On the subject of AI and "It's not hard, just do it"

[Deathlyhearts]

I get the feeling that they were back to their old practices of just painstakingly handcrafting those corridor sequences so they look great instead of building a functioning open world game system that actually functions. This is not the first CDPR has done something like this. They have a tendency of making demos that have little to do with the actual game, just so they could sell the product. Anyone remember the wall-climbing and riding of the 48-minute demo? Guess why we can't do that? Because they faked the demo to appear like they were actually playing.

Similarly we can only drink beverages in the cutscenes in 1st person , because those are animated by hand instead of using the game engine. Cant in open world cause no such system exists.

The gap between those uninteractable sequences and the open world part is massive.

ps. Im not going to say crafting a good AI is easy, but when you have 7+ years and 20-50 million budget, there are very few excuses allowed. Other game companies give the impression of doing better stuff in half the time.

They removed the wall-climbing during development and they were completely honest and open about it. I don't think those features were fake. I assume (and yes this is personal assumption) that they weren't fully tested across the board and ended up not working very well in the game for the most part, thus ending getting removed.

Also I see several people around keep pushing that false narrative that they had 7+ years of development, which is false. Can we already stop with that, please? It was already confirmed multiple times by official sources that -active- development started only in the late 2016 after Witcher 3 Blood and Wine DLC was complete. That's 4+ years of active development. Not 7+... You can absolutely see they needed more time to polish certain systems. If the game -actually- took 7+ years of developing it would've been way more polished on release I imagine.

 

[Netrunner2go]

so thats even more of a thread than - why have they stated advertising the game 3-4 years before they even started with development?... building the hype hm?

 

[kaotkbliss]

so thats even more of a thread than - why have they stated advertising the game 3-4 years before they even started with development?... building the hype hm?

I would guess lots of discussion on story (which would need to be written before any work on the actual game starts),
systems the will use for combat, leveling and everything else
What kind of game it would be
etc.
There are a ton of things that need to be known before actual work on the "physical" game can even begin.

 

[DC9V]

note-perfect guitar animations?

I don't want to get off-topic but ...are you kidding me? It's well done for a video game but far away from "note-perfect". I play the guitar for more than 20 years. You can pitch the note just by pressing on the string, you don't even need to move your hands if the frets are tall enough. I once saw a classical guitar player playing the same piece over and over again while slowly playing softer until there wasn't any music coming out of his guitar anymore but he remained to move his body and his hands in the same way without touching the strings and you could still hear the music in your head going on for minutes. There is so much more perfection to playing the guitar than you will ever see in a video game...

 

[PepeSmug]

I don't want to get off-topic but ...are you kidding me? It's well done for a video game but far away from "note-perfect". I play the guitar for more than 20 years. You can pitch the note just by pressing on the string, you don't even need to move your hands if the frets are tall enough. I once saw a classical guitar player playing the same piece over and over again while slowly playing softer until there wasn't any music coming out of his guitar anymore but he remained to move his body and his hands in the same way without touching the strings and you could still hear the music in your head going on for minutes. There is so much more perfection to playing the guitar than you will ever see in a video game...

Point being, they attempted to animate the motion required to play each note, regardless of the musicality, which was still far more than required

I'm also a classically trained musician

 

[DC9V]

Point being, they attempted to animate the motion required to play each note, regardless of the musicality, which was still far more than required

I'm also a classically trained musician

Ok.. Sorry for ranting.

 

[neatchee]

Then I don't see the point, at all.

Yes, game development is hard. So are a lot of things. Many companies in any number of fields undertake difficult, intensive, and complex projects, and produce an end product that meets or exceed expectations.

When a major studio is struggling to produce AAA-level quality work, then it speaks to deeper issues with that studio -- not just the fact that operating at a high level in a particular industry is difficult or resource-intensive.

Hi there! I'm the OP from Reddit I registered here specifically so I could respond to this comment!

I can say with 100% certainty that...

a) the context of my original post was very different, and was about a very specific fight in Outriders that was horribly tuned

b) even so, the point was to illustrate the *types* of feedback that are useful. Saying "just do it, it's not hard" isn't even close to the whole picture. It's good for players to know the process so that the conversation can focus on the right things (resource allocation, etc)

Criticizing CDPR for how they spent their money is very different than just saying "this should have been better." Let's have a conversation about their business choices! But not "they suck at X". I promise every person assigned to a given feature worked as hard as they could under whatever constraints they had. And the answer isn't always "throw more money at that department".

 

[PepeSmug]

even so, the point was to illustrate the *types* of feedback that are useful.

Was this ever a mystery? People who just say "make it better 4head" are not worth giving the time of day.

Criticizing CDPR for how they spent their money is very different than just saying "this should have been better."

You're kind of contradicting yourself, here.

The development of a game is a black box to the end consumer. Really, the best anyone can say is that X should have been better.

We can suggest that resources could have been allocated more efficiently, but at the end of the day it's all conjecture. Maybe the solution wasn't better AI, but instead tailoring the gameplay to make the most of the AI they had. But pointing out that one aspect or another of the development process is "hard" is needless deflection from the original criticism.

 

[neatchee]

Yes, game dev is a black box to consumers....which is why I'm here trying to make it less of a black box...

I don't think it's too much to ask that folks try to steer clear of prescriptive feedback ("add/remove/do this thing") and prefer I-statements ("I don't like this because it makes me feel...").

And secondarily, try to talk about the things you'd be happy to give up to make it happen (both planned and existing stuff). That gives designers a much better idea of how important things are to you.

"I'd really like to have feature X. I don't really care so much about feature Y. Hey CDPR, can you prioritize this over that going forward?"

"When I go to A and see B, it breaks my immersion much more than if they'd just skimped on C."

^These types of conversations are way more constructive *in the context of game dev*, and can expand into much more useful discussion.

 

[Sardukhar]

But pointing out that one aspect or another of the development process is "hard" is needless deflection from the original criticism.

It's a discussion forum. Discussion is the point. Pointing out some parts of development are "hard" (why in quotes?) is fair. That's not a "needless deflection" at all, it's another viewpoint, equally as valid.

It's okay to have opposing viewpoints. And pointing out the difficulty with developing features in games provides perspective to an audience, that frankly, tends to lack it.

"Make it better" is a common, common instruction, in one form or another. Perhaps it's obvious to you that is insufficient, but it is not at all obvious from forum, twitter and Discord posts.

There are multiple issues with 2077, and AI is a big one. It's not so simple to improve, which is why we still have quite limited AI in videogames. We really haven't gone far past Half-Life and F.E.A.R. AI, even now.

 

[BionicDance]

So, just to point it out since I'm in the industry:

I used to be in the industry, but as an artist, mostly cut-scene scripting and level design.
But I played around with some coding in high school, just enough to know how much I don't know...

tl;dr: making boss fights harder by making them 'smarter' is costly, and if you want that then you need to be okay with slimming back somewhere else in the game to cover the costs.
There is a (not very good) reason why devs don't do the things you recommended sometimes...it's easier and cheaper to add health than literally anything else.

I only got to this forum recently; I've been spending most of my time on the Steam forums until now. I say that only to point out that we've not been listening to the same people, and I'm not trying to contradict you.

That said, most of the folks I know who've said the AI is "stupid" mean it quite literally: the NPCs act like complete chowderheads.

The last patch seemed to make non-combatants a bit smarter in terms of whether they run away or cower, but the enemies are still idiots.
I mean, we're talking, "Hey, that person I'm shooting at just tossed a grenade at my feet! I'll stand here, right atop my present until it--" **BOOM!**
That level of stupid.

So, in my experience, AI-needs-to-be-smarter isn't necessarily so much that the baddies stink at killing me as it is that they have zero sense of self-preservation and are thus too easy to kill. At least in the context of this game, anyway.

It's not possible to "just add something, it's not hard" because I promise you there is no game studio in existence that releases a game with work hours to spare.

You're not wrong.

What if it were re-phrased as, "Just add something; games have been doing it for ages, now"...? Cuz I think that's a huge part of why people are even discussing it: CP2077's AI just isn't even up to today's standards for this sort of game, much less being revolutionary.

I think the question on many peoples' minds with regard to this issue is, "...whatinheck was CDPR thinking???"
And, all snark aside, I would love to pick their brains to know what steps led to CP2077 being what it is, the decisions or assumptions that led to thinking they'd made something of quality, release-worthy.

 

[PepeSmug]

It's a discussion forum. Discussion is the point. Pointing out some parts of development are "hard" (why in quotes?) is fair. That's not a "needless deflection" at all, it's another viewpoint, equally as valid.

It comes off more as a deflection and excuse as opposed to good-faith discussion. It's basically a strawman; taking the most naive criticism ("AI isn't hard, just do it") and knocking that down.

It's okay to have opposing viewpoints. And pointing out the difficulty with developing features in games provides perspective to an audience, that frankly, tends to lack it.

It provides "a" perspective. But, as I was getting at before, it deflects from the perspective that something is wrong somewhere in the chain if the AI in a top-shelf game is sub-par, or stands out particularly as a problematic aspect of gameplay.

Sure, a lot of games have NPC behavior that is awkward or results in combat with insufficient challenge. So why is it apparently such a pain-point with the reception of Cyberpunk 2077? Is it just that expectations are too high, or is there something more here? That is where some real discussion might be had.

"Make it better" is a common, common instruction, in one form or another. Perhaps it's obvious to you that is insufficient, but it is not at all obvious from forum, twitter and Discord posts.

Here's some perspective for you. Maybe the interpretation from the developer side of "make the AI better" is too literal. Of course the layperson has no idea how easy or hard the development process is. But guess what, that isn't their problem. It's the job of the developer to minimize the number of people that notice such limitations. If the answer to bad AI isn't to simply brute-force a better implementation, then there must be other aspects of the game -- balance, level design, mechanics -- that could have made better use of the available technology. Resorting to the excuse "game development is hard" isn't a discussion, it's a cop-out.

 

[neatchee]

Honest to god, @PepeSmug , you have some great feedback but it's presented poorly; and I don't mean that you're stupid, or that you shouldn't feel how you feel, I'm saying your presentation doesn't help solve the problem.

I understand how you are feeling, but posting "make it better" is literally no different than clicking the upvote button and moving on. They provide the same value. A "me too".

My entire OP on Reddit was about how to give useful feedback that helps designers and developers make the game better for you. Tempering player expectations - making sure people know what goes into the development process - is part of building that exchange of knowledge. If you know what it takes to get these things done, you'll be better equipped to make suggestions that accurately reflect YOUR assessment of cost/benefit.

@BionicDance

I think that's a big part of my OP - Presenting feedback in a way that enables the pick-your-brain kind of dialogue you want. "Just do it" doesn't do that as well as "...what were you thinking, CDPR? This is important enough that you should have prioritized it."

As a dev, I would much rather chat with the latter person over the former

 

[DC9V]

So why is it apparently such a pain-point with the reception of Cyberpunk 2077? Is it just that expectations are too high, or is there something more here? That is where some real discussion might be had.

We've already seen this discussion several times in this forum. It always results in personal attacks and skirmishes.
It would require scientific research to answer this question correctly.

 

[PepeSmug]

I understand how you are feeling, but posting "make it better" is literally no different than clicking the upvote button and moving on. They provide the same value. A "me too".

It would be very unfortunate if "just make it better" was all you took away from my remarks. And it would also be playing into exactly the issue I brought up with your OP.

My point, or perhaps suggestion, is that feedback from a naive source should be subject to interpretation. If someone wants "smarter AI," perhaps there will be ways to make the AI appear "smarter" without a literal refactoring of the logic. Example: combat styles in Creation Engine games can make enemies appear "smarter" by tuning the suite of parameters controlling combat behavior.

It's just my opinion, but responding to feedback with feedback on the feedback just stonewalls the discussion. The majority of people playing games are going to have no idea about the technical limitations behind the scenes, as frustrating as this may be. While I agree that improving understanding on all sides is a worthy goal, there is also value in considering uninformed feedback through an interpretive lens. There is also value in "me too" kinds of feedback. If people don't provide it, there is no way to know how widely held a particular view might be among a playerbase.

 

[GogRelvas]

I think one very important aspect of the original post reflects not on the investment (money and other recourse-wise) that was allocated to create the different systems but what the engine of the game is dealing with at any such time during the play of the game. Isn't it possible that a better AI is programmed right now but a simpler version, with less features, needs to be put in place in the actual game to support the city, lighting, sound, all the features that are being streamed at any given moment?
The sacrifices aren't just made when researching and creating code but also when allocating it all together or did I misread this part?

 

[neatchee]

It would be very unfortunate if "just make it better" was all you took away from my remarks. And it would also be playing into exactly the issue I brought up with your OP.

My point, or perhaps suggestion, is that feedback from a naive source should be subject to interpretation. If someone wants "smarter AI," perhaps there will be ways to make the AI appear "smarter" without a literal refactoring of the logic. Example: combat styles in Creation Engine games can make enemies appear "smarter" by tuning the suite of parameters controlling combat behavior.

It's just my opinion, but responding to feedback with feedback on the feedback just stonewalls the discussion. The majority of people playing games are going to have no idea about the technical limitations behind the scenes, as frustrating as this may be. While I agree that improving understanding on all sides is a worthy goal, there is also value in considering uninformed feedback through an interpretive lens. There is also value in "me too" kinds of feedback. If people don't provide it, there is no way to know how widely held a particular view might be among a playerbase.

I've been trying to tell you, man: the point of the post is education. You keep saying "some people are too stupid, just ignore their feedback, and include them as a basic statistic" but that's kinda crappy. Again, I have no idea what ya'll deal with over here in Cyberpunk - I'm not active on these forums - but educating the players on how things work is good, no matter which way you slice it.

I also fundamentally disagree that "feedback on giving feedback" is a bad thing; as long as it doesn't supplant, but rather expand upon the feedback itself, then you're building something positive.

I believe it is the responsibility of the speaker, just as much the listener, to make sure that communication is clear and well received. That goes from both sides.

I don't see how giving insight into the process, and encouraging players to give more concise, useful feedback is a bad thing.

 

[Jetro30087]

Usually when people are talking about "smart AI", they are talking about AI with the capability to interact with the game world in interesting ways. In Zelda BOTW for example, enemies exhibit a variety of contextual behaviors based on other items in their environment. They'll pick up weapons, sets weapons on fire, knock away bombs thrown at them, lookouts on towers that will blow horns to alert other monsters, ect. Their basic combat loop is pretty dumb, approach Link, hit Link. But the additional interactions the AI have with the environment make them more believable which is interpreted as smarter by the player.

If the AI in cyberpunk demonstrated more behaviors it help. For example, if a melee enemy is near a dead enemy with a gun, that AI would seem smarter if it picked up that gun and started using it. AI would seem smarter if it threw grenades back at you. Maybe the AI could grab a nearby civilian as a shield. Small interactions with the environment like that makes the AI seem smarter without tweaking it's stats and the more behaviors that are implemented like this, the better the AI tends to be perceived.

 

[PepeSmug]

I've been trying to tell you, man: the point of the post is education. You keep saying "some people are too stupid, just ignore their feedback, and include them as a basic statistic" but that's kinda crappy.

I said "uninformed" or "naive," not "stupid." What language would you prefer?

I also said above that this feedback should be considered, instead of ignored or deflected, but with an eye to the lack of technical expertise behind it.

Again, I have no idea what ya'll deal with over here in Cyberpunk - I'm not active on these forums - but educating the players on how things work is good

I said as much in my comment you quoted.

no matter which way you slice it.

Adding "no matter what" to a statement doesn't make it stronger or harder to argue against.

I also fundamentally disagree that "feedback on giving feedback" is a bad thing; as long as it doesn't supplant, but rather expand upon the feedback itself, then you're building something positive.

But in this case it does supplant the feedback. What I read in your post was a lot of "sorry, too hard."

My counterpoint is that you do better service to the people sharing feedback if you try to understand the spirit of what is asked for, and consider how it might be accommodated within the available technical framework. I'm not saying that there is an obligation to consider all feedback or to treat it all equally. But it is far more constructive to look for compromise as a part of the process.

I believe it is the responsibility of the speaker, just as much the listener, to make sure that communication is clear and well received. That goes from both sides.

What about the communication wasn't clear or concise? How much education does the layperson need until they can offer feedback you'd consider valuable or actionable?

I don't see how giving insight into the process, and encouraging players to give more concise, useful feedback is a bad thing.

I didn't want to belabor the point, but frankly I don't see where you explained how to make feedback more useful. What I saw was a long list of "no can do." Which gets back to my original point, that the entire premise of this thread comes off more as deflection and making excuses for flaws in the game, rather than a basis for a more constructive dialogue going forward.

 

[greeks]

In Cyberpunk it's more of a case of "a pedestrian forgot how to walk all of a sudden"...
I doubt that the developers were like "coding AI is difficult so we focused on pretty graphics instead. But it turned out that pretty graphics is difficult too, so we focused on adding Keanu Reeves to the game."
It's not like they don't know that their AI is broken. It's obvious the moment you walk out of Megabuilding 10. "Fix your AI" is not that bad of a feedback as in its current state literally anything you'd do to it would make it better.

Following on that above:
Feedback of "fix the game" is still important as the devs would say "they are not complaining about that element of the game anymore, case closed guys, well done!"
Also if only one person gives constructive feedback on a broken element of the game and the rest of 10 million players sits quitet because they have no higher knowledge on the subject, the devs would be like "ok, there is one unhappy guy out there, the rest seems ok with it, let's ignore it as it's too much effort to please one dude".

If we are not vocal about the state of the game, it won't be fixed. I'm sure that devs would love to make as good of a product as possible, but also they have to keep in mind that time and effort is money and perfecting certain aspects on a merit of "I can do better" is not very profitable.
I'm sure that after the game is "finished", there will still be a couple unhappy gamers on this forum complaining about a bug that they are still experiencing. And that bug will never be addressed as it seems to occur in only this guy's case. And if the bug is really hard to fix or management decides to not work on improving the game anymore...