2024 Off-season trade rumours and speculation thread

[butterfly]

Originally Posted by ComixZone

https://twitter.com/user/status/1832438056567370090

Giordano seems to be taking too much heat in this thread. He's still an effective NHL player. He'd be an excellent target for the Oilers.

Those player cards are kind of useless. WAR is a counting stat, not a rate stat. What does 82% WAR mean mathematically? A 7th D as suggested by JFresh?

A 7th D is, by definition, replacement level.

[calculoso]

Originally Posted by Monahammer

Treliving and the coaches take more heat in retrospect, but the forums were criticizing in room leadership and gio specifically back to when he played here as a top reason for lack of success during that era.

Some examples of obviously weak in room leadership off top of my head :
-the Hartley super bowl fall out. Why was the coach policing the activities of rookies so hard and where were the vets in the room to reign the kids in before coach got pissed?
- gulutzan train beers. Why was the coach so distant from the guys in the locker room and why wasn't the captain working from the other end to bridge the gap?
- years of rumors of rifts or sides in the dressing room. Groups of players who supported each other
- edmonton series playoff collapse
- Sutter comments about sorry state of team and leadership when he comes in

You’re suggesting that the coaches need to helicopter parent their players and that every dressing room is 100% cohesive?

[Jay Random]

Originally Posted by Goriders

Which team would you like to own if the Flames make the playoffs? An expansion team?

The Rangers, of course. Forbes says they're worth $2.65 billion, second only to the Leafs, and I would give up the extra bit of money to keep the Leaf stink off me.

Then I would sell the team, change my name, buy a private island with a volcano lair, and plot to take over the world. Wouldn't you?

[Jay Random]

Originally Posted by butterfly

Those player cards are kind of useless. WAR is a counting stat, not a rate stat. What does 82% WAR mean mathematically? A 7th D as suggested by JFresh?

A 7th D is, by definition, replacement level.

The WAR percentile is how Giordano ranks compared with all other NHL defencemen. He's better on that formula than 82% of the D in the league – with the caveat that he was playing in a third-pairing role, which probably inflated his stats. (That's why JFresh gives you that context on his cards.)

Based on his performance the last three years, he's an exceptionally good third-pairing defenceman, the kind that can cover for a poor partner. Based on last year, he's less than that. Based on his age, he shouldn't be playing 82 games or anywhere near it, so he's probably best suited for a #7 role.

[Oil Stain]

Originally Posted by butterfly

Those player cards are kind of useless. WAR is a counting stat, not a rate stat. What does 82% WAR mean mathematically? A 7th D as suggested by JFresh?

A 7th D is, by definition, replacement level.

The WAR stat is broken because it can't differentiate between guys getting sheltered, vs guys that are better but playing bigger roles on bad teams.

Connor Timmins had even better advanced stats than Giordano last year, and William Lagesson had similar stats. Are they all top notch 3rd pairing guys playing 13-14 minutes per night?

I would presume that they just got heavily sheltered on a good team which led to their good underlying numbers.

Duncan Keith had like a 4% WAR in his last season in Chicago playing first pair on a bottom feeder. He was alright/serviceable in a second pairing role in his one season in Edmonton.

I've always felt that underlying numbers for players are better at comparing within teams than across teams, and the stats guys have always had great difficulty trying to balance quality of teammates and opponents with zone starts, etc.

For instance something some NHL coaches will do is throw their third pairing over the boards if Nathan Mackinnon has been on the ice for 40+ seconds, and the puck is headed towards the Avalanche end of the ice. If a coach is consistently changing his players like that, then a third pairing guy most likely will end up with positive underlying numbers against good competition.

Very tricky to come up with a one number overall ranking of a player like WAR when things like this are common in hockey.

[Monahammer]

Originally Posted by calculoso

You’re suggesting that the coaches need to helicopter parent their players and that every dressing room is 100% cohesive?

Not at all. One or two of them would be hand waved away easily. But the fact is we had like 5 coaches in 6 years who couldn't get any effort out of the room consistently. That comes back to inside leadership and at the time Gio was captain. Imo those teams were totally bereft of internal leadership. I have the receipts about complaining about it back then if you want to look at my post history.

Gio was a leader in name and not in practice.

[butterfly]

Originally Posted by Jay Random

The WAR percentile is how Giordano ranks compared with all other NHL defencemen. He's better on that formula than 82% of the D in the league – with the caveat that he was playing in a third-pairing role, which probably inflated his stats. (That's why JFresh gives you that context on his cards.)

Based on his performance the last three years, he's an exceptionally good third-pairing defenceman, the kind that can cover for a poor partner. Based on last year, he's less than that. Based on his age, he shouldn't be playing 82 games or anywhere near it, so he's probably best suited for a #7 role.

So if they put me in for a few games everyone else would be 99-100%?

It's simply wrong, because only the worst player in the league can get 0% WAR, which JFresh thinks is the same as replacement level, which also means that no below-replacement level player has ever played in the NHL.

[Reggie Dunlop]

Signing Mark Giordano as a player is a separate issue from signing Mark Giordano as a management trainee. His agent is doing him a disservice by seemingly linking the two roles.

[Hockey-and_stuff]

Originally Posted by butterfly

So if they put me in for a few games everyone else would be 99-100%?



It's simply wrong, because only the worst player in the league can get 0% WAR, which JFresh thinks is the same as replacement level, which also means that no below-replacement level player has ever played in the NHL.

Replacement level on the JFresh scale is around 20% WAR. Below that you are below replacement level.

Sent from my Pixel 7 Pro using Tapatalk

[butterfly]

Originally Posted by Oil Stain

The WAR stat is broken because it can't differentiate between guys getting sheltered, vs guys that are better but playing bigger roles on bad teams.

Connor Timmins had even better advanced stats than Giordano last year, and William Lagesson had similar stats. Are they all top notch 3rd pairing guys playing 13-14 minutes per night?

I would presume that they just got heavily sheltered on a good team which led to their good underlying numbers.

Duncan Keith had like a 4% WAR in his last season in Chicago playing first pair on a bottom feeder. He was alright/serviceable in a second pairing role in his one season in Edmonton.

I've always felt that underlying numbers for players are better at comparing within teams than across teams, and the stats guys have always had great difficulty trying to balance quality of teammates and opponents with zone starts, etc.

For instance something some NHL coaches will do is throw their third pairing over the boards if Nathan Mackinnon has been on the ice for 40+ seconds, and the puck is headed towards the Avalanche end of the ice. If a coach is consistently changing his players like that, then a third pairing guy most likely will end up with positive underlying numbers against good competition.

Very tricky to come up with a one number overall ranking of a player like WAR when things like this are common in hockey.

The reason for this is because the data isn't available. The technology is, but the NHL doesn't like curious people having it for whatever reason.

They put these things in everyone's jerseys and presumably the puck, so we would be able to know how long MacKinnon has been out there and the location and velocity of the puck to be able to build models that neutralize it:

[butterfly]

Originally Posted by Hockey-and_stuff

Replacement level on the JFresh scale is around 20% WAR. Below that you are below replacement level.

Sent from my Pixel 7 Pro using Tapatalk

Then that level of performance should be 0 and everything progressively worse should be negative.

Am I interpreting this correctly? Mark Giordano has 82% "WAR" according to that scale, so pick 6 NHL defensemen randomly and give them Giordano's usage, and 5 of them would be worse?

[EVERLAST]

I think it's interesting how certain agents have done a huge disservice to their clients.

Mouthing off for the player not only ticked off the coach , but the GM and the team. I can count numerous Ex Flames and 1 current 1 who never moved thr needle ...complained in thr media about the player not getting fair treatment and ended up causing issues behind the scenes.

I'm glad Conny got rid of unhappy guys who didn't want to to be here a day longer than they had to be and I hope he gets rid of any more trouble that causes damage .

Prospects, call ups see this S going on and I cannot imagine the damage done. I'm glad we have the prospects we have coming up because it sounds like #1 they're already leaders #2 they come from winning pedigrees #3 these are smart kids raised by good parents...Zayne Parekh is a prime example of that #4 These kids wanna win and have done so at every level and we're all needle movers

So glad Craig has been so so very involved behind the scenes and isn't picking up future head cases ( fingers crossed) . I Mean we just lost the greatest example of that ( The whole world did) in #13 RIP

Winters needs to shut his yap.

[Oil Stain]

Originally Posted by butterfly

The reason for this is because the data isn't available. The technology is, but the NHL doesn't like curious people having it for whatever reason.

They put these things in everyone's jerseys and presumably the puck, so we would be able to know how long MacKinnon has been out there and the location and velocity of the puck to be able to build models that neutralize it:

https://puckiq.com/woodmoney?season=...er_season_team

The data is kind of out there. This site tracks the % of ice time that players play vs elite, middle, and gritensity for example, and their results against those opponents.

It gets pretty complicated though. Is playing with the best players against the best players more of a boost than playing with the second best players against the worst players? How much is a zone start worth? What is the raw ability of Mackinnon 30 seconds into a shift, vs any other star player? Is the 3rd pairing D man getting matched with Mackinnon with his 4th line in front of him or his first?

https://www.naturalstattrick.com/lin...v3=0&v4=0&v5=0

Here you can look at McDavid and Draisaitl vs Mackinnon and Rantanen, but when you break it down that much the sample sizes get very small. McDavid played 2 minutes without Drai against the Colorado duo and Drai played 5 minutes alone. Can you get any useful data out of 2 minutes?

Further to that, maybe one team is playing their third game in four nights at the end of a road trip. Maybe one player is playing injured. Maybe the score is 5-2 halfway through the third and the winning team is just trying to coast to victory. There are so many variables, I'm not sure WAR rankings are effective for individual players.

I suspect teams are breaking things down to look more at specific events that they believe lead to winning rather than overall WAR rankings for players, but I don't know for sure.

The public has more data now than ever before even with so many analytics sites going dark. Some teams probably have some pretty sophisticated models, but they are not going to share them for obvious reasons, and the NHL itself doesn't have any real incentive to develop and publish their own models.

[Jay Random]

Originally Posted by butterfly

So if they put me in for a few games everyone else would be 99-100%?

It's simply wrong, because only the worst player in the league can get 0% WAR, which JFresh thinks is the same as replacement level, which also means that no below-replacement level player has ever played in the NHL.

Wrong. JFresh himself puts ‘replacement level’ at the 18th percentile for skaters, if memory serves.

[Jay Random]

Originally Posted by butterfly

Then that level of performance should be 0 and everything progressively worse should be negative.

That's what you would see if JFresh published his raw WAR numbers and not just the percentiles. It's one of the objections I have to his cards.

[Enoch Root]

If Giordano's WAR percentile is 84%, then the stat is flawed. Full stop.

Or it is measuring something useless and irrelevant

[butterfly]

Originally Posted by Jay Random

Wrong. JFresh himself puts ‘replacement level’ at the 18th percentile for skaters, if memory serves.

Then it's equally as misleading as NHL Edge who won't even tell you anything about anyone in the bottom half of the league, laughably.

If you think I'm an 18% NHL player, well...that should tell you all you need to know about what that metric is evaluating.

[Jay Random]

Originally Posted by butterfly

Then it's equally as misleading as NHL Edge who won't even tell you anything about anyone in the bottom half of the league, laughably.

If you think I'm an 18% NHL player, well...that should tell you all you need to know about what that metric is evaluating.

I don't think you're an 18th-percentile NHL player, and neither does JFresh.

What he's saying is that the top players in the AHL and other feeder leagues are better than the bottom 18% of NHL players. This stands to reason, because an organization doesn't benefit by having its 23 most skilled players on the major-league roster. A waiver-exempt player will usually be sent to the minors ahead of a slightly worse player who requires waivers, for instance. Teams aren't equally deep at all positions, so a team might (for instance) have a defenceman in the minors who's above replacement level, but no job for him, while they have to play a winger on the NHL roster who is below replacement level. And of course talent isn't evenly distributed among all teams.

None of this is to say that you would have any WAR rating at all. The system isn't designed to account for people who don't play the game and would never be offered an NHL contract.

[Jay Random]

Originally Posted by Enoch Root

If Giordano's WAR percentile is 84%, then the stat is flawed. Full stop.

Or it is measuring something useless and irrelevant

It's a composite measure of a whole bunch of factors, and the weighting can throw them out of whack. The first problem is that it takes an average of the past three years. Giordano was still a very good defenceman three years ago, and that raises his average. He is still a really good penalty killer and can generate more offence than a typical third-pairing D, so those things also raise his rating. But he's declining fast due to age and becoming a liability at even strength.

Most players don't have really lopsided skill sets like that, so most of the time the WAR model works reasonably well. Giordano is one of those cases that go outside the limits of the model.

[gvitaly]

I find it a bit funny that Gio's WAR with SEA was 65% per JFresh, he obviously slowed down over the last 2 years, but his WAR is higher due to his team, and sheltered usage.

There was a good reason TOR didn't play Gio or Brodie in the playoffs.

EDIT: found it