He also could have extended the elbow just as part of the animation they're going into. I really don't think the elbow is being extended to throw a hit. Sure, if these were real world players, there's always the possibility of trying to get away with a dirty hit. Obviously, being AI players, that don't make those decisions. As far as the game is concerned, this is incidental so there's no hit and hence, no penalty.
@EA_AljoHere is another example of AI not being sure what to do. Can you please tell me how my 2 forwards are getting in any position to open up for a pass for a play at the net or even screen the goalie?
I'm really not able to diagnose why the AI are behaving the way they do. I'm not a developer. I don't know what all the conditions are for their reactions. It looks like a player skates into the slot for a pass, but doesn't recognize the lane isn't open. I think we've already established they are often not in the best position. However, there are many times they are. It's certainly easy to cherry pick when they do something that negatively impacts your play when there are also many times they're getting open for good scoring chances.
@EA_AljoI know you cant explain in detail why it is the case. These instances were in an game and a half of play today. To me, I'm unsure of how frequent they are actually in a correct position or following the selected strategy.
Also I'm not sure how this is cherry picking? I was driving the puck wide to swing around for a possible pass in the slot or back to the Blueline.
@EA_AljoI know you cant explain in detail why it is the case. These instances were in an game and a half of play today. To me, I'm unsure of how frequent they are actually in a correct position or following the selected strategy.
Also I'm not sure how this is cherry picking? I was driving the puck wide to swing around for a possible pass in the slot or back to the Blueline.
I just mean, it's always easy to focus on the bad things they do when there also are many times they play really well.
@EA_AljoI know you cant explain in detail why it is the case. These instances were in an game and a half of play today. To me, I'm unsure of how frequent they are actually in a correct position or following the selected strategy.
Also I'm not sure how this is cherry picking? I was driving the puck wide to swing around for a possible pass in the slot or back to the Blueline.
I just mean, it's always easy to focus on the bad things they do when there also are many times they play really well.
How do you define "really well" in terms of AI competence. I find the underlying strategies to be rather faulty. I mean, 0/3 strategies result in "net front presence" on a consistent basis. You would think "crash the net" would have 1 player always screening the goalie, yet they usually stand 2-3 feet off the crease which is effectively useless if your main goal on offense is to establish a low-to-high cadence with lots of pucks and traffic on net. The non-negotiable rule of an offense centered around quantity rather than quality is taking the goalies eyes away which does not happen on a consistent basis.
Overload when starting down low always sees the closes sport wing actively leaving the center of the ice to go stand literally on the boards. This not only does nothing offensively, it actually eliminates the puck carrier's option of passing it to the strong side point or skating up the wall to try and attack the tops of the circle as there's literally a teammate in the way forcing you to the middle which is where the defender is which takes time and space away from their own teammate.
Behind the net sees the play start centered behind the net and then has both supporting wings literally stand stationary off the posts of the net and never sees the AI take their handedness or the puck position into account. They actively will not move into the high middle slot even if wide open which is a saving grace because I can almost guarantee no defensive player would adequately track said offensive player to that spot if it did happen.
I actually think it's much more likely that "cherry picking" would be coming from the "AI is actually pretty good" side compared to the "what are the AI doing?" side of the house based on the fact that the in-zone strategies for both offense and defense simply don't align with modern hockey and also that many of the FC and NZ strategies are either fictional or not executed properly. Now, maybe the AI are able to execute these flawed plans correctly at a decent clip, but they're simply not good strategies and they don't produce good hockey results because of it.
@Limp_KidzKit I can't remember his name but on NHL 23 he was very into making a lifelike experience and with rosters too. He had the game playing very sim-like and said it required you to play your positioning almost exactly how you would need to be. He also position locked to defense too. So with tuning the AI can play very great it just requires you to play exactly how they want.
@EA_Aljo @Yes it can but honestly I was not trying to focus on any negative and went in trying to find the fun again in it (seeing I have stepped away multiple times)
I spent time relocating and creating a team. Worked on sliders too.
It's more the fact that these AI issues stand out so much, it makes me step back, pause the game and try to figure out what was going on, which breaks the gameplay/immersion.
Thus making it hard to turn a blind eye to it and in a sense force myself to ignore the negative.
@MasterB89 Look up PernellKarl027 on the NHL 23 forums. He went above and beyond creating a sim-like offline experience. He edited everything including stats of players and such.
@KlariskraysNHL wrote: @Limp_KidzKitI can't remember his name but on NHL 23 he was very into making a lifelike experience and with rosters too. He had the game playing very sim-like and said it required you to play your positioning almost exactly how you would need to be. He also position locked to defense too. So with tuning the AI can play very great it just requires you to play exactly how they want.
My entire point is that the current strategies in the game are not "sim-like" even with perfect execution. Sure, I don't doubt their ability to play their programmed strategies to perfection, but the strategies when executed perfectly are fundamentally flawed.
This is very apparent when the puck is in the zone. There is zero AI understanding of assignments and how to keep said assignments. A simple "3 high" look will not see the defensive center taking the "extra" guy. Everything is very obviously a "zone" foundation when high-level hockey is primarily M2M for all intents and purposes.
So no matter the sliders, the game is unable to be even "sim-like" when you see very obvious positional errors in all zones that I believe can be mitigated by a different approach to how the AI execute forechecks and mark assignments in their own end.
Not to "double dip" on you, but heres a screenshot, 26 seconds into the period:
Notice:
Poor wing path to D with puck. Lane to net extremely open (this is half due to poor faceoff setup)
Original strong-side D collapsing to net while allowing his mark to stand open on the weakside...now, the wing is helping momentarily right now, but if this frame were to progress further, the defensive wing would continue going high and the LA wing would be left wide open.
Another grab, 30 seconds into the game. PK is running "Weak Side Lock" as seen at the start of the video.
Weakside Lock Reminders:
A weakside lock is a "1-2-2" variant. This means D pinching shouldn't be a main part of a forecheck
Should be forcing the puck to the weakside
Notice:
The game accounts for a pinching D according to the description...this is a fundamentally flawed "weakside lock" You can tell this by the fact that
The game has Kopitar (LA #11) force the puck towards the strong-side
The game has Kempe (LA #9) grab the COL center
The game has LA #55 take the COL weakside wing
Thus leaving the puck carrier to LA #44 who starts moving that way at 1:21 of the video.
Quick recap...we didn't force the puck to the weak side and we wanted the D to pinch in a strategy that already sees 3 forwards low. Not a great strategy even when executed properly. It goes against every foundational element of a weakside lock
Here's what should've happened (screenshot taken a second before erroneous WSL started being executed)
Kempe should've started the forecheck as the closest forward. Also, already had strong-side leverage which is an added bonus
Koptiar should've stayed with the Center he was already next too
55 should've taken his wing as originally planned
No D should be expected to pinch
This simple change would be a proper weak side lock which would not only remove confusion about the game's current weak side lock which sees the puck being forced to the strong-side and a D pinch, but is also just a much more fundamentally sound hockey strategy that better accomplishes puck pressure while maintaining a better defensive structure.
Now, this is 30 seconds into this video. Regardless if anyone finds his game speed and sliders realistic is irrelevant, the fact that the AI have made these simple mistakes in such a short amount of time proves that there's some fundamentally improper AI programming right now which means no amount of slider changes can make for a "sim" experience.
About NHL 24 General Discussion
Discuss the latest news and talk with us about your experiences in NHL 24.4,284 PostsLatest Activity: 2 years ago
Community Manager
