|
|
LukaOo
Hired Hero
|
posted December 10, 2015 07:58 PM |
|
Edited by LukaOo at 20:03, 10 Dec 2015.
|
gomaki said:
Lets take another example. AI.I have spoken to one of the programmers who is involved in this process and he's written a reply for you to give you a bit of insight of why TBS (Turn based Strat) AI is much harder to programmer than say Call of Duties AI (FPS)
'FPS AI works a certain routine: search, destroy, (optional: die trying), recover. TBS AI works with hundreds of variables thinking what move to do next. We run an utility-based AI meaning that the AI as a whole is working kind of like an organic being. It has sensors (like living beings have a few of those.. you know.. skin.. nose.. ears.. gut feeling, too) and these sensors determine the "need" or "usefulness" of a certain action to perform. The flow is as follows: sensor->utility->action scoring->decision. our AI takes the top decision, even though it could have a random factor inside, serving as a sort of "indecision" factor, but we decided against it. Actions can affect each other and the scoring, so it's an ever-changing thought process. Every time an action is performed, everything needs to be re-evaluated. Changing just one single thing ripples across the system like you would expect from the classic butterfly effect. Sometimes it surprises us, and I mean a positive surprise. Sometimes it starts doing retarded snow. it's a question of finding the right balance between values, wants and needs and objectives we want the AI to achieve that defines its actions. '
Your guys work in the wrong place obviously. By description I decided that you try to create something like this http://asimo.honda.com/. It took a lot of years to develop given device. It sounds like an excuse: because it is very difficult therefore we can not create good AI. I prefere to get the cheated playable AI than mystical not cheated one in five years. You look on world like your company is unique, and constantly try to convince us that it solves unrealistic tasks, with which nobody faced before. It is not so. There are a lot of software companies which develop games, and how I can see solve similar problems. Your opus, Gomaki, convinced me once again that you do not have specialists in the company who can make a good game.
|
|
Conduit
Tavern Dweller
|
|
posted December 10, 2015 08:06 PM |
|
|
LukaOo said:
gomaki said:
Lets take another example. AI.I have spoken to one of the programmers who is involved in this process and he's written a reply for you to give you a bit of insight of why TBS (Turn based Strat) AI is much harder to programmer than say Call of Duties AI (FPS)
'FPS AI works a certain routine: search, destroy, (optional: die trying), recover. TBS AI works with hundreds of variables thinking what move to do next. We run an utility-based AI meaning that the AI as a whole is working kind of like an organic being. It has sensors (like living beings have a few of those.. you know.. skin.. nose.. ears.. gut feeling, too) and these sensors determine the "need" or "usefulness" of a certain action to perform. The flow is as follows: sensor->utility->action scoring->decision. our AI takes the top decision, even though it could have a random factor inside, serving as a sort of "indecision" factor, but we decided against it. Actions can affect each other and the scoring, so it's an ever-changing thought process. Every time an action is performed, everything needs to be re-evaluated. Changing just one single thing ripples across the system like you would expect from the classic butterfly effect. Sometimes it surprises us, and I mean a positive surprise. Sometimes it starts doing retarded snow. it's a question of finding the right balance between values, wants and needs and objectives we want the AI to achieve that defines its actions. '
Your guys work in the wrong place obviously. By description I decided that you try to create something like this http://asimo.honda.com/. It took a lot of years to develop given device. It sounds like an excuse: because it is very difficult therefore we can not create good AI. I prefere to get the cheated playable AI than mystical not cheated one in five years. You look on world like your company is unique, and constantly try to solve unrealistic tasks, with which nobody faced before. It is not so. There are a lot of software companies which develop games, and how I can see solve similar problems. Your opus, Gomaki, convinced me once again that you do not have specialists in the company who can make a good game.
I completely agree with you. luk.
Gomaki,
Your AI is NOT WORKING! Out of 7 games your AI is the worse! Why can't you guys see this! We never asked for this, You knew better just like you thought you did with everything else in this train wreck. I don't play MP and I know many others who don't as well, so a non working AI is game breaking for most people. Ubisoft is a narcissistic company and I can't wait for the day when it bites them in the ass and they come crashing down. Preferably sometime soon.
On a side note Gomaki, THANK YOU for all your communication. You give it to us straight even if its not what we want to hear. I'm sorry you're in bed with the devil but hopefull ubi will let you make the game great.
|
|
frostymuaddib
Promising
Supreme Hero
育碧是白痴
|
|
posted December 10, 2015 08:20 PM |
|
|
Conduit said:
I don't play MP and I know many others who don't as well, so a non working AI is game breaking for most people.
Even in MP, AI is important. Fighting neutrals that act as idiots is not appealing and fun at all.
@LukaOo
100% agree with you.
____________
"Occam's shuriken: when the answer is elusive, never rule out ninjas." -- Dr. Gordon Freeman (Freeman's Mind)
"lol" -- VERRIKER VON ERWINSSEN
|
|
EnergyZ
Legendary Hero
President of MM Wiki
|
|
posted December 10, 2015 08:34 PM |
|
|
|
Seeing all these discussions makes me wonder how well are doing those that just ignored all the comments that were writtenon this thread before and after H7 got launched. As well as the H7 site.
|
|
TD
Promising
Famous Hero
|
|
posted December 10, 2015 09:08 PM |
|
|
Antalyan said:
I say nothing like 'this game's development is great and Ubi did it right". Definitely no. Actually I hate the way of development "do the game asap and then fix it somehow". The only thing I am trying to say is: not to judge the whole game according to its development. The game might be finally great even with the horrible situation after the release.(I hope this is the case of H7). Consider, after releasing patches, you will not play the version 1.0, you are always interested in the final version only. Especially after some time after the development actually ends.
See that's kind of problematic to do. They had history of previous games and still copied the steaming pile of snow known as h6 for most part. Ironically with h6 and h5 I could really see there being potential. Here with h7 it's very hard to see any kind of potential because they've already stated they won't re-work the skill-system for example.
As for the judging, why should h7 be exempt from that? We know they won't change the many core problems of the game(they've stated so themselves!). What does it matter if they weed out every single bug in the game if it stays so boring that you never wanna play it? Honestly if Ubi announced that h6 or h7 will get new expansion depending who wins the vote I would place my money on h6. The game has sooooooooo much more potential than h7 and I don't even like nor believe in h6. It's just to show how little potential I see in h7.
This game would need to go back to the drawing board and re-launch the game in a year or two after many necessary changes and fixes. What they do now with just patching is same as putting band-aids to a leg that's gonna need an amputation anyway.
|
|
Brukernavn
Hero of Order
|
|
posted December 10, 2015 09:28 PM |
|
|
gomaki said:
Do remember though that Heroes Community, Celestial Heaven, Acid Cave or who ever else are the hard core community, you speak for a VERY small majority of the playing community. That doesn't make the opinions in-valid and in fact I wouldn't even be here if I thought what all of you guys said was irrelevant. In a lot of cases you're spot on with some assessments and ideas. But you have to understand we're not making a game just for you, the hardcore fan base. If some things don't make sense from the perspective of a hardcore player, it might be because it's better for a majority. However I will accept some things might just not be good, and in those cases we will aim to change what can be changed.
I just wanted to comment briefly on the "hardcore" thing. I consider myself a casual player, but I have spent many an hour with H3 and maybe even more with H5. Among the people I know in real life, I am the person with the largest interest in the Heroes franchise. For this reason my opinion of the game will directly influence whether my friends, family, co-workers etc. even consider buying the game. So far I have prevented maybe 10-15 people (IRL) from buying H7, just by comparing it to previous titles. I think you can assume the same for most of the people in the "hardcore" category. That's not mentioning the online effect. If the fans don't like it, you've lost your most fervent advertisers.
And in this case it's even worse. Not only do many of the fans not like the game, but they are actively working against it. I have previously argued that this is caused largely by the advertisement and communication from Ubisoft. And as JJ mentioned a few pages back, the fans warned about a lot of the problems long in advance, but were ignored.
Regarding the sales numbers for H6, that game had a lot of things going for it. It followed the eventual success of H5, the fans were not as clearly negative as with H7, and the reviews were generally positive. The real test of it's success is how many - after buying the game - bought the DLCs and expansion? Being the only game since H2 with only one expansion, and the early decision to end the support, indicates that it was not so successful in this regard.
And when we're already discussing sales numbers, let's include Heroes 7 and Heroes 3 HD. First of all, according to these numbers Heroes 3 HD has sold 6 times as many copies as H7 so far. Secondly, for modern video games about a third of the total sales are made during the first 24 hours (including pre-orders). If we generously let H7 sales so far represent a third of the total sales, it will only sell 1/4 as many copies as H6. This is another indication that many of the people who bought H6 were not pleased enough to buy the new iteration. Although the numbers are also affected by all the bad reviews. Once the bad reviews are out there, it's very, very difficult to repair the damage, however unjustified some of them might be.
|
|
olorin
Adventuring Hero
tophatchild.blogspot.com
|
|
posted December 10, 2015 09:32 PM |
|
|
JollyJoker said:
gomaki said:
You know, there IS a way to make both noobs and hardcore fans happy - just make a good game, is all.
Amen to that.
I just CAN'T believe Gomaki's words. Is he serious? 
A suggestion: hire the modding fellows lurking around here (concretely magnomagus)
|
|
Antalyan
Promising
Supreme Hero
H7 Forever
|
|
posted December 10, 2015 10:19 PM |
|
|
I don't see the problem of cheating AI too. The main thing for more than 90% of fans is to have AI which is fast and smart. If it cheats or not is much less important for theem, if they care at all. So if you and Ubi want to follow the majority of fans, which in this case involves the opinion of hardcore fans as well as the other players' you should definitely have to end ul with the same conclusion as I did.
Btw. I remember H6 AI very well, many times really annoying. For example, there was a mission in campaign where I was attacked every week by the same enemy's hero with exactly the same army, even using exactly the same skills every time. It's clear that I hated that mission a lot - it is just na example how important is to have AI smart (and it is not important if cheating or not). Hopefully this game will finally work better than H6 did.
|
|
Gryphs
Supreme Hero
The Clever Title
|
|
posted December 10, 2015 10:22 PM |
|
Edited by Gryphs at 22:37, 10 Dec 2015.
|
gomaki said:
Lets take another example. AI.I have spoken to one of the programmers who is involved in this process and he's written a reply for you to give you a bit of insight of why TBS (Turn based Strat) AI is much harder to programmer than say Call of Duties AI (FPS)
'FPS AI works a certain routine: search, destroy, (optional: die trying), recover. TBS AI works with hundreds of variables thinking what move to do next. We run an utility-based AI meaning that the AI as a whole is working kind of like an organic being. It has sensors (like living beings have a few of those.. you know.. skin.. nose.. ears.. gut feeling, too) and these sensors determine the "need" or "usefulness" of a certain action to perform. The flow is as follows: sensor->utility->action scoring->decision. our AI takes the top decision, even though it could have a random factor inside, serving as a sort of "indecision" factor, but we decided against it. Actions can affect each other and the scoring, so it's an ever-changing thought process. Every time an action is performed, everything needs to be re-evaluated. Changing just one single thing ripples across the system like you would expect from the classic butterfly effect. Sometimes it surprises us, and I mean a positive surprise. Sometimes it starts doing retarded snow. it's a question of finding the right balance between values, wants and needs and objectives we want the AI to achieve that defines its actions. '
Now it's very easy to look at what things we don't have, as opposed to what we do have. The fact that AI has sped up nearly 400% since release was not some 1 line of code change that everyone believes game development to be 'What do you mean you can't make this unit do that it's 1 line of code!!' is a common email I get.
Why, I must ask, was the A.I not good from release? Why does it have to be patched now? Why should we be grateful for the addition of a "feature" that should have already been there? Also it is not just the "hardcore" fans(I am not one by the way; I really am not even that good at the game at all)even the people who loved H6 are taking issue with H7. If you had appealed to any kind of majority the steam reviews would have showed it. Just look at Rome Total War 2 a game thrashed by its fanbase and had a poor launch but still has mixed to positive scores.
____________
"Don't resist the force. Redirect it. Water over rock."-blizzardboy
|
|
frostymuaddib
Promising
Supreme Hero
育碧是白痴
|
|
posted December 10, 2015 10:31 PM |
|
|
Gomaki will have plenty of reading material 
____________
"Occam's shuriken: when the answer is elusive, never rule out ninjas." -- Dr. Gordon Freeman (Freeman's Mind)
"lol" -- VERRIKER VON ERWINSSEN
|
|
MoritzBradtke
Known Hero
|
|
posted December 10, 2015 10:51 PM |
|
|
People want quality and not quantity means no cheating AI i hated HVI for its AI it always had more trops than me and was deadly and campaign maps always respawned
The same dull hereos with the same armies doing ecxactly the same like there was no tomorrow, THAT is boring only way to go is a clever thinking AI and not giving AI free gold and armies like u wish if they did it u all would still cry about how bad the AI plays and therefore it is still too easy to beat u cant make an AI challenging by these things why is it so hard to understand for you?
so please let em work and dont try to confuse em if u think u can do it better u can always try to make ure own AI
|
|
DIEGIS
Supreme Hero
power of Zamolxis
|
|
posted December 10, 2015 11:06 PM |
|
|
Antalyan said:
I say nothing like 'this game's development is great and Ubi did it right". Definitely no. Actually I hate the way of development "do the game asap and then fix it somehow". The only thing I am trying to say is: not to judge the whole game according to its development. The game might be finally great even with the horrible situation after the release.(I hope this is the case of H7). Consider, after releasing patches, you will not play the version 1.0, you are always interested in the final version only. Especially after some time after the development actually ends.
totally agree with you, BUT they need to fix multiplayer/sim-turns, otherwise I've seen this before....and my trust in these people is ZERO. ONLY fixing this will work for most of us; no time for more bla-bla and semi-Sim-turns; there can be only one: solution! 
____________
dacian falx behind you
-knowledge itself is power-
www.cabinet-dentaire-malaunay.fr
|
|
verriker
Honorable
Legendary Hero
We don't need another 'eroes
|
|
posted December 10, 2015 11:11 PM |
|
|
MoritzBradtke said:
so please let em work and dont try to confuse em if u think u can do it better u can always try to make ure own AI
as mentioned already, someone literally already did, and it is a hundred times better than Limbic's lol
____________
|
|
Datapack
Famous Hero
|
|
posted December 10, 2015 11:30 PM |
|
|
Is the H4 campaign still dropping in December?
____________
|
|
Karanshade
Adventuring Hero
|
|
posted December 11, 2015 07:14 AM |
|
|
MoritzBradtke said:
People want quality and not quantity means no cheating AI i hated HVI for its AI it always had more trops than me and was deadly and campaign maps always respawned
The same dull hereos with the same armies doing ecxactly the same like there was no tomorrow, THAT is boring only way to go is a clever thinking AI and not giving AI free gold and armies like u wish if they did it u all would still cry about how bad the AI plays and therefore it is still too easy to beat u cant make an AI challenging by these things why is it so hard to understand for you?
so please let em work and dont try to confuse em if u think u can do it better u can always try to make ure own AI
Hopefully someone is making sense in this concert of unconstructive critiscism. But you se, those past 20 posts or so are written by guys who are not interrested a tidy bit in heroes , nor do they want to have the game go forward. They hate the game and they just want it to fail at all cost for the only pleasure of saying ' I told you so' . It is both despicatable and not a little bit sad.Like this verriker guy who is using the photo of a guy he's never met once in his life
EDIT: Alrigh't kinda angry mood today , I might be trolling there since my post is also 100% not helping at all either.
|
|
JollyJoker
Honorable
Undefeatable Hero
|
|
posted December 11, 2015 07:59 AM |
|
|
Karanshade said:
Hopefully someone is making sense in this concert of unconstructive critiscism.
Dude, the game is out 10 weeks. The time for constructive criticism was a year+ ago, and don't believe for a second there hasn't been enough of it.
Now the game is there, and the things that can be changed are quite limited, for quite obvious reasons.
So, this isn't meant as constructive criticism, it's voicing of disappointment.
I can't imagine, that there are people who are satisfied with the game they bought for 50+ bucks.
|
|
Elvin
Admirable
Omnipresent Hero
Endless Revival
|
|
posted December 11, 2015 09:09 AM |
|
|
Karanshade said:
Hopefully someone is making sense in this concert of unconstructive critiscism. But you se, those past 20 posts or so are written by guys who are not interrested a tidy bit in heroes , nor do they want to have the game go forward.
I care about the game and want it to move forward but I still agree with most of the points mentioned.
AI? Sounds great on paper, sucks in practice.
Visuals/animations/UI? Disappointing to various degrees each.
Magic system? Disappointing.
Skill system? Even more so.
Amount and quality of unit abilities? Disappointing.
Town development? Annoyingly slow and restrictive.
(Not so) Sim turns? Oh dear.
Flanking? Next to pointless.
No the game is not bad, it's just lackluster. Hit or miss on many accounts. But the exasperating thing is that none of those are a matter of budget, the team simply didn't know what to aim for. What we liked. What irritated us. By team I do not refer to Limbic or Ubi in particular, simply whoever is responsible for the decisions that matter.
The game can still pull off a tote but only if the team starts re-evaluating their priorities and set some clear, uncompromising goals of what they should be aiming for. Limbic has proven hard-working and fast but they will need better direction. And from the looks of it, more resources and workforce.
____________
H5 is still alive and kicking, join us in the Duel Map discord server!
Map also hosted on Moddb
|
|
Elvin
Admirable
Omnipresent Hero
Endless Revival
|
posted December 11, 2015 10:10 AM |
|
|
And regarding the hardcore fanbase, what image does that bring to the team's mind? Crazy multiplayer guys who can kill 356 champions on week 2 with their hands tied? People who know the mechanics of each game by heart? People too stuck in the past to appreciate anything new?
It's like Brukernavn said. Most people have simply spent a lot of time playing the previous games and have a good idea of how things work. They influence all their friends on whether to get the latest game or not and inform them on new releases. Calling them hardcore is something that can be misunderstood.
____________
H5 is still alive and kicking, join us in the Duel Map discord server!
Map also hosted on Moddb
|
|
Karanshade
Adventuring Hero
|
|
posted December 11, 2015 10:38 AM |
|
|
Yah well JJ and elvin you were not really my frist target , and I said a bit of an angry post.
In a better world , games would come out fine at release. It s not in nowadays trend unfortunately. I can only regret it but I also came to accept it. Out of topic , I ve seen one game who was quite fine at release Mordheim but it benefited a lot from early access and a very nerdy (in a good way) community who analysed the game at every step.
But yeah in general , games tend to be good one year after initial release. And I don't think that trashing the game over and over (or limbic ftm) just for the sake of trash helps a lot during this year of improvement. Again not directed to you two since they are numerous constructive post about alternatives recently from both of you , I m thinking idea for skills and spell from Elvin , and your post about defense vs offense not being equal JJ. Those are definetly usefull posts.
The AI is not a consensus. They are trying to have a not cheating AI. I'm saying it with all the time to reflect on it possible , it s ok that the AI is not good at release. They took the hard way , they should be given some time to improve on that front. Especially since the turn calculation does not suffer too much so far (at least not on my machine ,the highest turn time I had was in the order of 25s).
I can not judge on their approach tough. As a data scientist , I'd say get player's data and analyse them through nexus but that s just me. Also as Verriker mentionned on his uniquely usefull comment , they should hire the guy who redid the AI from H5 as a consultant.
Yes I regret Ubisoft not throwing enough budget for the franchise , but I still think that Limbic is really trying to get the game straight and right. All we can do as player is help instead of feeling all powerfull because we threw some money in the game already and keep rubbing that in their face. Not helping.
Let's face a simple fact , everyone here who paid for the game knew beforehand it would not be ready ! I can not believe for one second that the ppl here are idiots with the memory lapse of a goldfish.We knew. We all knew.
And for those who post about the game without having bought it and hence are not playing it at all ....well they are section of the forum dedicated to other heroes, maybe you'd make a more informed contribution by posting there.
Now the simturn and stuff for MP, I guess it is important. I don't play turn based pvp much except maybe blood bowl so I can not speak for it. I know that the game pausing during a fight in simturn was the recurrent issue on heroes-like game. I know I could not stand it when playing MP with AoW3 (maybe it has changed since vanilla , JJ may tell about that) .
Tl dr : Stop the trash for trash sake , and get cracking on the possible solutions (or let the post about possible solution to be visible , right now you have to dive in a mud of acid idiocy to find them)
PS and OOT : Btw , the issue of offense vs defense has been around since heroes 3 and I m amazed it has not been resolved in a simple manner yet (like boosting defense scaling ) . H6 was very tentative and probably the closest to some true mathematical solution but then the clarity was completely lost (I know I could not compute the damages as f(creature damage ,my attack,his defense) , while right now I can evaluate easily both the average damage and I can appreciate how dispersive my results are gonna be. They are some deliciously nerdy maths about the homothetic behaviour of damage spread but that s for another post (...or not).
|
|
Aionb
Known Hero
|
|
posted December 11, 2015 11:23 AM |
|
|
I'm confused, are we talking about H6 or H7? I've read almost the same things several years ago.
Which reminds me: in a nutshell
|
| |
|
|
This MEGA THREAD is pages long: 
