Putting the 'role' back in role-playing games since 2002.
Donate to Codex
Good Old Games
  • Welcome to rpgcodex.net, a site dedicated to discussing computer based role-playing games in a free and open fashion. We're less strict than other forums, but please refer to the rules.

    "This message is awaiting moderator approval": All new users must pass through our moderation queue before they will be able to post normally. Until your account has "passed" your posts will only be visible to yourself (and moderators) until they are approved. Give us a week to get around to approving / deleting / ignoring your mundane opinion on crap before hassling us about it. Once you have passed the moderation period (think of it as a test), you will be able to post normally, just like all the other retards.

KickStarter Solasta Pre-Release Thread [GAME RELEASED, GO TO NEW THREAD]

Lawntoilet

Prophet
Patron
Joined
Oct 5, 2018
Messages
1,840
Some of the iconic classes missing of 5E are paladin, those got really revamped and no longer lawful goodie two shoes,the oath are very worth it . Ranger is the lovechild of wotc, a powerhouse in 5E i am surprised its not in .Bards is very versatile and OP when it comes to social game, in a dungeon delve not so much. Warlock a lackluster wizard with less options, just spam eldritch blasts, monks aren't good either.
Paladins who take Oath of Devotion are still classical LG Pally (not restricted to LG but their oath reflects that they should act like it).
Rangers are a "powerhouse?" Lol what are you talking about, Beastmaster is maybe the worst archetype in the PHB (it's that or 4 Elements Monk). Other archetypes are OK but I would not really call them "powerhouses."
Bards are strong in combat, not just social situations - Cutting Words is a great ability and Magical Secrets lets you steal the best spells from any list.
Warlocks could be considered "Wizards with less options" but also you could think of them as basically an Archery fighter, who uses EB instead of a bow, with extra options.
Monks can be useful for Stunning Fist CC or Shadow Monk sneaking around, but are probably best left to a party of 5 or 6 and a dungeon crawl might also not be the best place for them to shine.
 

Myzzrym

Tactical Adventures
Patron
Developer
Joined
Apr 12, 2019
Messages
168
Did you lurk at all before beginning to post here? I don't ask derogatorily; the above quote makes me think you may not fully understand the nature of this forum. ArchAngel has been getting shit slung at him like a champ for years. And he's flung it right back with the best of them. Having the closest thing to completely unmoderated discourse that you'll find on the internet these days means we tend to attract very opinionated and strong-willed posters, of all stripes. Calls for even-handed conversation are wasted here (again, not trying to be off-putting).

I do lurk quite a lot around here, don't worry. Don't get me wrong, I'm not asking everyone to play nice and proper - y'all can flame and troll as you want, it's your turf not mine. It's just easier for me to go through the feedback when it doesn't devolve into a poo-slinging shitfest, but eh I'll manage regardless. Early in my career I was a Producer on a P2W game, so I can tell you shit-talking & insults are pretty much par for the course - not that I enjoy them, but I've heard worse. Plus I've been playing DotA for the longest time, and good lord if you ever dipped your feet in that cesspool you know how toxic MOBA communities can be (on the plus side, I got to learn Russian)
 

Mortmal

Arcane
Joined
Jun 15, 2009
Messages
9,158
I am sorry you are disappointed we are not throwing feces around anymore, but its a thread about a real rpg, when it comes to that the codexers posting are gentlemen of refined tastes speaking about build and mechanics...
 

Abu Antar

Turn-based Poster
Patron
Joined
Jan 19, 2014
Messages
13,514
Enjoy the Revolution! Another revolution around the sun that is. Shadorwun: Hong Kong Divinity: Original Sin 2 Pillars of Eternity 2: Deadfire Pathfinder: Wrath I'm very into cock and ball torture I helped put crap in Monomyth



AI-Cover-pinned.jpg

DEV DIARY #5: AI BASICS

Greetings adventurers,

Have you ever wondered how creatures think in video games? Why did this orc decided to attack my warrior, while the other one decided to just hide behind cover? If you have, you came to the right place as I will be shedding some light on what sort of dark thoughts drive the AI of our vicious creatures.

When designing the AI system for Solasta, we wanted to make sure that it would best fit a set of objectives that crystallized what's important for us:
  • The AI must be able to choose from a large variety of activities. The 5th edition ruleset is quite expansive, and creatures can have a plethora of actions available to them at any given turn. Attacking, casting spells, getting to cover, dashing, dodging, disengaging, shoving are some examples.
  • The AI must be able to evaluate its environment properly. The verticality of our levels, the importance of light and covers must all be considered thoroughly when making decisions.
  • The AI must be easy to create and modify. We should not have to alter the code every time we want to tweak some creature’s behavior.
  • The AI system must be usable in both turn-based combat and real-time exploration. Maintaining two separate systems should be avoided wherever possible.
  • The AI must be adaptable to different creatures' behaviors. Intelligent creatures should not behave the same as creatures that are more feral and sensitive to light for instance.
tQgfZ5FkBockxm4G1TJggID5iHbzOL5aCMr6yfQx.jpeg


Making good choices in life is important, such as avoiding to get on the wizard's nerves.

All those objectives made it clear that we would use a variant of Utility AI. Utility AIs are very popular in tactical RPGs, 4Xs, Grand Strategy games where large amounts of possibilities are offered to the agents.

Basically, what this system does is giving creatures a set of decisions to evaluate. Each decision will be scored independently following a set of designed rules. From these scores the AI will choose which decision best fits its current situation. The AI will then repeat that process until it runs out of options or until it decides that ending its turn is the best decision.

But let’s skip the theory, shall we? Let me show you what it means with a simplified example.

xCiUbHQ2mSCPU0nNqEfxhcnx1VsRLHH4vJSTjrl4.png


Do AI Orcs dream of AI adventurers?

Let’s say we have a very young and inexperienced orc that we’ll call Bob from now on. What decisions could Bob be faced with during battle? Being the newbie of the group, Bob is very eager to prove himself.

First things first, can Bob make use of his axe right now? If so, which foe would be the best target? Is said foe a real threat for Bob? Bob sure doesn't want to be made fun of for attacking the weak ones after all.

What about moving? Bob can move to all those places - how does he choose the best one? Being an aggressive fellow, Bob wants to prioritize locations where Bob can HIT PEOPLE IN THE FACE.And as stated earlier, preferably someone that is a match for his brute strength. Not only that, but Bob wants to show his valor - the more foes around, the happier Bob will be. Now Bob, some would tell you that it's dangerous and not very tactically sound - but you do you buddy.

As a last resort, Bob could make a ranged attack - he does have a javelin at hand after all. But that could be seen as cowardly, so Bob only does that if he can't reach anyone. And Bob never was that good at throwing javelins, so Bob will go for someone that's easy to hit instead of a target cowering being cover.

4pWQ5nTbJPT64dAPzmzxiLQJa1rqQl9bXcdlEKkv.png


In truth, Bob is controlled by his AI overlord. But don't tell that to Bob, we don't want him to deal with an existential crisis as of now.

As you can see, there are a lot of things to consider for an AI in Solasta, and that was just a much-simplified example (no offense Bob). Giving the AI a correct set of rules, with the appropriate ways to score decisions is what makes or breaks a behavior. And the complexity does not end here as we also must balance the behaviors so that the AI does not become frustrating to fight against. That’s all the joy that’s AI design and programming!

I hoped you enjoy this sneak peek into our thought and design process, and I will see you soon for more.

Keran,
AI Overlord
 
Last edited by a moderator:
Joined
Jul 8, 2006
Messages
2,930
the most successful D&D-type games usually were those which stayed the most faithful to the ruleset.

This cannot be hammered home enough.
:lol:
Yeah, when I think of successful D&D games you know what pops into my mind? Definitely ToEE.

Not talking about commercial success. I thought that was obvious. There is right and wrong, there is good and bad. And projects like ToEE or Knights of the Chalice are successful at being far superior adaptations than any AAA streamlined casual filth.
but what about 49 billion hamburgers sold?

https://www.thomaskeller.com/tfl

French Laundry? Michelin star restaurant? lol, what a failure. Who has heard of it? Like less than .01% as many hamburgers sold and its named after a fucking laundromat and won an award from a tire store.... lol.... 49 billion sold, son..
 
Joined
Jan 14, 2018
Messages
50,754
Codex Year of the Donut
the most successful D&D-type games usually were those which stayed the most faithful to the ruleset.

This cannot be hammered home enough.
:lol:
Yeah, when I think of successful D&D games you know what pops into my mind? Definitely ToEE.

Not talking about commercial success. I thought that was obvious. There is right and wrong, there is good and bad. And projects like ToEE or Knights of the Chalice are successful at being far superior adaptations than any AAA streamlined casual filth.
but what about 49 billion hamburgers sold?

https://www.thomaskeller.com/tfl

French Laundry? Michelin star restaurant? lol, what a failure. Who has heard of it? Like less than .01% as many hamburgers sold and its named after a fucking laundromat and won an award from a tire store.... lol.... 49 billion sold, son..
Most rich people food is shit so this is a good post
 

Quantomas

Savant
Joined
Jun 9, 2017
Messages
260
Funny that devs still think an AI with a one turn lookahead would be challenging for veteran RPG players. It's easy to outwit and exploit. Utility AI is something that works for RTS or DotA.

For TB, that is if you want challenging combat not only because of enemy numbers, the AI needs to recognize the significance of taking out an opponent, in which order and essentially in which turn, and what forces it can afford to loose on the way. This requires multi-turn lookahead.

You can get away with an Utility AI if you mask it well, by meticulously tuning the encounters, environment, enemy numbers, stats and so on. But it would be a big missed opportunity.

[inconspicuously moving the bets from Solasta to Gloomhaven]
 

Myzzrym

Tactical Adventures
Patron
Developer
Joined
Apr 12, 2019
Messages
168
but what about 49 billion hamburgers sold?

https://www.thomaskeller.com/tfl

French Laundry? Michelin star restaurant? lol, what a failure. Who has heard of it? Like less than .01% as many hamburgers sold and its named after a fucking laundromat and won an award from a tire store.... lol.... 49 billion sold, son..

I'm so confused right now. It's like I entered the wrong thread.

Funny that devs still think an AI with a one turn lookahead would be challenging for veteran RPG players. It's easy to outwit and exploit. Utility AI is something that works for RTS or DotA.

For TB, that is if you want challenging combat not only because of enemy numbers, the AI needs to recognize the significance of taking out an opponent, in which order and essentially in which turn, and what forces it can afford to loose on the way. This requires multi-turn lookahead.

You can get away with an Utility AI if you mask it well, by meticulously tuning the encounters, environment, enemy numbers, stats and so on. But it would be a big missed opportunity.

[inconspicuously moving the bets from Solasta to Gloomhaven]

Well if Gloomhaven ends up even half as good as its tabletop version, I'd say it's a safe bet your making here. As for our AI, I do not necessarily think what you're highlighting is incompatible with Utility AI - it all falls down to making sure the AI weights his options properly.

That being said, I want to clear up a potential misunderstanding here - we want our AI to behave according to the nature of the creature. A group of wandering orcs is not the equivalent of a highly-trained Rainbow 6 Squad, but aggressive tribal warriors who are likely to rush the first opponent they find. On the other hand, fighting against battle-hardned mercenaries might prove to be a little more tactically challenging, as they will have a different set of priorities and may end up trying to take down your healers first.

What I mean by that is - Utility AI in itself does not prevent most of what you're suggesting (yes, even multi-turn lookahead). However, we may not have the same priorities in mind (pure combat challenge seems to be a priority for you, varied behaviors is more important for us so that fights don't end up being "same AI different skin")
 
Joined
Jan 14, 2018
Messages
50,754
Codex Year of the Donut
For TB, that is if you want challenging combat not only because of enemy numbers, the AI needs to recognize the significance of taking out an opponent, in which order and essentially in which turn, and what forces it can afford to loose on the way. This requires multi-turn lookahead.
Where did you get the idea that utility AI is incapable of this? It's a fairly generic concept.
 

Quantomas

Savant
Joined
Jun 9, 2017
Messages
260
The difference between Utility AI (UT) values and path sequencing for example is that UT values are based on what you know at time of evaluation, whereas path sequencing uses the lookahead for path (variant) comparison and selection.

There are a number of reasons why predicting variants based on what you know at any point of evaluation (as in UT) is unsatisfactory, most importantly that it doesn't differentiate between the possible actions opponents can take (in the lookahead). I.e. arriving at an UT value that reflects different possible actions a few turns ahead, e.g. for an orc attacking the next melee grunt and then moving on as opposed to alternatively charging a wizard further away directly, is tricky at best. In principle you can run UT predictions for multiple turns but the inherent problem of the missing lookahead is not solved and the result is unefficient and poor. If you want a more involved answer, in principle UT values do not reflect recursive dependencies, i.e. that the proper selection of actions depends on the actions performed by other actors. In the lookahead you have to track these.

On the other hand path sequencing can solve the problem.

Hence my comment that using UT is very much like an AI without lookahead. Especially if you look at the post and the tools shown in the Solasta blog which conspicuously looks like a one-turn-lookahead. I know some people who use the very same tools for MOBAs.

You should think twice before labelling a post Fake News if you do not know your stuff.
 

_Vic_

Educated
Joined
Aug 15, 2019
Messages
56
I always wanted to make a "Belkar´s halfling rage jumping attack"
oots0176.gif


In other party-based games like POE your barbarian and rogue can jump, in DaO you can do somersaults, the assassin and barbarian in diablo 2 can also jump and kick; but in fact, you are really playing in 2D, the jumps are only aesthetic.

The game sounds promising to me. Just keep it up and do not care about make the game "simpler" or "cater to a large audience". That does not usually go well in the long run.
 

LESS T_T

Arcane
Joined
Oct 5, 2012
Messages
13,582
Codex 2014
https://www.enworld.org/threads/a-chat-with-solasta-5e-video-game-developers.666696/

A Chat With 'Solasta' 5E Video Game Developers

About six weeks ago, I mentioned an upcoming 5E-powered video game called Solasta: Crown of the Magister. Developers Tactical Adventures have been in touch, so I asked them a few quick questions about the upcoming game, wich will be coming to Kickstarter soon.


What can you tell us about the world in which this game takes place?

We will be giving more information on the world fairly soon (we plan to introduce our lore during our Kickstarter). We cannot reveal too much, as we want discovering the world of Solasta to be part of the final experience. That being said, Solasta is a High Fantasy universe where players will find most of their favorite playable races, with one notable caveat: they may behave / interact differently than you think, as an ancient cataclysm forged Solasta to its current state.

You might also be interested to know that one of writer is Graeme Davis, one of the original author of Warhammer Fantasy Roleplay back in 1986 - and that he won the Best Writing ENnies at Gen Con this year!

You mention some character types such as a Darkweaver Rogue and a Mountaineer Fighter. Which 5E classes and races can we expect to see in the game?

We would like to include the races and classes found in the System Reference Document of the OGL – though we may not be able to include all of them for our first game. Note however that the SRD only details one subrace per race and one archetype per class, so we will be adding more of those according to our own lore - hence the Darkweaver & Mountaineer. currently we have 4 races / Classes available on our Website: Human / Dwarf / Elf / Halfling, and Fighter / Cleric / Wizard / Rogue. More will be added through Kickstarter.

How much of the 5E ruleset made it into the game, and how much did you need to change for the new medium?

I actually wrote an article about that particular topic. Long story short I would say that we want to include as much as possible of the ruleset. Some successful games included 70%-80% of the rules, we want to try to go as high as 90-95%. Not because we are fanatics, but because these rules (such as flying and lighting) create interesting gameplay situations. We do not want to pretend to be smart and replace the rules with our own just because. We want to limit rules adaptation to when there is no other choice for the video game media. For example, when vision distance represents 20 monitor screens (longbows…), we have a problem and must adapt it so that the gameplay and visual feedback stays consistent.

Is this game a single-player game or just multiplayer?

We want to focus on building a solid single player campaign first. Multiplayer would be great, but we are a small independent team and we have to remain realistic.

Is the combat turn-based or real-time?

It is turn based, as in my opinion it is the sole way to properly adapt 5E rules. For example, main action, bonus action and reactions do not make sense if you have a real time with pause simulation. Exploration are real time though, with reactions pausing the game.

You mention the term "verticality" a few times when describing the game. Could you give us a couple of examples of what you mean by that?

Our Gameplay Director has written an article about this topic. In a nutshell, verticality encompasses a number of 5E features, ignored or poorly adapted in other games: climbing / jumping / flying, shoving enemies, using ropes or climbing kits, and much more. To enable all these cool features, we built a 3D simulation based on cubes where most games use a plane which is extruded vertically in some place.

This game uses the 5th Edition ruleset as its core; are you utilizing the Open Gaming Licence?

Yes absolutely. This is a great initiative, which allows all fans to develop or play 5e content. The content of the System Reference Document (SRD) is massive, as it contains 400 pages of character classes, race, monsters, spells, items, etc.

Could you speak a little on the challenges involved with transcribing the 5E rules to a video game environment?

When I started playing 5E on tabletop, apart from having a blast with my friends, I realized that this was the best edition so far to adapt on video game. It surprised me that no one managed to do it correctly, so I guess this was the inception for the creation of the new studio. As I started developing our prototype, everything seemed to make sense and was easily integrated. The only issue we had so far, and which required some adaptation, concerned the lighting and vision system. When playing a team of four adventurers, a video game applies some constraints in terms of art direction, feedbacks, etc. (We wrote an article about that particular issue here).

There have been very few video games which leverage rulesets like 3.x, 4E, 5E, and so on. Do you have any thoughts on why that is?

I would say it requires passion. There are many great game designers / developers, but sometimes reskinning your game with a license and fighting with the franchise to adapt it will disappoint everyone. I’ve been playing D&D for 30 years, I am not trying to reinvent it - we are doing our best to replicate the same experience, but in a video game.
 

pm_675

Arbiter
Joined
Mar 14, 2017
Messages
228
Myzzrym , what will the max level be in the game?. The full 20 levels or less?
 
Last edited:
Joined
Jan 14, 2018
Messages
50,754
Codex Year of the Donut
The difference between Utility AI (UT) values and path sequencing for example is that UT values are based on what you know at time of evaluation, whereas path sequencing uses the lookahead for path (variant) comparison and selection.

There are a number of reasons why predicting variants based on what you know at any point of evaluation (as in UT) is unsatisfactory, most importantly that it doesn't differentiate between the possible actions opponents can take (in the lookahead). I.e. arriving at an UT value that reflects different possible actions a few turns ahead, e.g. for an orc attacking the next melee grunt and then moving on as opposed to alternatively charging a wizard further away directly, is tricky at best. In principle you can run UT predictions for multiple turns but the inherent problem of the missing lookahead is not solved and the result is unefficient and poor. If you want a more involved answer, in principle UT values do not reflect recursive dependencies, i.e. that the proper selection of actions depends on the actions performed by other actors. In the lookahead you have to track these.

On the other hand path sequencing can solve the problem.

Hence my comment that using UT is very much like an AI without lookahead. Especially if you look at the post and the tools shown in the Solasta blog which conspicuously looks like a one-turn-lookahead. I know some people who use the very same tools for MOBAs.

You should think twice before labelling a post Fake News if you do not know your stuff.
I'm not sure you understand what you're proposing.
How do you propose to use look-ahead on a game as complex as D&D? The state space for D&D combat is simply massive, it would result in combinatorial explosion making it impossible to search more than a couple moves ahead(And I'm being generous here.)
But here's another issue: D&D is not deterministic unlike Chess or Go, nor is it perfect information(nobody likes a cheating AI.) That means that your AI has no way to know what spells a spellcaster has memorized, what feats the player has learned, what items they have in their inventory, so on and so forth.
And I suspect you mistranslated whatever "path sequencing" is supposed to be because that term doesn't mean anything.
 

Quantomas

Savant
Joined
Jun 9, 2017
Messages
260
Are you on the payroll of Tactical Adventures? You have ridiculed every critical post in this thread.

Proper critique can help you to make a game better. There are still a few Codexers here who can be helpful. But there are many more who pretend strong language is a sign of competence.

As regarding advanced AI, it is not discussed openly on the web like graphics for instance. Path sequencing had been used to great effect by Gus Smestad for Heroes III at NWC. You would need to look at the source or talk to a pro who knows the relevant techniques from other sources. Likewise there are techniques to cope with the exponentially growing variants for lookahead computation. You can't process these all.

These days many indie devs seem to believe that they can do AI on a shoestring budget. But dealing with the recursive dependencies typical for AI is truly complex and requires skill. That is the reason why finding a game with good AI is rare. But if you want to do a turn-based game with real depth, good AI will truly enhance it. As I said to Myzzrym you can get away with it if you invest in encounter design, but if you want to do it properly you need to hire a pro. If an interviewee talks to you about neural networks or how Utility Theory can be used for TB games, you know he is clueless.

varied behaviors is more important for us so that fights don't end up being "same AI different skin"
Another reason to do things properly. Competent AI lets you shape behaviours, like the aggressivity of an orc or the stealthy demeanor of a spider that will change to rush once it thinks its victim is in range. The games with samy AI typically are the incompetent ones built with UT and/or weak lookahead.
 

_Vic_

Educated
Joined
Aug 15, 2019
Messages
56
These days many indie devs seem to believe that they can do AI on a shoestring budget. But dealing with the recursive dependencies typical for AI is truly complex and requires skill. That is the reason why finding a game with good AI is rare. But if you want to do a turn-based game with real depth, good AI will truly enhance it. As I said to Myzzrym you can get away with it if you invest in encounter design, but if you want to do it properly you need to hire a pro. If an interviewee talks to you about neural networks or how Utility Theory can be used for TB games, you know he is clueless.
XDD We are NOW absolutely convinced of the need for a pro for this. Just give them your CV already and your tariff for a freelancing job. It is a free forum, we do not mind. Good sales pich, congrats!
 

Quantomas

Savant
Joined
Jun 9, 2017
Messages
260
Nah. Indies are hard to work with and don't pay well.

I trust Solasta will do just fine. But you have to instill a sense of realism in indie devs.
 

As an Amazon Associate, rpgcodex.net earns from qualifying purchases.
Back
Top Bottom