) and video event data tagging (PC connected to the OHL video feed). Unfortunately, it hasn't occurred to me to keep track of what teams these have been.

I'd like to give thanks and a shout out to @3hayden3 for the analytics data he curates here: prospect-stats.com
Click to read more...
 
Torts

Torts

Registered User
Aug 21, 2009
2,686
318
Ontario
It wouldn't surprise me if all teams have been at least starting to in recent years with the increased popularity and effectiveness/results.
 
PrioritySelection

PrioritySelection

Hi. I watch hockey.
Dec 22, 2015
101
1
www.priorityselection.ca
There are a lot of OHL teams that have adopted analytics in some capacity; some more publicly than others.

As the OP mentioned, www.prospect-stats.com is an excellent resource for fans that are interested in combing through data that goes beyond Points, Shots, and Plus/Minus. The age adjustment data is particularly interesting, but there's a lot of quality information that can be found there.

One problem that teams face in the OHL is that the development differences are so stark between players who are starting out as 16 or 17-year-old's, and those who are more established at 18, 19, or even 20. The age adjustment goes a long way towards contextualizing these discrepencies but it does add a bit of a grey area when you consider the application of analytics in a league like the NHL, where by and large the players have reached the peak of their development.

Another area of particular interest, IMO, is draft analytics as picks are incredibly valuable in junior hockey given the number of 16-year-old's that make teams immediately after being drafted. Teams that are able to assess the value of picks gives them a bit of a competitive advantage when it comes to maximizing value in trades throughout the season to ensure they're getting equitable returns.

Definitely an interesting topic to discuss, though!
 

sfan

Registered User
Jun 26, 2013
573
0
Ottawa
PrioritySelection said:
There are a lot of OHL teams that have adopted analytics in some capacity; some more publicly than others.

As the OP mentioned, www.prospect-stats.com is an excellent resource for fans that are interested in combing through data that goes beyond Points, Shots, and Plus/Minus. The age adjustment data is particularly interesting, but there's a lot of quality information that can be found there.

One problem that teams face in the OHL is that the development differences are so stark between players who are starting out as 16 or 17-year-old's, and those who are more established at 18, 19, or even 20. The age adjustment goes a long way towards contextualizing these discrepencies but it does add a bit of a grey area when you consider the application of analytics in a league like the NHL, where by and large the players have reached the peak of their development.

Another area of particular interest, IMO, is draft analytics as picks are incredibly valuable in junior hockey given the number of 16-year-old's that make teams immediately after being drafted. Teams that are able to assess the value of picks gives them a bit of a competitive advantage when it comes to maximizing value in trades throughout the season to ensure they're getting equitable returns.

Definitely an interesting topic to discuss, though!
Click to expand...

Yes, Prospect-stats is a great resource for interested fans, and perhaps for scouting, but it is not a substitute for a team properly leveraging analytics to maximize either team performance or player development.

For this, it seems to me, any credible OHL hockey operations needs to start by gathering its own TOI and event data. No scraped-data analytics site has this. For example, Prospect-stats (I believe) estimates TOI is estimated based only on OHL gathered events like shots, +/-, goals, assists (ES and PP) and face-offs. A season may be good enough sample size but eTOI is very unreliable for performance assessment as the season starts and unfolds.

Besides the challenge of comparing younger vs older players, there are also the challenge of comparing player performance corrected for linemates, correcting performance based on the quality of competition, zone starts, etc.

A few years ago the 67s were very public about hiring Matt Pfeffer to do analytics but he's now with the Canadians and it is not clear to me, either from their hockey ops staff, or my casual observations at the rink, that they currently leverage analytics. (Maybe I am wrong about this, these are just my inferences.)

My assumption is that at least some teams remain just old-school, while other teams systematically leverage analytics. It would be very interesting to know more about the teams that have adopted advanced analytics, what they are doing, and how it is working out for them.
 

Mayday3*

Guest
Windsor tweeted out their corsi rating after every single period this season so obviously they're tracking that
 

sfan

Registered User
Jun 26, 2013
573
0
Ottawa
Draft list of analytics use in OHL

Here is a list of what has been found or observed/reported so far. Please post if you have any additional information and insights and I will update.

Eastern
Barrie
  • Unknown
Hamilton
Kingston
Mississauga
  • (2016) Zac Urback, Coordinator of Hockey Analytics at Mississauga Steelheads
Niagara
  • (2014) Using Pointstreak's Performance analytics platform
North Bay

  • (2014) Using Pointstreak's Performance analytics platform
Oshawa
  • Unknown
Ottawa
  • (2013-14) Misha Donskov and Matt Pfeffer
  • (2015-16) Unknown
Peterborough
Sudbury

Western
Erie

  • (2014) Using Pointstreak's Performance analytics platform
Flint
  • Unknown
Guelph
  • Unknown
Kitchener
London
  • (2010-13) Misha Donskov started London's analytics program
Owen Sound

  • (2014) Using Pointstreak's Performance analytics platform
Saginaw
[*] Unknown
Sarnia
Sault St Marie
  • (2010-14) Kyle Dubas pioneered analytics in OHL
Windsor
  • (2014) Using Pointstreak's Performance analytics platform
  • (2015-16) Tweet CORSI stats per game
 
Last edited:
Ward Cornell

Ward Cornell

Registered User
Dec 22, 2007
6,396
2,620
Analytics is a great tool for almost every sport but what a fuss it create when used in policing! ;)
 

Big Punisher*

Registered User
May 23, 2014
82
1
Torts said:
It wouldn't surprise me if all teams have been at least starting to in recent years with the increased popularity and effectiveness/results.
Click to expand...

What effectiveness? Every year we have plenty of examples that prove these stats are flawed. And every time I hear "(something) adjusted" I have to laugh a little. I guess when the stat doesn't give you the answer you want, just adjust it a bit

While teams are "tracking numbers", those numbers are based off real stats and the team's scouting evaluations. This is a little bigger in the pro ranks where salary cap implications are involved. Hockey is not baseball. A player's numbers will change dramatically based on the players he plays with, or the caliber of team he's on.
 

sfan

Registered User
Jun 26, 2013
573
0
Ottawa
Big Punisher said:
What effectiveness? Every year we have plenty of examples that prove these stats are flawed. And every time I hear "(something) adjusted" I have to laugh a little. I guess when the stat doesn't give you the answer you want, just adjust it a bit

While teams are "tracking numbers", those numbers are based off real stats and the team's scouting evaluations. This is a little bigger in the pro ranks where salary cap implications are involved. Hockey is not baseball. A player's numbers will change dramatically based on the players he plays with, or the caliber of team he's on.
Click to expand...

I'm not religious about stats, but I see them as an essential tool to be used in combination with the best observations and judgement of the coaching staff. I believe that willful dismissal of them is a sign of a a coaching mindset that is closed to reason, ignorant of bias, more concerned about alpha opinion than doing everything possible to build team and player competitiveness.

The key stats that I care about, and the OHL doesn't seem to care to provide on behalf of the teams and players, are:
- Who is on the ice when? This covers all players on both teams for every second of play.
- What relevant events take place, when and with whom? Zone, zone entry and exit, turnover. pass made/received, turnover, shot position. dangerous shots, blocks, hits, puck battles, etc...
- All of the above in a readily analyzable spreadsheet or database format, integrated with OHL's traditional stats, and synchronized with game video.

How is this information not relevant? It is essentially just tracking the building blocks of the game, captured and evaluated game by game as the season unfolds. Without this then coaching decisions and player development is inherently subjective and therefore sub-optimal.
 
bobber

bobber

Registered User
Jan 21, 2013
8,628
6,412
Kitchener Ontario
A lot of teams use these stats but in most cases it comes down to having the top skilled players. They just score more goals on average and if you have them ala London and Erie you win.
 

sfan

Registered User
Jun 26, 2013
573
0
Ottawa
bobber said:
A lot of teams use these stats but in most cases it comes down to having the top skilled players. They just score more goals on average and if you have them ala London and Erie you win.
Click to expand...

Sure, I understand that. And a really good goalie can cover many sins and steal games too. So what do you do with the rest of the players on those teams that are not (yet) "top skilled'? How do you maximize the effectiveness of their utilization and development? What do you do to make the most of a team roster that isn't blessed in any given year?
 
bobber

bobber

Registered User
Jan 21, 2013
8,628
6,412
Kitchener Ontario
sfan said:
Sure, I understand that. And a really good goalie can cover many sins and steal games too. So what do you do with the rest of the players on those teams that are not (yet) "top skilled'? How do you maximize the effectiveness of their utilization and development? What do you do to make the most of a team roster that isn't blessed in any given year?
Click to expand...

I think the analytics are a good idea. Helps teams put the right kind of players in their systems.
 

sfan

Registered User
Jun 26, 2013
573
0
Ottawa
I agree with a lot of what you say but you haven't really made an argument against advanced stats. I am not trying to convince you, I don't really care and that is not what I am trying to explore in this thread. What I hope to learn here is simply what OHL teams have decided to invest in analytics.

I tend to believe the sport of hockey produces stats, not that stats produce hockey. I get a chuckle when people will point at the made up stats of good teams, and think that having good corsi makes them a good team, instead of being a good team means they have good corsi
Click to expand...

I agree.

The game is far too fast and dynamic for the calculation of stats to really matter to the degree some want to believe it is. Changing the spot a player is on the ice by as little as 5' can make a huge difference in the offensive chance he can create. I don't believe you can say the same in other sports. Other sports are also based on percentages, but hockey isn't quite the same. In football, if a team is 3rd and long, it's more likely to call a passing play. If it's 2nd and short, expect a running play. Lets not even start with baseball and pitch selection. A player has a split second to make a decision on where he will pass or shoot, and a lot of that is based on what his teammates are doing, and what the other team is giving him. Other sports have time to set up plays and dictate to the defence what they are going to do
Click to expand...

Of course the hockey is incredibly fast and dynamic. Now think of an entire season of fast and dynamic. Now convince anyone that your second and third line has been optimized, both in composition and utilization. What specifically has improved or deterioriated over the last ten games on those lines leveled for the quality of competition they faced and the game situation characteristics of their ice time?

Hockey (and basketball) are two fast and dynamic team games. I agree baseball and football aren't so they are irrelevant to this discussion.

A good coach will know how his players are doing, and doesn't need a excel sheet to tell him.
Click to expand...

Sure. And all coaches have perfect memory and are mental mathematical geniuses.

Baseball is a boring sport that needs something to keep you interested. I don't believe hockey is the same way, and if you'd rather count shot attempts than enjoy the game, I don't think you appreciate the greatness of the sport
Click to expand...

Again this isn't about baseball.

I remember when Winnik went from the Ducks to the Leafs, and people didn't understand why his possession numbers changed. Like the talent level of your teammates has no barring on the play.
Click to expand...

Understanding a players performance relative to linemates (among other variables) is a core use of stats.

I use to coach a kid (defenceman) who had a great snap shot for a midget player. Easily one of the hardest I have ever seen at that age. But you know what, couldn't hit the net, let along keep it below the cross bar. Didn't matter about his shot attempts or misses if that was the result. No one was going to stand in front of it, and zero chance at it going in, being deflected, or creating rebound chances. But his corsi stats would have been awesome. Same year had another kid who had a very average shot as far as velocity, but had a knack of getting it on net, and creating opportunities for deflections and rebounds. That's why I don't subscribe to some of these stats. Some players have a knack for finding the net, or for getting their stick on loose pucks. Some find ways to win (Toews), while others come up short (Ovechkin). I just believe the sport is about watching what a player does, and when. Rather than tracking it.
Click to expand...

I'm not a huge fan of corsi or fenwick and certainly using any one stat in isolation is meaningless.

I agree, advanced stats is exactly about "watching what players do and when". Except it is for every player, every event, in context, with complete recall and the ability to find insight that is otherwise lost in the fog of even the best human minds. It doesn't negate or replace a coach's understanding of the game or a player, it complements it.
 
Hammer9001

Hammer9001

Registered User
Apr 1, 2015
848
436
Hamilton
Big Punisher said:
I tend to believe the sport of hockey produces stats, not that stats produce hockey. I get a chuckle when people will point at the made up stats of good teams, and think that having good corsi makes them a good team, instead of being a good team means they have good corsi(ignore the fact this "stat" is flawed and was created by a goalie to try and make goalies look better)

The game is far too fast and dynamic for the calculation of stats to really matter to the degree some want to believe it is. Changing the spot a player is on the ice by as little as 5' can make a huge difference in the offensive chance he can create. I don't believe you can say the same in other sports. Other sports are also based on percentages, but hockey isn't quite the same. In football, if a team is 3rd and long, it's more likely to call a passing play. If it's 2nd and short, expect a running play. Lets not even start with baseball and pitch selection. A player has a split second to make a decision on where he will pass or shoot, and a lot of that is based on what his teammates are doing, and what the other team is giving him. Other sports have time to set up plays and dictate to the defence what they are going to do

A good coach will know how his players are doing, and doesn't need a excel sheet to tell him. Baseball is a boring sport that needs something to keep you interested. I don't believe hockey is the same way, and if you'd rather count shot attempts than enjoy the game, I don't think you appreciate the greatness of the sport

I remember when Winnik went from the Ducks to the Leafs, and people didn't understand why his possession numbers changed. Like the talent level of your teammates has no barring on the play.

I use to coach a kid (defenceman) who had a great snap shot for a midget player. Easily one of the hardest I have ever seen at that age. But you know what, couldn't hit the net, let along keep it below the cross bar. Didn't matter about his shot attempts or misses if that was the result. No one was going to stand in front of it, and zero chance at it going in, being deflected, or creating rebound chances. But his corsi stats would have been awesome. Same year had another kid who had a very average shot as far as velocity, but had a knack of getting it on net, and creating opportunities for deflections and rebounds. That's why I don't subscribe to some of these stats. Some players have a knack for finding the net, or for getting their stick on loose pucks. Some find ways to win (Toews), while others come up short (Ovechkin). I just believe the sport is about watching what a player does, and when. Rather than tracking it
Click to expand...

I tend to agree, save that I feel that it is worthwhile to go through the numbers, literally, so you can focus on where a player is struggling. It helps to also back away from the team's successes and failures and look at the data in a bit more of a raw form. I feel it is another tool that should be used, but shouldn't be your primary tool, especially at the Jr. level. The downside is that it can lead to second guessing of the coaching staff and management looks at the numbers and decides maybe a key puck grinder just isn't scoring enough, etc etc.
 
You must log in or register to reply here.

Ad

Latest posts

Upcoming events

Ad

Ad

Top Bottom