2024-25 Player Cards discussion

[Phil-Cornwall]

They look good. Pretty accurate on first look.

Well done Big man.

[Phil-Cornwall]

Just took a look at the FA skater list....wow...a TON of really good free agents to pick from this year. A bumper crop.

[Jon-Seattle]

www.stat.cmu.edu/cmsac/sure/2023/showcase/hockey_saves/report.html

Save percentage has declined year over year almost every single year for the last decade. It doesn't show 2024 but i believe when i looked it up it stayed in that .904 mark

This is in response to some of the discussion regarding goalie cards on the twitch stream. I missed the first half hour or so of discussion but the talk of a 50-50 goalie card being representative of average (Vasilevsky and his 900 S%). 

There is some crazy statistical analysis going on in that article but i just wanted to grab it for the first graph which shows save percentage over the last few years. I remember talking with people and thinking that .920 was a realistic average to expect from a top 10-15 goalie as can be seen in their 4th graph presented. (It's still around .917 but has also steadily decreased from a high of .923 just a decade ago). 

So this information can be used in several ways but I think the discussion revolved around ensuring goalies are more accurately represented by their cards in comparison to their save % and trying to figure out what number is good/bad/average. 

I would thus argue that a 50-50 card for Vasilevskiy should be considered below average just based on his save % being below the average. Does this accurately reflect how he will perform in the GHL though is a good question that was brought up. Would a 50-50 card accurately replicate his .900 Save%? Theoretically it should, but as we all know there are other factors involved such as GAA, shot placement, shooter proficiency, stamina, etc. will alter his GHL stats but lets say all things being equal I would hope the sim would portray a .900 Save%. 

To test this, an interesting solution could be to replicate NHL team rosters and run the sim and see what sort of differences we see. Not sure if thats practical or feasible but could shed some light on what tweaks could be made to more closely align real players performance and their cards. 

For me, I think it would be nice to have some sort of weighted stat system within the dice roller to ensure players stay within a certain threshold so that a goalie like Vasilevskiy stayed within something like an 895-905 save % threshold or at least something within reason. There would be less of those 8-0 games and more 4-1 games. He would also more consistently give up 3-4 goal games while his big save % or whatever it is could still allow for a shutout here and there. Not sure if any of that helps or is realistic but a thought i had when ya'll were talking.

The other issue is that we dont just rely on Save% but also GAA for goalies and so there is more than just 1 factor to look at when determining card values. GAA could completely contradict how good/bad a goalies save % is and then you have to figure out which one should be more highly valued.

So in truth I may have a ton of ideas or thoughts on how we could alter things but I think many of those changes could cause drastic unforseen consequences that completely wrecks what we already have and I dont want to mess all that up just because I think Shesterkin should be a 85-90 goalie instead of his 62-65 goalie card

Phil sorry for steppin all over your thread but i figured this was as good a spot as any to settle in and takeover with my ridiculously incoherent thoughts.

[Jon-Seattle]

After all of this you also have should backups have significantly lower cards just based on them being a backup compared to their teams "starter". As this is a sub-theme(?) in the article. In addition if you start looking at capping games played or trying to evaluate where AHL goalies should be at you have so many layers of ranking players that even I get lost. I wanted to add this as I've been trying to talk with Suffolk on figuring out how you could cap players based on games played and isnt necessarily a part of this conversation but an extension that breaks off down its own perilous path but one I'm sure we'll touch on at some point in this discussion