Some thoughts about the Edge AI or whatever they are calling it...

[onlybygrace]

Seems kind of pointless to 'use' this data in a game that has sliders for skating speed, shot power, shot accuracy, among many other things that obviously effect gameplay while claiming that players will perform according to this NHL Edge information.

Also, I saw a YouTube video today by Tougie I believe where they looked at real NHL Edge data for players and compared it to their ratings in NHL 26. There was no correlation even in the ratings to the data.

So where is this data going to actually be of any relevance? Especially when sliders alter the scale by which everything is measured anyways. Hell, in years past, I have to edit shot power and accuracy sliders WAY DOWN to make the goalies in franchise mode gameplay worth a damn.

If this and the new Be A Pro is all that theyve got to tell us, then whats the point.

Its confirmed we STILL have no goalie stats pre game and the player/team intros ONLY happen in BAPro. What!?!?!!

Lame.

Sorry, I digressed. lol

[The EVHL]

It would be cool if they used the edge data to actually create a baseline for speed, acceleration, agility, etc.

game is way too fast and loose by default, there’s no weight to the players in skating.

[zbguy]

I think the issue with tech like this is it's being applied to a game that is not and really has never been an attempt at modeling actual NHL play or being a hockey simulator, like EHM or HLS except played out on screen. It's more of an arcade, physics-based game, hence the sliders to allow for flexible gameplay. But actual hockey doesn't allow for the kind of flexibility in games like these; if actual hockey were as fast as EA's default sliders, it would be played completely differently.

That's why I'm skeptical that this tech can be used for what EA hockey is like. A game like this can't be played "realistically" for various reasons, so should the AI instead be geared towards consistency? Flexibility? And that's even assuming the sliders/attributes work correctly/do anything at all, which it sounds like some still don't.

[MiracleMet718]

It’s going to be more for tendencies and if used correctly, getting the behind the scenes AI decision making to be more accurate from a setup perspective. The physics and all the other stuff aren’t going to be impacted much as that is driven by the game engine itself (which needs tweaking and that’s a whole other story). But it should help to make the AI decision making actually make more sense.

[The EVHL]

I think the issue with tech like this is it's being applied to a game that is not and really has never been an attempt at modeling actual NHL play or being a hockey simulator, like EHM or HLS except played out on screen. It's more of an arcade, physics-based game, hence the sliders to allow for flexible gameplay. But actual hockey doesn't allow for the kind of flexibility in games like these; if actual hockey were as fast as EA's default sliders, it would be played completely differently.

That's why I'm skeptical that this tech can be used for what EA hockey is like. A game like this can't be played "realistically" for various reasons, so should the AI instead be geared towards consistency? Flexibility? And that's even assuming the sliders/attributes work correctly/do anything at all, which it sounds like some still don't.

I’d argue this game has always been able to be tuned fairly well (well, at least since NHL 19 when they expanded the sliders) but the AI used to really let it down with poor decision making.

From 23 and on, the engine is worse and the AI are worse, so a double whammy.