Heroes of Might and Magic Community
visiting hero! Register | Today's Posts | Games | Search! | FAQ/Rules | AvatarList | MemberList | Profile


Age of Heroes Headlines:  
5 Oct 2016: Heroes VII development comes to an end.. - read more
6 Aug 2016: Troubled Heroes VII Expansion Release - read more
26 Apr 2016: Heroes VII XPack - Trial by Fire - Coming out in June! - read more
17 Apr 2016: Global Alternative Creatures MOD for H7 after 1.8 Patch! - read more
7 Mar 2016: Romero launches a Piano Sonata Album Kickstarter! - read more
19 Feb 2016: Heroes 5.5 RC6, Heroes VII patch 1.7 are out! - read more
13 Jan 2016: Horn of the Abyss 1.4 Available for Download! - read more
17 Dec 2015: Heroes 5.5 update, 1.6 out for H7 - read more
23 Nov 2015: H7 1.4 & 1.5 patches Released - read more
31 Oct 2015: First H7 patches are out, End of DoC development - read more
5 Oct 2016: Heroes VII development comes to an end.. - read more
[X] Remove Ads
LOGIN:     Username:     Password:         [ Register ]
New Server | HOMM1: info forum | HOMM2: info forum | HOMM3: info forum | HOMM4: info forum | HOMM5: info forum | MMH6: wiki forum | MMH7: wiki forum
Heroes Community > Dimension Gates > Thread: Q&A with Erwan Le Breton
Thread: Q&A with Erwan Le Breton This thread is 5 pages long: 1 2 3 4 5 · «PREV / NEXT»
blob2
blob2


Legendary Hero
Blob-Ohmos the Second
posted August 29, 2018 12:00 AM
Edited by blob2 at 00:12, 29 Aug 2018.

What tanked Heroes 7 was bad game design. The whole "take what's best from past games" was a buzzword, PR slogan. I doubt they were trying their best when they knew the franchise was sinking and the budget was at it's lowest. Ubisoft is some indie company that does not know how things work, yeah right...

And I'm not talking about not attracting new players at all. I'm talking about not selling the soul of a game just to get well with trends. Remind me, didn't Ubisoft do just that with their innovative and fresh ideas? You know, to attract a new player base? Cool WoW overblown design? Dark Elves? Half-naked ladies? Cool shiny artefacts you can get by spamming exp points?

(HINT: They failed miserably)

Also, read this article. It basically tells why a franchise on the brink of being tossed away has become one of the most recognized Nintendo IP's. And hey, it went back to its roots!

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
bitmaid
bitmaid


Adventuring Hero
posted August 29, 2018 01:08 AM

Oddly you just agreed with me there. Them looking back and recreating bits of what worked in the past is resistance to change. They did not innovate with h7 at all. Well I hope the franchise can come back to life, too. But if not, there almost has to be new medieval fantasy tbs to prove (at least) this slice of the genre isn't dead.

 Send Instant Message | Send E-Mail | View Profile | PP | Quote Reply | Link
Elvin
Elvin


Admirable
Omnipresent Hero
What if Elvin was female?
posted August 29, 2018 06:19 AM

Stevie said:
But honestly, I'd much rather have them fold on Heroes and start a new franchise, at which point I wouldn't buy it anyway because it's Ubisoft and no Might & Magic.

I wouldn't mind a new franchise, I might actually welcome it. Just not by ubi..

These days I am playing regalia of men and monarchs and I love its tbs combat. But the absence of hexes vexes me ^^
____________
DON'T BE A NOOB, JOIN A.D.V.E.N.T.U.R.E.

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Quantomas
Quantomas


Responsible
Famous Hero
AI Wizard
posted August 29, 2018 09:39 AM
Edited by Quantomas at 09:39, 29 Aug 2018.

bitmaid said:
The game itself is not that costly to make compared to other AAA titles, but the player base as well as talent pool for this specific genre is dwindling. [..] So as a result, programmers in training are going to be focused on the action aspect, edge detection, physics, that sort of thing.


The talent pool is not the issue, because these days way more programmers are working in the games industry than back in 2000.

Innovation does not happen automatically though. There has been solid progress in graphics because people desired to do more. They wanted more complex models requiring more polygons, and experimented with all kinds of approaches to make things work, like anti-aliasing in software and smarter ways to process the data. Graphics card manufacturers listened and implemented more advanced features into the hardware, and hand-in-hand innovation proceeded.

Not that much happened in terms of AI. There were some companies who tried to provide AI middleware but these are rudimentary at best. The AIs built for Master of Orion II and Heroes III are still the best, no one bothered to build on the skill and tech on display there. The know-how wasn't preserved as the programmers with the skills went on.

Today AI research isn't exactly an easy discipline. What I am doing is to figure out how intelligence truly works to inform sentient decision making, and to arrive at lessons how to build software that mirrors this. There is no point in going on to explain how big the hurdles are and what the benefits for the games will be, but I will say that AI has the potential to transform games, the future of games really, how game worlds become interconnected. My work isn't just a TBS AI, it is more universal and in fact a real-time AI that adjusts its own focus of processing on the fly. But it's best if you can experience this firsthand once H5 gets its new AI and you can try it out.

Regarding a future for HoMM, naturally there was always the opportunity for someone from Ubisoft to contact me in order to make something from all this skill and available tech, like a remaster of H4 or simply taking the assets form H6/H7 and building something solid on a sound technological basis from it. It would have cost less than H7. But this didn't happen and once the new H5 AI is out, this window of opportunity will close, as in the current market there are much better opportunities than those Ubisoft typically offers.

Regarding a future for a HoMM-like, maybe, it's tempting with a technological basis like the one I am building, as it adds much more to the game world than strategic competency, features like a dynamic game world in which all things are connected with choice and consequence. But it is in doubt whether that will happen.

There is something else on my radar that also features turn-based strategy, a fascinating game world and something akin to the quality gameplay we know from HoMM, and what I suspect many of you would like as well. But we will talk about it once we have the new H5 AI running on your PC.

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
blob2
blob2


Legendary Hero
Blob-Ohmos the Second
posted August 29, 2018 10:15 AM
Edited by blob2 at 10:16, 29 Aug 2018.

That is some interesting stuff @Quantomas!

bitmaid said:
Oddly you just agreed with me there. Them looking back and recreating bits of what worked in the past is resistance to change. They did not innovate with H7 at all.


They might've been so. But in my opinion first you need to get the base right, then you can think of innovation. What I wanted to show you with Fire Emblem example is that the base game didn't change, it's the things that were build around it that added value. But again, you need a solid base for that to work.

Let's face it, Heroes 7 was a low-budget experiment: it was like selling shoddy product for a lot of $$. It worked until people actually found out what they bought (not counting people who knew what was going on with the sham). What good is innovation, when a game is a complete disaster...

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
JollyJoker
JollyJoker


Honorable
Undefeatable Hero
posted August 29, 2018 10:49 AM

HoMM 3 is more like an example for the psychology of players what is supposed to be a "good" AI and what is supposed to be a bad one. Keep in mind we are dealing with a two-part AI here, one for adventure map and one for battles.

H3 simply works that way that the AI in connection with the cheats is able to halfway believably emulate human player behavior on the adventure map. That is good, because it simulates a good build-up. However, the hero skilling routines are crap, because the AI always comes with crap skills on their heroes (in blatant disregard of what are "good" and what are "bad" skills), and the battle AI is crap as well - or at least it has their blindsides.

Combined it delivers the pleasure of looking at Titans with legs of clay. Phrased differently, what will face you on the AV map looks formidable, but the player will still persevere on the battlefield, which seems to be very pleasing, psychologically.

Conversely - if the AI canNOT emulate a human player halfway believably, immersion goes away, and the human player gets the feeling they are not playing a real game.

MoO 1 (the original) can be easily compared with HoMM 3, because it has stack-based combat as well (as opposed to MoO 2), and MoO 1 has probably the best AI of them all (or should I say the best cheating/ability combo). It handles diplomacy, espionage, and framing quite well, their random AI player personalities work flawlessly - and they are pretty darn good in combat and ship-building tactics.

Since Spaceward Ho! is an NWC game a well, that has an excellent AI, while the battle part is missing (since that is solved automatically), my guess is, that the HoMM (AV Map) AI has the same "foundation".

In any case, the AI is of course a problem a designer has to solve (look at Civ 6). What use has the best game if the AI can't handle it? It will not even be useful as an MP game, if it involves battling against AI forces (like neutral stacks).

On the other hand, though, a good AI doesn't mean a bad game gets any better.

You need the game to be good and the AI to be servicanle, at the least.

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Quantomas
Quantomas


Responsible
Famous Hero
AI Wizard
posted August 29, 2018 03:03 PM

Well, nothing feels more like a time waster than a strategy game in which the opponents are pushovers or don't abide by the rules. Awareness matters, naturally.

I didn't get into playing MoO1 because it had this odd quirk that you could easily loose your ships after retreating and/or reaching an already occupied system. In MoO2 you shouldn't make the mistake to build large fleets, with fleets of 12 - 15 ships you can already conquer everything, including the battles at Antares and Orion. Arguably MoO2 would have been better received, if the game simply restricted the fleet size.

Sure, in MoO on higher difficulties the opponents had advantages, but afaik they didn't cheat. The perk system of MoO is by far the most elegant system to grade difficulty that I ever have seen. And it actually makes sense, opponents that are stronger/smarter because they are!

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
JollyJoker
JollyJoker


Honorable
Undefeatable Hero
posted August 29, 2018 03:47 PM

MoO (1) is a game I've played intensively and I actually bought it at GoG again a couple of years ago and still fire it up now and then. I played MoO 2 as well, but the original is - in my opinion - infinitely better.

For your "problem" there is a solution called "Hyperspace Communications", a level 34 Computers Technology that allows you to alter the course of your ships while they are in flight. That in combination with a good scanner Tech solves it.

However, with MoO you'll lose ships anyway, since you have only six designs (as you probably know). Once you have used up every design, if you want a new one (which you will) you'll have to scrap an old one and with that all existing ships of that design.

The stackbased combat makes for the right feeling. When the Alkaris attack you with a stack of 32K small bomb-carrying ships you can virtually feel the panic creeping into your colonists.

I agree that a cool game loses a lot with a bad AI. It's vital that game designers understand their games and code an AI in the knowledge of its limitations, helping with cheats where necessary.

I've been fooling around with AI-modding mysele for AoW3 the last year or so, and it's interesting to see the big effect small changes can have.

Still, you need a game that is worth playing in the first place, and that's essentially the problem.

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Quantomas
Quantomas


Responsible
Famous Hero
AI Wizard
posted August 30, 2018 09:33 AM
Edited by Quantomas at 09:36, 30 Aug 2018.

MoO2 is absolutely worth playing, not only because it has fixed the flaws you mentioned. If you restrict yourself to a limited fleet size, which is easy to do, it plays very much like a stack-based TBS. You could even say that it exceeds HoMM in respect of lots of tight battles, of which TBS fans always desire more. It's downright great to see your last cruiser survive a battle for an important system with a few hitpoints remaining, and you know in the next turn a new battleship will emerge from your shipyard. It's good fun to play MoO2 on its second highest difficulty, to nodge up a win for each of its different victory conditions. It's also playable on its highest difficulty, although you need a bit of luck with your starting position, and the game becomes more long-form and strategic. Plus I like how the Antarans are done.

AoW3 has a fair AI that keeps up the illusion of competent opponents. But the turn times on the largest strategic maps! Ten minutes in the later stages. If you can easily change the perceived AI strength by some modding or scripting, it's a sign that there is something wrong with the underlying AI architecture. If you are just up for the delusion though, it's fine.

But AI as I understand it is much more. A fully fledged AI will transform the game and enable services that were not possible before. AI would not only be useful for bots to speak coherently or NPCs going sensibly about their tasks and daily lifes, it is also very useful in the context of a strategy game. People are not looking in this direction now, but if they will see what's possible, they will slap their head and say, obviously, of course, that can be done!

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
blob2
blob2


Legendary Hero
Blob-Ohmos the Second
posted August 30, 2018 09:34 AM
Edited by blob2 at 09:39, 30 Aug 2018.

JollyJoker said:
Still, you need a game that is worth playing in the first place, and that's essentially the problem.

Yeah, sometimes I think all debates around Heroes 7 should amount to a sentence like that.

Quantomas said:
But AI as I understand it is much more. A fully fledged AI will transform the game and enable services that were not possible before. AI would not only be useful for bots to speak coherently or NPCs going sensibly about their tasks and daily lifes, it is also very useful in the context of a strategy game. People are not looking in this direction now, but if they will see what's possible, they will slap their head and say, obviously, of course, that can be done!


Well yeah, because people are into multiplayer atm. For many, the first question a new game releases is "can I play it in multi?". And by "play" I don't mean the option alone (cus almost every game has multi now), but a full-fledged multiplayer. People don't even need bots nowadays, they jump into the fray at once... and I even see this tendency in strategy/tbs games. Most discussions look like "the game is great, but you should work on multiplayer aspect more" etc.

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Stevie
Stevie


Responsible
Undefeatable Hero
posted August 30, 2018 10:37 AM

Quantomas said:
But AI as I understand it is much more. A fully fledged AI will transform the game and enable services that were not possible before. AI would not only be useful for bots to speak coherently or NPCs going sensibly about their tasks and daily lifes, it is also very useful in the context of a strategy game. People are not looking in this direction now, but if they will see what's possible, they will slap their head and say, obviously, of course, that can be done!


Reading on this AI tangent, I guess the next "high" the TBS genre is going to experience is when Quantomas finishes his Self-Thought AI™ and unleashes it in all of its glory upon the gaming world. ETA: Soon™.

On a serious note, AI is important, but you need a game for it. We don't have a game and we surely don't have an AI either.
____________
Guide to a Great Heroes Game
The Young Traveler

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
JollyJoker
JollyJoker


Honorable
Undefeatable Hero
posted August 30, 2018 10:46 AM

Quantomas said:

AoW3 has a fair AI that keeps up the illusion of competent opponents. But the turn times on the largest strategic maps! Ten minutes in the later stages. If you can easily change the perceived AI strength by some modding or scripting, it's a sign that there is something wrong with the underlying AI architecture. If you are just up for the delusion though, it's fine.

An AI can never be more than a delusion, because it's not a sentient opponent. Winning against it doesn't mean more than winning a fame of solitaire. (That doesn't mean, it wouldn't be more fun against a competent AI, but "competent" is a relative term in that regard.)
AoW 3 - as well as HoMM 3 - is NOT a game that is recommended to be played on the biggest possible map size (that can even be modded bigger at that in AoW 3). Games like these need to be tight affairs, and if the maps are too big the games become repetitive and the endgame to tiresome. (You can mitigate that in AoW3 by adjusting a couple of settings, but the point stands.)

I also don't agree that easy moddability is a sign of a weak or wrong underlying structure. How and what in which situations the AI prioritizes, for example, isn't a matter of how the AI evaluates the importance of something, but how important it is factually. OR, how it influences the behavior of the AI. This in turn may depend on the settings the map is based upon, so you may end up with different AIs, so-to-speak, for different sets of settings, map sizes, and so on.

The problem is, a game is all the better the less obvious winning strategies are (and the more there may be). However, that also makes devising an AI more difficult.

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Quantomas
Quantomas


Responsible
Famous Hero
AI Wizard
posted August 30, 2018 11:04 AM

Well, that is a firm statement against innovation.

It's hard to argue with that!

I would let that stand, after all it's a viable opinion, but back on topic, this is exactly the type of logic that led to at best middling games, if you don't count the bugs, poor exectution and DRM hassle, that are Heroes 6 and Heroes 7.

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
blob2
blob2


Legendary Hero
Blob-Ohmos the Second
posted August 31, 2018 12:24 PM
Edited by blob2 at 12:30, 31 Aug 2018.

I wonder how original Heroes V (NWC/3DO) AI would look like. Supposedly they wanted to build it from the ground up. Who knows, maybe if things went different we would be referring to it as a fine example these days?

Damn I just... can't get over with how things turn out. Each time I look on those renders and concept art it makes me wanna cry.

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Quantomas
Quantomas


Responsible
Famous Hero
AI Wizard
posted September 01, 2018 10:41 AM

AI design has to deal with the recursive feedback loops that pop up everywhere in processing deep and nestled variants as well as the tactical resolution of what's best. If you don't do that, the AI will be unable to adapt if the context changes, and will feel flat and unchallenging for the player. Or requires cheats, but that is no gain for players who can see through this.

Gus Smedstads design for Heroes III is elegant and beautiful to behold, but you can't just read it like a book and it reveals a lot if its mechanics only after in-depth and profound analysis. Nival build on it, and while their programmers definitely were skilled, even they failed to grasp the subtle threat mechanics. They tried to mitigate this by increasing the lookahead from 3 to 3.5 turns, but performance quickly degraded on the more open areas. Though not that big, these were necessary to produce the feel of a living world with their bigger 3D objects. NWC knew this and kept spaces tight, or isolated. After my first AI patch in 2008, I knew how to fix the H5 AI in the vein of H3. Alas life chose a different road, and seeing what will come next, I believe it was worth it.

And yes H6 and H7 had fairly good assets, not consistently, but it would have been nice if it wasn't such a waste in the end.

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Stevie
Stevie


Responsible
Undefeatable Hero
posted September 01, 2018 02:46 PM

Quantomas said:
After my first AI patch in 2008, I knew how to fix the H5 AI in the vein of H3. Alas life chose a different road, and seeing what will come next, I believe it was worth it.


I'm not really sure what you mean by this. Can you elaborate? Is this somehow related to that project you were trying to get companies interested in for funding? Because I've been looking forward to that but nothing happened for years now. An update in that regard would be greatly appreciated.
____________
Guide to a Great Heroes Game
The Young Traveler

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Quantomas
Quantomas


Responsible
Famous Hero
AI Wizard
posted September 01, 2018 04:36 PM

That referred to the work done on the AI. If you cannot see the difference it makes, once you get to play H5 with it, I would consider it a failure.

Let us talk of the future once we have something to talk about.

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Stevie
Stevie


Responsible
Undefeatable Hero
posted September 01, 2018 05:06 PM

I played Heroes 5 with your AI and the difference was more than noticeable indeed, no failure there. I was merely inquiring about the work on that prototype game, can't really recall the name... Heroes of the Ancient Order or something similar? But if there's no advance, then I guess that's just how it is. All we can do is wait, humming the good ol' tune about how time keeps on slipping into the future. 2020 is just around the corner and there's literally nothing better to talk about with regards to Heroes than Erwan's musings on what went wrong for the past decade and how there's nothing else to hope for. We're scrapping the bottom of the barrel here.
____________
Guide to a Great Heroes Game
The Young Traveler

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
imid
imid


Hired Hero
posted September 01, 2018 07:31 PM

Quantomas said:
After my first AI patch in 2008, I knew how to fix the H5 AI in the vein of H3. Alas life chose a different road, and seeing what will come next, I believe it was worth it.


hey Quantomas,

Since you find some time to get back on forums, I expect that, indeed, you're close to an end with AI for homm5 Even if ubi is going to produce a new homm, people  won't be interested. I guess, these days, many of homm fans are playing aow 3 and won't get back to a homm game of low quality. Once a good/challenging AI is done, there will be many options to construct an interesting homm-like game.

I would have fun playing against an aggressive AI team (that can act as a *team*) where my decisions are going to be indeed important. In homm 3 and aow 3 all bad decisions can be repaired by "resurrection"  
I would like a game where some very strong magic spells to depend on the level of the heroes. We need some good ideas to change the mechanics of the game: developing is important, but if I can access early in the game strong magic it's not fun. Players who can develop the heroes faster should have some advantages, but I would not like that one hero to be able to get all good stuff. In aow 3 the leaders and heroes develop sufficiently different. On top of that the independents don't let you develop , that's why is fun playing.

Did you form yet a vision on the mechanics of a future game or you have invested all the time in AI?

Cheers!



 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Galaad
Galaad

Hero of Order
Li mort as morz, li vif as vis
posted September 01, 2018 07:50 PM

imid said:
I guess, these days, many of homm fans are playing aow 3


I wouldn't know how much actually. I think AoW3 is great but not as good as H3. My main issue being the slow pace.
____________
Even Italian restaurants seemed to betray me. -Neraus

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Jump To: « Prev Thread . . . Next Thread » This thread is 5 pages long: 1 2 3 4 5 · «PREV / NEXT»
Post New Poll    Post New Topic    Post New Reply

Page compiled in 0.0525 seconds