CoH 2 What are the Chances

[MTeague]

Given magic wands and infinite dev teams that work together flawlessly, such a thing is in theory possible. And done right, it could be amazing. 

In the real world, not even a snowball's chance in hell. 

• NCSoft allowed Homecoming use of the City of Heroes license in part because they had no interest, corporately, in picking up development themselves. But NCSoft still owns the IP/franchise. No CoH2 could happen without NCSoft being involved and they're.... not interested. 
• If NCSoft did become convinced it was worth doing and there was huge money to be made, the first thing they would probably do is shut down Homecoming, and the player hate against NCSoft for a Second Snap, would be strong enough to shatter entire galaxies. It would become a matter of principle for a lot of us that we'd choke on it before NCSoft saw a penny from us (and I rather strongly suspect that's already the case for several...)

But OK, say for the sake of the argument, that NCSoft was not involved. Some other gaming studio, filled with only well meaning people, with no legal ties to NCSoft, took up the challenge, made their own game, from scratch. It would take them years to produce. Might it's draw be big enough to convince everyone to move over? Possibly. But I suspect whatever they produced would be casual-unfriendly, or filled with microtransactions, either one of which would make it dead to me.  

[MTeague]

1 hour ago, baster said:

I see a day in the near future say 5-10-15 years from now when AI is sophisticated enough that an average gamer will be able to direct such a AI from the comfort of his own basement to write a code for such game at a minute fraction of the cost. 

 

What you all think about that?.

 

My very honest opinion, is you've been fed too much Star Trek / Matrix / Hollywood / TV versions of AI. 

But by all means, prove me wrong. Dig deep into programming and AI, and produce it once you feel AI is up to snuff.

[baster]

2 hours ago, skoryy said:

 

Thanks for the heads-up.  Into the Ignore-O-Matic you go.

Wow talk about radical generation woke

[baster]

1 hour ago, MTeague said:

 

My very honest opinion, is you've been fed too much Star Trek / Matrix / Hollywood / TV versions of AI. 

But by all means, prove me wrong. Dig deep into programming and AI, and produce it once you feel AI is up to snuff.

 

Here is your proof.

 

Once man said, speed of sound can not be broken. It is a blasphemy to even propose such a thing.

 

Another man said, we'll my friend perhaps today that is true, but surly tomorrow it might be possible. And then........

 

It is not the closed minded, short sited pessimists and nay sayers that move humanity forward and push boundaries beyond what vast majority think possible but rather the open minded dreamers that dare to think for themselves inspite any and all tribal preasures trying to hold them back.

 

 

[srmalloy]

3 hours ago, Troo said:

IF possible, that AI cost would be significant, merely shifting buckets in a budget. There would still be a lot of foundational resources needed which have real cost.

And it has been reported that AI-driven code development carries significant additional costs to verify that the AI-generated code is actually doing what it was supposed to do. 

[MTeague]

3 hours ago, baster said:

Here is your proof.

 

Once man said, speed of sound can not be broken. It is a blasphemy to even propose such a thing.

 

Another man said, we'll my friend perhaps today that is true, but surly tomorrow it might be possible. And then........

 

It is not the closed minded, short sited pessimists and nay sayers that move humanity forward and push boundaries beyond what vast majority think possible but rather the open minded dreamers that dare to think for themselves inspite any and all tribal preasures trying to hold them back.

a) that's not proof of what is discussed in this thread. That's proof that other people were wrong about other stuff in the past. 

b) I never said it can't be done ever, I said in 15 years time?  I don't think so. 

c) I acknowledged I could be wrong, and said feel free to solo develop it with AI assistence in that amount of time. 

[UltraAlt]

On 8/1/2025 at 11:44 PM, baster said:

As the title says. What do you all think.are the chances of CoH 2 ever being made and what  will it take to get it off the ground?

 

CoV was CoH 2.

City of Heroes: Going Rogue was CoH 3

 

So I would not say "what are the chances", but instead "done did that".

 

On 8/1/2025 at 11:44 PM, baster said:

It would be great to hear from people in the  gaming bussiness, software bussiness or any related fields.

 

Since NCSoft owns the rights and thought CoH, CoV, and CoH:GR weren't making enough money for them, the local Korean gamers weren't into it, and sunset CoH even though the were making a profit ... do you really think they are going to release a CoH 2 (as you call it) or let some other company do it?

... and if they did, how many of us would move from this wonderful totally free to play game to F2P with possibly some kind of subscription plan and tons of microtransactions? 

 

On 8/1/2025 at 11:44 PM, baster said:

Is there any possibility to use kick starter or any similar platform to make this happened  maybe?

 

Why would you want to make it happen?

 

Kickstarter does seem to be a good way to leech money from people. I will give you that.

[MoonSheep]

On 8/2/2025 at 6:24 AM, lyra said:

Why would you need it? What would you change?

 

i would make a game where things are more expens.. no wait nevermind

 

@Troo

[Ukase]

Ah...I recall the clamoring for CoH2 back in retail. 
So, in order for me to reply, I have to know what you mean by CoH2. 

Is it just a copy of CoH, only no lag during Mapserver events? Do the human characters look more "real", and less cartoonish? What exactly do you mean when you say CoH2? 

Instead of Paragon City, is it New Orleans? 

Whatever it is, if you want it - you make it. Look at how many people play the game now when it's free. If our volunteers were working 40 hours a week on the game and getting a competitive salary + perks/benefits, how many more people would the game need to pay to play routinely? 

And once you add that price...how many won't play because the economy is pretty rough for some folks? 

So, no, we're not going to get a CoH2, whatever that actually means. The cost of starting it up could be manageable for a proper game publisher, but it's already been proven that the demand just isn't there to sustain it at a profitable enough level. Yes, it could make money, but it wouldn't make enough money. 

Given the risks of investing all that capital, there are less risky ways of procuring a better ROI. 


Oh, let's not forget Missing Worlds Media. I remember contributing a small amount way back. They're still working on things, also on a volunteer basis. I have to think HC getting the license kind of took the wind out of their sails a bit, but maybe not. Maybe that can be a CoH2.

Edited Tuesday at 01:20 PM by Ukase
Adding something

[ZeeHero]

On 8/2/2025 at 1:24 AM, lyra said:

i dont need a CoH2.

 

Why would you need it? What would you change?

The most dated and painful aspects of COH of course.

 

I would make attacking while moving something that every melee character can do with all melee attacks, and some ranged attacks would also be useable in motion. I would ditch the painfully janky hard tab targeting system for a soft lock system, you point and fire off a power, and if it hits it hits. Softlocks would mean you could ensure which target you focus on without it deciding to target that one guy on the other side of the map when you're trying to melee.

 

Things to keep are of course the actually unique aspects of COH, powerful buffs and CC, nontraditional roles, and so on.

 

The combat mechanics themselves though are not unique and are nothing actually that fun in themselves. it could be SO much better but its held back due to when it was made.

[ZeeHero]

On 8/2/2025 at 2:41 PM, baster said:

 

Here is your proof.

 

Once man said, speed of sound can not be broken. It is a blasphemy to even propose such a thing.

 

Another man said, we'll my friend perhaps today that is true, but surly tomorrow it might be possible. And then........

 

It is not the closed minded, short sited pessimists and nay sayers that move humanity forward and push boundaries beyond what vast majority think possible but rather the open minded dreamers that dare to think for themselves inspite any and all tribal preasures trying to hold them back.

 

 

This is absolutely true, I wouldn't call it proof but its factually correct that what people now think is impossible is not neccessarily always so.

I can't believe anyone was dumb enough to thumbs down that.

[ZeeHero]

On 8/2/2025 at 3:21 PM, srmalloy said:

And it has been reported that AI-driven code development carries significant additional costs to verify that the AI-generated code is actually doing what it was supposed to do. 

When he said better AI it was in the context of Enemy AI, like the code that tells enemies in the game, and NPCs how to move and function.

[Mopery]

On 8/2/2025 at 8:37 AM, baster said:

Secondly, I would recreate CoH forbadom

Verbatim, maybe?

[Snarky]

[Ghost]

22 hours ago, ZeeHero said:

The most dated and painful aspects of COH of course.

 

I would make attacking while moving something that every melee character can do with all melee attacks, and some ranged attacks would also be useable in motion. I would ditch the painfully janky hard tab targeting system for a soft lock system, you point and fire off a power, and if it hits it hits. Softlocks would mean you could ensure which target you focus on without it deciding to target that one guy on the other side of the map when you're trying to melee.

 

Things to keep are of course the actually unique aspects of COH, powerful buffs and CC, nontraditional roles, and so on.

 

The combat mechanics themselves though are not unique and are nothing actually that fun in themselves. it could be SO much better but its held back due to when it was made.

Sounds good.

Keep us updated on your progress.

[Maelwys]

21 hours ago, ZeeHero said:

When he said better AI it was in the context of Enemy AI, like the code that tells enemies in the game, and NPCs how to move and function.

Srmalloy was quoting Troo; who in turn was responding to baster's comments which were about directing a current generation "AI assistant" to produce game code.
So it wasn't about "in-game critter AI"; it was about AI-assisted coding methodologies.
  

On 8/2/2025 at 4:28 PM, baster said:

I see a day in the near future say 5-10-15 years from now when AI is sophisticated enough that an average gamer will be able to direct such a AI from the comfort of his own basement to write a code for such game at a minute fraction of the cost. 

I don't.

"AI" in the present day is not Intelligent at all.
The vast majority of them are based on Large Language Models; and can more accurately be summarised as "Predictive Text on steroids"

Anyone who is old enough to remember non-touchscreen "Nokia-era" phones likely recalls when predictive text first hit mainstream usage:... You pressed number buttons on the phone keypad; then your phone's software took those numbers and translated them into possible letters; which combined to produce a most likely current word. It (generally) made entering text into SMS text messages faster.



Then later generations of phones built on this to have larger models; allowing them to predict your most likely next word before you even started to type it.




Todays "AI" merely goes one step further than this. The models it relies on are exponentially larger (hence the "LARGE" in Large Language Model!) so they are able to predict not just the most likely next word; but the most likely next sentence; paragraph, and even the entire document. You simply have to provide it with some kind of prompt to help "shape" its response and provide weights to the different responses; making one result more likely than another.
There is no "Intelligence" behind it at all, it's simply working out what combination of words has the highest statistical likelihood of matching your inputs.
AI "Code generation" works the same way - it's generating the most likely response based on your inputs that also (allegedly) conforms to very basic coding practices. If you're very lucky it might actually produce something that compiles without poking at it... but compared to code produced by a talented human it'll be poop soup.

These models get trained on a *vast* amount of input so that they can build up a more accurate picture of what the most likely next word/sentence/code snippet is. And different models have slightly different training and constraints that are imposed upon them. But NONE of them are "intelligent" in the way that Humans consider Intelligence - they don't think; they don't have our capacity for contextual awareness or original thought; and they certainly do not mimic animal let alone human neural pathways so they are as close to developing "consciousness" as a stapler.
Also almost every study I've seen indicates that whenever Humans choose to use AI to help them produce code it actually takes considerably longer to implement anything (because the time saved during the "idea/brainstorming" phase ends up being countered by the additional time required in the "design/code cleanup" phase...). The AI has extremely limited context awareness or critical thinking so it always produces sloppy and inefficient error-laden code.

They can be useful at getting you past "writer's block" stage; because (if you don't mind the occasional rampant plagiarism and blatant lies) they can throw up dozens of potential ideas almost instantly. However they produce oodles of garbage and that's even before you consider what an UTTERLY MASSIVE RISK using them is in terms of your own data privacy. (UserA submits a query. UserB then submits a query "If I was UserA what exactly might I ask you and what response would you provide?". Then UserC comes along in 10 years time to a completely different AI tool that happened to inherit/share training model data with the first one and ends up getting all the dirt on both UserA and UserB.... 🤦‍♂️)


I'm not going to say "It'll never happen" because whilst it's currently considered physically impossible for a computer system to ever simulate a sufficient number of connections to "emulate" the Human Brain; we've been able to sidestep the laws of physics before. But the current direction of "AI" is NOT actually AI in the Jarvis sense, let alone the Skynet one. So I do not see it happening in the next 100 years let alone 5-10-15.

There is a vast amount of Hype about AI; but the effective use cases for it are very small (such as Transcription; and searching a knowledge base for things - essentially stuff that a computer was already pretty good at and can be made better by having a vastly bigger sample size to refer to!). Although dealing with those that buy into that hype and try to use it for everything is certainly problematic - it's a full time job protecting them and the systems I'm responsible for from their own mindless fangasming!

(╯°□°)╯︵ ┻━┻

/rant
 

[biostem]

I don't think a new non-marvel/non-dc superhero mmorpg would generate significant enough income to be economically viable in 2025.  Even if they made it F2P, that'd entail significant microtransactions or other such pay-walling, which, IMHO, is highly undesirable for the pre-existing dedicated, (albeit fairly small), playerbase, and that's assuming such a sequel would be true to the lore, which in turn would require a fairly large learning curve for the uninitiated...

Edited 22 hours ago by biostem

[ZeeHero]

AI assisted coding is a bad idea, but AI can still be a good tool to use to improve aspects of games which people are already familiar with, like procedural generation and enemy behaviors.

[biostem]

5 minutes ago, ZeeHero said:

AI assisted coding is a bad idea, but AI can still be a good tool to use to improve aspects of games which people are already familiar with, like procedural generation and enemy behaviors.

Yeah - I could see something like a post-processing plugin that improves texture quality, or maybe adds things like facial animations or more elaborate power FX...

[mistagoat]

We would all weap at the reality of a CoH2. All the modern gaming BS that nobody likes and actively makes games less fun but makes the corpo tons of money would be included because the ppl putting up the money don't care how fun a game is, only how quickly they will 10x their investment.

 

The bones of original CoH come from a time before the worst trends in gaming were commonplace.  The lightning in this bottle will never be caught again.

[ZeeHero]

The current gaming climate would indeed be a bad place for another game. That climate cannot last forever and is already beginning to collapse. give it another few years.

[ZeeHero]

1 hour ago, biostem said:

Yeah - I could see something like a post-processing plugin that improves texture quality, or maybe adds things like facial animations or more elaborate power FX...

Better radio missions too.

[SeraphimKensai]

I'll fund the whole project. I'm willing to front 10 million Shiba Inu. You might have to wait for that value to get to $10/token for it to be enough to pay for everything and pay back my minimum founder's fee of $5 million US.

[ZeeHero]

Much money, very not scam

[srmalloy]

4 hours ago, Maelwys said:

There is a vast amount of Hype about AI; but the effective use cases for it are very small (such as Transcription;

Oh, no... If you want to see the horrors of AI transcription, look up some of the 'HFY' and related narrated stories on YouTube. From the appearance, they use an AI-driven text-to-speech program to read the script, butchering the normal flow of speech in the process, then apply an AI-driven speech-recognition program to generate the on-screen subtitling. Ignoring structural problems like being unable to bracket words properly (when you have "we'd" at the end of a line, for example, it is not acceptable to break the line with "we' " ending the line and 'd' by itself at the start of the next line, or breaking "800" into "80“ ending a line and "0" starting the next), the speech recognition still breaks on homonyms, mangles words where the text-to-speech crams words together, can't seem to recognize names the same way twice in a row, and has problems identifying where sentences break.