2024-25 Roster Thread #2: Midseasonnar

[BiggE]

Zev Buium projected to be the #1 D on the US World Junior squad.

But hey, we got a future middle 6 forward.

 

[JojoTheWhale]

Based entirely on feel from what I have watched this season it seems like teams are working harder to maintain control until they can produce high danger shots. It seems like there is a lot less "getting pucks on net" then there was say 5-10 years ago.

There is quite a bit of data to support this. There was a shift 2-3 years ago and it’s only accelerating.

 

[BiggE]

There is quite a bit of data to support this. There was a shift 2-3 years ago and it’s only accelerating.

Except for the Flyers, they still think it’s 1984

 

[deadhead]

This change in strategy has increased the value of PFs and big D-men.

It's one thing to hang on on the edge of the High Danger zone and take better shots, it's quite another to park in front of the net and screen the goalie and pounce on rebounds.

If you look at the Flyers heat maps, they struggle at both ends of the ice, both creating traffic in front of the net and clearing the porch.

Which is why you should always account for the "Red Queen effect," you can see this in the NFL where the shift to smaller, quicker defenses has revived the power running game. In the NHL, it is why size will still matter, not your whole roster, but a strategic group of players that allow you to play a "power" game to complement quick, skilled players.

 

[VladDrag]

There is quite a bit of data to support this. There was a shift 2-3 years ago and it’s only accelerating.

Did a quick analysis myself :

A few notes:
It's only 5v5 data.
For 19-20 season, I used the first 68 games so it was an easier equation in excel
For the most 24-25, I only used first 19 games for the same reason.
Shot Danger -- xG/FF*100 -- gives you a snapshot of how dangerous each shot is. It's something I made up, but it makes sense to me.
Data is from NST

Year	CF	FF	xG	Corsi per Team per game	Shot Danger
10-11	105709	78713	4383.13	42.97	5.57
11-12	105549	78393	4360.12	42.91	5.56
2013	61581	45426	2535.44	42.76	5.58
13-14	106050	78906	4340.55	43.11	5.50
14-15	107256	79176	4386.05	43.60	5.54
15-16	106114	78732	4372.53	43.14	5.55
16-17	109124	81095	4474.75	44.36	5.52
17-18	117750	87685	4891.67	46.32	5.58
18-19	117188	87482	4959.3	46.10	5.67
19-20	94616	71336	4044.69	44.88	5.67
20-21	74572	56311	3183.65	42.96	5.65
21-22	117826	89129	5307.31	44.90	5.95
22-23	119786	88764	5584.94	45.65	6.29
23-24	125473	89353	5418.26	47.82	6.06
24-25	28640	20256	1216.1	47.11	6.00

You see a pretty large increase in shot attempts over the last few seasons, and the shots are certainly more dangerous over the last 3-4 seasons.

This is why I've prioritized players like Brink, Frost and to a lesser extent Foerster. These are guys who can cycle the puck in the zone better than others.

 

[dragonoffrost]

Except for the Flyers, they still think it’s 1974

FTFY

 

[JojoTheWhale]

This change in strategy has increased the value of PFs and big D-men.

It's one thing to hang on on the edge of the High Danger zone and take better shots, it's quite another to park in front of the net and screen the goalie and pounce on rebounds.

I am not aware of any recorded increases in Rebound rates.

 

[deadhead]

Did a quick analysis myself :

A few notes:
It's only 5v5 data.
For 19-20 season, I used the first 68 games so it was an easier equation in excel
For the most 24-25, I only used first 19 games for the same reason.
Shot Danger -- xG/FF*100 -- gives you a snapshot of how dangerous each shot is. It's something I made up, but it makes sense to me.
Data is from NST

Year	CF	FF	xG	Corsi per Team per game	Shot Danger
10-11	105709	78713	4383.13	42.97	5.57
11-12	105549	78393	4360.12	42.91	5.56
2013	61581	45426	2535.44	42.76	5.58
13-14	106050	78906	4340.55	43.11	5.50
14-15	107256	79176	4386.05	43.60	5.54
15-16	106114	78732	4372.53	43.14	5.55
16-17	109124	81095	4474.75	44.36	5.52
17-18	117750	87685	4891.67	46.32	5.58
18-19	117188	87482	4959.3	46.10	5.67
19-20	94616	71336	4044.69	44.88	5.67
20-21	74572	56311	3183.65	42.96	5.65
21-22	117826	89129	5307.31	44.90	5.95
22-23	119786	88764	5584.94	45.65	6.29
23-24	125473	89353	5418.26	47.82	6.06
24-25	28640	20256	1216.1	47.11	6.00

You see a pretty large increase in shot attempts over the last few seasons, and the shots are certainly more dangerous over the last 3-4 seasons.

This is why I've prioritized players like Brink, Frost and to a lesser extent Foerster. These are guys who can cycle the puck in the zone better than others.

I don't see Brink and Frost as good at cycling the puck, they're good passers, but cycling the puck requires winning board battles on a regular basis to maintain possession. They need to play with linemates who can do the dirty work and get them the puck so they can make plays.

Laughton - Poehling - Hathaway are probably the best at this, but only Laughton can finish on a regular basis.

Foerster was good at this last season, but has regressed a bit.

The other aspect of a successful cycle are forwards who have a good feel for the "rotation," knowing when to fall back and cover a D-man pinching to keep the puck in the O-zone.

 

[deadhead]

I am not aware of any recorded increases in Rebound rates.

Don't know, and don't know if that would capture the effect unless you map the rebounds.

Heat maps might, if more shots are coming within 5 feet or less, probably players getting better at turning rebounds into shots.

If the increase in high danger shots were coming from 5-15', then it's more likely to be the result of getting pucks to players in the high danger areas.

In either case, you would want to know the source of the shot, from a pass, rebound, recovery of the puck from a blocked shot, etc. It would be interesting to see shot release data (i.e. something akin to pop up data for baseball catchers, how quickly shot is released once shooting motion is initiated).

The other data that would be interesting would be akin to the QB in the pocket, and separation for WRs - how much time between a forward receiving a puck in various spots before he's challenged by a defender. Which forwards make plays quickly, which teams are quick to challenge potential shooters in the HD areas.

I suspect this is the kind of data that tends to be proprietary, since it would require coding each shot (off video), and each player shooting - which is time consuming and therefore expensive. I wonder how much of an analytics budget goes to actual analysis, and how much is spent on gathering raw data.

 

[JojoTheWhale]

Don't know, and don't know if that would capture the effect unless you map the rebounds.

Heat maps might, if more shots are coming within 5 feet or less, probably players getting better at turning rebounds into shots.

If the increase in high danger shots were coming from 5-15', then it's more likely to be the result of getting pucks to players in the high danger areas.

In either case, you would want to know the source of the shot, from a pass, rebound, recovery of the puck from a blocked shot, etc. It would be interesting to see shot release data (i.e. something akin to pop up data for baseball catchers, how quickly shot is released once shooting motion is initiated).

The other data that would be interesting would be akin to the QB in the pocket, and separation for WRs - how much time between a forward receiving a puck in various spots before he's challenged by a defender. Which forwards make plays quickly, which teams are quick to challenge potential shooters in the HD areas.

I suspect this is the kind of data that tends to be proprietary, since it would require coding each shot (off video), and each player shooting - which is time consuming and therefore expensive. I wonder how much of an analytics budget goes to actual analysis, and how much is spent on gathering raw data.

Well, you were positing that this manifests in netfront players. That means the burden of proof is on you. I’m just saying I have never seen data that supports it, which is of course not the same thing as proving it false.

The way Rebounds are currently defined I don’t believe has anything to do with distance. It’s just a low time period between Shots without a break in play.

 

[Ghosts Beer]

I would imagine the league-wide reduction in save-percentages has much to do with equipment changes.

Most particularly, the sticks these days are so ridiculously light and powerful. It's like if you let MLBers switch to non-wood bats.

 

[deadhead]

I would imagine the league-wide reduction in save-percentages has much to do with equipment changes.

Most particularly, the sticks these days are so ridiculously light and powerful. It's like if you let MLBers switch to non-wood bats.

I'd think this would show up in shot velocity (though the data doesn't seem to distinguish between slapshots and wrist shots). And that would be reflected in higher Sh% rates outside the high danger areas (i.e. higher velocity shots from longer distances beating goaltenders).

It might also show up as quicker releases if "better" sticks allowed you to garner velocity with a more compact shooting motion. Watching games, a quick release is more deceptive (and makes it easier to use a defender as a screen) and thus should be effective at lower velocity (the key is not how fast a puck is shot, but how much time the goalie has between recognizing a shot and it arriving at the goal).

Again, probably requires detailed and thus proprietary data.

 

[VladDrag]

I don't see Brink and Frost as good at cycling the puck, they're good passers, but cycling the puck requires winning board battles on a regular basis to maintain possession. They need to play with linemates who can do the dirty work and get them the puck so they can make plays.

Laughton - Poehling - Hathaway are probably the best at this, but only Laughton can finish on a regular basis.

Foerster was good at this last season, but has regressed a bit.

The other aspect of a successful cycle are forwards who have a good feel for the "rotation," knowing when to fall back and cover a D-man pinching to keep the puck in the O-zone.

I think it's important to explain what I believe is cycling the puck.

When a team cycles the puck they already have possession = they don't need to engage in board battles when cycling. They already have 100% possession. Why would you put the puck in a 50/50% change to gain or loose the puck? Doesn't make sense. A well cycling team moves the puck around defenders so that they don't loose possession. The point is being able to possess the puck. Viewing board battles as a requirement to be a good cycling team is somewhat antiquated.

Now, that doesn't mean there is no value in board battles. Dump and chase is still viable, applying forechecking pressure to turn the puck over is important. Board battles are a means to gaining possession so you can move the team can move to the cycle. But you can be good in the NZ and enter the zone with possession and either get a rush scoring chance or go to the cycle. I think the prime example of this is when McKinnon, Toews and Makar were on the ice for the Avalanche. That's supreme cycling at it's best -- they didn't have to do any board battling. Just pass it around defenders.

 

[Cody Webster]

I think it's important to explain what I believe is cycling the puck.

When a team cycles the puck they already have possession = they don't need to engage in board battles when cycling. They already have 100% possession. Why would you put the puck in a 50/50% change to gain or loose the puck? Doesn't make sense. A well cycling team moves the puck around defenders so that they don't loose possession. The point is being able to possess the puck. Viewing board battles as a requirement to be a good cycling team is somewhat antiquated.

Now, that doesn't mean there is no value in board battles. Dump and chase is still viable, applying forechecking pressure to turn the puck over is important. Board battles are a means to gaining possession so you can move the team can move to the cycle. But you can be good in the NZ and enter the zone with possession and either get a rush scoring chance or go to the cycle. I think the prime example of this is when McKinnon, Toews and Makar were on the ice for the Avalanche. That's supreme cycling at it's best -- they didn't have to do any board battling. Just pass it around defenders.

We're going to cycle the shit out of the puck tonight!

 

[VladDrag]

We're going to cycle the shit out of the puck tonight!

3 full lines tonight

 

[Cody Webster]

3 full lines tonight

I know, so sad

 

[VladDrag]

Don't know, and don't know if that would capture the effect unless you map the rebounds.

Heat maps might, if more shots are coming within 5 feet or less, probably players getting better at turning rebounds into shots.

If the increase in high danger shots were coming from 5-15', then it's more likely to be the result of getting pucks to players in the high danger areas.

In either case, you would want to know the source of the shot, from a pass, rebound, recovery of the puck from a blocked shot, etc. It would be interesting to see shot release data (i.e. something akin to pop up data for baseball catchers, how quickly shot is released once shooting motion is initiated).

The other data that would be interesting would be akin to the QB in the pocket, and separation for WRs - how much time between a forward receiving a puck in various spots before he's challenged by a defender. Which forwards make plays quickly, which teams are quick to challenge potential shooters in the HD areas.

I suspect this is the kind of data that tends to be proprietary, since it would require coding each shot (off video), and each player shooting - which is time consuming and therefore expensive. I wonder how much of an analytics budget goes to actual analysis, and how much is spent on gathering raw data.

Did another quick one for you. Just a few notes:

IT's 5v5 data. It's from NST. NST defines a rebound as a corsi attempt that is taken within 3 seconds after a previous corsi attempt. IT does not specify 'rebound' the way we think of one.

	iCF	iFF	iSCF	iHDCF	Rebounds Created	% of total CF that are Rebounds
18-19	117185​	87480​	54388​	22500​	6875​	5.87%​
19-20	97091​	73216​	45787​	18484​	5907​	6.08%​
20-21	74461​	56230​	35478​	14028​	4346​	5.84%​
21-22	117819​	89124​	58697​	23545​	7212​	6.12%​
22-23	119783​	88762​	60202​	25117​	9675​	8.08%​
23-24	125472​	89352​	57553​	23475​	11352​	9.05%​
24-25	31366​	22162​	14140​	5657​	2877​	9.17%​

Rebounds increase xG for sure. Unfortunately, this specific dataset doesn't really answer your question tho. I'd imagine that one of the sites somewhere may let you filter shots by rush vs rebound and see where the heat maps are for them. I can look into that in a bit.