• We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
I’m obviously an MP addict and I’d like to better understand the reasons why you’d rather stick with SP, given that you’re all here complaining about how bad the AI is.

What is it you don’t like about MP and what is it you’re looking for?
 
I’m obviously an MP addict and I’d like to better understand the reasons why you’d rather stick with SP, given that you’re all here complaining about how bad the AI is.

What is it you don’t like about MP and what is it you’re looking for?

I don't like playing strategy games against people online. A board game is fine but too many chances to cheat online. A board game, I can see you. Not online. I just don't find it fun at all.

When playing a strategy game, I am usually playing something I want to play alone. I want to experiment with tactics, different empire building strategies. MP tends to devolve into all out skirmishes. I am not interested in that. Yeah, you may find some people interested in longer games but that is rare. Your pace is not my pace, and I want to play at my pace.

For me, a fighting game is fine for competition. So is a FPS. I am just personally not a fan of hotseat or simultaneous turns.... Hell, it is hard enough to get people to play a 6-8 hour game of the Twilight Imperium boardgame. There are not many people who want to sit and play the length I do.

What am I looking for? That is easy. A competent AI that can play the game. There are plenty of games where the AI can at least compete. I want an AI that knows how to cross water (Which should not be that hard.) I want an AI that knows how to efficiently build up an empire. I want AI personalities that approach the game differently. I want a competent AI. Not much to ask.

And it is not complaining. The AI deficiencies are quite apparent. It is pointing out something that should have improved between WL1 and WL2. I mean, it could be worse. It could be Age of Wonders III where the AI was DESTROYING its own empire if you came even close to it with a big army.
 
What am I looking for? That is easy. A competent AI that can play the game. There are plenty of games where the AI can at least compete. I want an AI that knows how to cross water (Which should not be that hard.) I want an AI that knows how to efficiently build up an empire. I want AI personalities that approach the game differently. I want a competent AI. Not much to ask.

Unfortunately it is a lot to ask :/

If it was easy, more tbs games would have good AI.

Consider that Civ5, a game with a massive budget/staff in comparison to most smaller 4x titles still has stupid AI problems. The AI, even on the highest difficulty levels, simply won't pursue certain victory conditions because doing so with its huge production bonuses would be a near certain loss for the player (e.g. spending its huge gold reserves to buy all city states and win diplo victory). It even has the exact same problems that Warlock had a patch or two ago where you could ally, trade off a ton of stuff for its massive gold reserves, declare war severing the agreements and then use the huge gold pile to churn out dozens of lategame units from all of your cities and crush it militarily.

And it still sucks at naval combat and (largely) tac combat, especially when terrain chokepoints are involved.

Age of Wonders 3 is another good example of that same issue - hamstringing the AI because if it played with its resource advantages, it renders an aspect of the game unusable by the player. Specifically dispelling - with its production bonuses to mana, the AI could permanently and continuously shut down any strategic spells you tried to cast and maintain, rendering an entire set of spells totally useless for the player. Their solution? Giving the AI a chance that it would not dispel on any given turn. That is, it simply randomly ignores your strategic spells.

The sooner devs abandon the conceit of having the AI play the game the same way as the player, the faster we'll have challenging tbs games. It'd help if players weren't so adamant that the AI play the same game as well :laugh:

See the game AI War for an example of how completely tossing out the concept of having the game 'play fair' can result in a really tough, really fun experience.
 
Unfortunately it is a lot to ask :/

If it was easy, more tbs games would have good AI.

Consider that Civ5, a game with a massive budget/staff in comparison to most smaller 4x titles still has stupid AI problems. The AI, even on the highest difficulty levels, simply won't pursue certain victory conditions because doing so with its huge production bonuses would be a near certain loss for the player (e.g. spending its huge gold reserves to buy all city states and win diplo victory). It even has the exact same problems that Warlock had a patch or two ago where you could ally, trade off a ton of stuff for its massive gold reserves, declare war severing the agreements and then use the huge gold pile to churn out dozens of lategame units from all of your cities and crush it militarily.

And it still sucks at naval combat and (largely) tac combat, especially when terrain chokepoints are involved.

Age of Wonders 3 is another good example of that same issue - hamstringing the AI because if it played with its resource advantages, it renders an aspect of the game unusable by the player. Specifically dispelling - with its production bonuses to mana, the AI could permanently and continuously shut down any strategic spells you tried to cast and maintain, rendering an entire set of spells totally useless for the player. Their solution? Giving the AI a chance that it would not dispel on any given turn. That is, it simply randomly ignores your strategic spells.

The sooner devs abandon the conceit of having the AI play the game the same way as the player, the faster we'll have challenging tbs games. It'd help if players weren't so adamant that the AI play the same game as well :laugh:

See the game AI War for an example of how completely tossing out the concept of having the game 'play fair' can result in a really tough, really fun experience.

I am not saying a game to play fair. I want a game that has a challenging AI.

And I don't think it is that hard. I just think they have been going about it wrong for a long time and don't want to try new techniques. Civ 5 has been around for years now and it tends to be pretty good at playing the game, despite the flaws you pointed out.
 
I don't like playing strategy games against people online. A board game is fine but too many chances to cheat online. A board game, I can see you. Not online. I just don't find it fun at all.
Not sure what you mean here, could you develop? How can you cheat at W2?
When playing a strategy game, I am usually playing something I want to play alone. I want to experiment with tactics, different empire building strategies. MP tends to devolve into all out skirmishes. I am not interested in that. Yeah, you may find some people interested in longer games but that is rare. Your pace is not my pace, and I want to play at my pace.
I agree that it is hard to find players with similar interests and skill at the moment but I and a few other veterans are personally trying to make it more enjoyable for new players that come to play MP games. Korog for example is fairly into his long games on large map. I usually prefer shorter games but we started a dual on Large map because I was curious to know how it's like. I would only invest in a longer game with people I already know and who I'm comfortable with. I usually sample new players on small map to see if they have enough skill to make it worth my time on bigger maps. Given the current size of the community MP is definitely not a consumer oriented experience and feels more of taking part in a community. In this context it's good to keep an open mind and be ready to compromise.
For me, a fighting game is fine for competition. So is a FPS. I am just personally not a fan of hotseat or simultaneous turns.... Hell, it is hard enough to get people to play a 6-8 hour game of the Twilight Imperium boardgame. There are not many people who want to sit and play the length I do.
I would say that all serious Warlock players are willing to commit 6-8 hours+ for a game. That's probably what defines a serious player in the first place. Then again there is no mechanism in place to identify these players. It can also be a bit of a shock for players who are new to the genre and there is a certain amount of education that they need to undertake.
What am I looking for? That is easy. A competent AI that can play the game. There are plenty of games where the AI can at least compete. I want an AI that knows how to cross water (Which should not be that hard.) I want an AI that knows how to efficiently build up an empire. I want AI personalities that approach the game differently. I want a competent AI. Not much to ask.
xeren has already summed up all I had to say about this.
Thanks for the AI-War plug, looks very interesting btw.
And it is not complaining. The AI deficiencies are quite apparent. It is pointing out something that should have improved between WL1 and WL2. I mean, it could be worse. It could be Age of Wonders III where the AI was DESTROYING its own empire if you came even close to it with a big army.
I agree entirely about the AI being underpowered as well and all that. I've been trying to make suggestions to improve AIs for years, across a variety of games. I've also been consistently disappointed. Warlock is the very reason why I'm back to actually caring about a game and getting involved in a community. It's got great potential as a multiplayer game. It's only up to us to make sure this gets on top of the agenda with the devs/producers. IMHO investing in multiplayer will pay off far better than trying to fix a broken AI because this is what sets Warlock apart from the competition.
 
You can cheat at anything online with a PC and programming skills. Even easier with the mod tools. Increases in production, early access to late units, etc.... I don't care how good you have a game locked up, with the right tools and want, you can cheat in online play with little problem.

And with a community that is small, I am betting the anti-cheating mechanisms are not the best.

So yeah, cheating online. I mean heck, people can cheat on consoles. Imagine how much easier it is on a PC.
 
You can cheat at anything online with a PC and programming skills. Even easier with the mod tools. Increases in production, early access to late units, etc.... I don't care how good you have a game locked up, with the right tools and want, you can cheat in online play with little problem.

And with a community that is small, I am betting the anti-cheating mechanisms are not the best.

So yeah, cheating online. I mean heck, people can cheat on consoles. Imagine how much easier it is on a PC.
I don't know the nuts and bolts of how the game is implemented but I thought that the game now checked that all clients have the same files versions and mods when starting new games. I can see how the host can cheat but not the connected clients. I don't know if the clients even have access to the entire environment. The assumptions you're making for FPS and such don't necessarily hold for 4x games.

Also the advantage of a small community is that it's easy enough to uncover and shame cheaters. If I got caught cheating at Warlock for example that would be the end of my online games.

I don't think cheating is an issue at the moment but I agree that steps should be taken to prevent it where possible. Perhaps the devs could shed some light as to what degree of security we can expect from the current implementation. Do clients have access to the global game state? Do servers check that client actions are valid (i.e have enough money to recruit a unit, etc..)?
 
Can I just state that my original point was not simply that the AI in the game is poor, but broken. I'm not looking for anything spectacular from the small dev team - AI on a par with the AI in W1 would do, tbh. Right now, it's just painful to play, where W1 was a bit of fun at least.
 
Can I just state that my original point was not simply that the AI in the game is poor, but broken. I'm not looking for anything spectacular from the small dev team - AI on a par with the AI in W1 would do, tbh. Right now, it's just painful to play, where W1 was a bit of fun at least.

I agree. This is so true for Sandbox mode at least. W1 AI is much better, or at least can be.
 
@Man-At-Arms: Jörgen is well aware of the AI thread that you linked. Thank you for putting in your feedback.
@BjornB:
That's good to read, but hope to see some real improvements in next weeks and months. I think when you read the forum and official reviews and critics, you must know that AI is big point. So I hope you address all these basic AI problems in the list.


@All:
And Yes, some of us want to stick to normal gaming offline-matches against AI. :) And I think there will be a big crowd of player which actual want to play challenging matches versus AI. And like I said hundred times, we are not talking about super special player strategy or tactics and strategy in detail, we are talking about bad AI behavior and that AI is not understand challenging basic game-play and game-engine. And the AI is not needed to be good as best player, that's not the point, but AI should not look like doing stupid things and sacrificing all the time. See my big collection posting for details. There is no talking about "best special strategies" or about "what best player would do". It's really basic challenging gameplay and game-engine understanding.

When somebody think this list is not complete, and I forgot big AI problems, good ideas oder threads, please let me know. But I will only consider ideas without getting AI cheating, and without making changes in game design that the gameplay fit only actual AI gaming without being good for overall gameplay. Some feedback would be nice.
 
Last edited:
I’m obviously an MP addict and I’d like to better understand the reasons why you’d rather stick with SP, given that you’re all here complaining about how bad the AI is.

What is it you don’t like about MP and what is it you’re looking for?

It's quite simple: they can't stand losing. The first response you got was the excuse of cheating. Just like why people enjoy lots of random factors. They need to blame something else for their loss other than themselves. It bothers me to no end that what people really want is a role playing game, not a strategy game. They want the game to be 8 hours or longer and they ultimately want to win for doing little more than putting in time. A game will likely never appeal to both strategy gamers and rpg gamers, and, unfortunately for the strategy gamers, the number of rpg gamers completely eclipse the strategy gamers. It should be no surprise to anyone why the market is dead.

The best thing a TBS game could do is simply not have AI and force everyone online. This would cull your market to just the strategy gamers and your reviews will be better.
 
The best thing a TBS game could do is simply not have AI and force everyone online. This would cull your market to just the strategy gamers and your reviews will be better.

Please don't. I do not want my spare time to depend on others, but me myself. Otherwise I would play chess more.

And nothing prevents SP game to be challenging or strategic. Any AI deficiency can be compensated by asymmetrical challenge. There is a reason why is Deity in Civilization games challenging despite not so stellar AI.
 
It's quite simple: they can't stand losing.

I am not sure if it's necessary to talk about, why a bunch of player want to play this game in SP. I think it's normal that not everybody want to play games in MP. MP and SP gaming is different in many things

As I said, with some additions...

Competing in MP is an other style of competing, other playing style is needed,
- other playing time management is needed,
- other kind of frustration with winning and losing is coming up,
- more questioning balancing is coming up (and frustrating about losing games thinking of balancing issues)
- human relations and behavior of gamer could be frustrating (Yes, don't have to be)
- It's a completely other style of enjoy the game, and game-play / strategies needed to win the game

Yes, SP is more about easily gaming with your playing-time management exactly like you want to. And yes, you are right, it's more about winning. But also SP player want a bit of challenge, SP player don't need a AI which beats the players all the time and is good as best players, but it's more about challenging enough to not looking like doing stupids things all the time... see my last post. It's simply more fun in SP when AI play a good style, use all game features and seems to understand basic challenging game engine and needed strategy.

I think all MP-player can talk about the balancing or MP related issues, and the SP-player can talk about the SP issues like bad AI behavior. And let's talk about AI related issues here in the thread. :) Or should we do talking about "why not playing SP" in MP-related issues (or feature) threads? :) We can make a new thread with "SP vs MP", but we are in the "AI non-existent" thread.

- For Example, what AI issues do you think are most important to get fixed?
 
I am not sure if it's necessary to talk about, why a bunch of player want to play this game in SP. I think it's normal that not everybody want to play games in MP. MP and SP gaming is different in many things

I entirely agree with you, sorry if I came across as belittling SP experience, that wasn't my intent. I used to play SP a lot myself and I've done my share of suggestions to improve AI, implementing house rules and even modding in Warlock1 (and plenty of other games for that matter). Now a year later, things have gone even worse according to you SP players (personally I thought the AI was performing a bit better if anything, at least in Exiled). Again I don't want to shatter your dreams of a decent AI that doesn't make stupid mistake but I doubt you'll still be hopeful in a year's time at this rate.

Again don't take me wrong, I've enjoyed SP immensely, or I wouldn't be here now talking about it. In my opinion there just comes a point where no matter how much you stack the odds again yourself, the AI is so full of exploits it's not even fun. I think that when players reach that point, they're ready to move on to MP (or try another game). To me it's the normal lifecycle of any strategy game player. That's all I'm saying. Think about it.
 
Hi all,

After playing through the shard game, I went straight to sandbox on a large standard map, Impossible difficulty. It was... not impossible.

I don't want to whine about it too much, and I don't want to just drone on about how you should "Improve AI!", because you're already doing that.

I do have some other suggestions, mostly in the form of the inclusion of new game options:


1: Option to disable diplomacy
Add an option to disable diplomacy between you and the AIs on hardest difficulty, where you would be in a permanent state of war with them all the time (This option could give you a lot of points when you finish the game). This wouldn't just make the game more difficult, but also rid you of the AI constantly pestering for you for an alliance, or trading that damn 67 gold for 45 mana every damn turn.
Alternatively, you could add a "diplomacy difficulty level" instead of a downright disabling option. This would have the AI be more adverse to being allied with you, and more likely to ally with other AI against you.
Additionally, when playing multiplayer coop, add the option of creating AI teams. Maybe have all allied against the human team, or teams of 3 AI vs 3 AI vs Human team, something like that. (I haven't actually been able to play MP yet, so I don't really know how possible this is to implement.)

2: Monster Spawns
The AI is having a lot more trouble with this than I am. Maybe add an option that gives the AI a "soft start" somehow, making them able to grow in strength instead of being totally subdued by monsters, which seems to be the case of some of the AI players in the "impossible" game I just played.

3: Turning off water
Simple. The AI has problems with water, so add an option to disable ALL water in your sandbox map.

4: Give AI advantages
In most games, such as Civ, the AI gets great benefits on harder difficulties (more income, less unhappiness, etc.), so why not do the same in this game?
The most obvious here would to give the AI a higher City Limit than you. This could be added as an option box, again giving you more game points if you win on harder settings.
Additionally, the AI could be given all sorts of other advantages, such as gold/food/mana income, or unit strength. This kind of "direct/obvious AI cheating" might not be as fun to play against as some of the other things, but really... ANYthing to make the game more challenging.


I know very little about game design, so I don't know how hard these things would be to implement, but... in all my ignorant glory, it sounds kind of simple.


The best thing about these suggestions is that they are optional, so if someone doesn't like playing with them, they don't have to turn them on.


So... how about it? :)
 
AI already gets numeric bonuses, look at their gold/mana in an impossible game.

Their deficiencies are a combination of poor tactical/strategic ai, and general game design issues that cause them severe problems (water, terraforming, chokepoints in outplanes mode, efficient monster lair hunting, and especially hero/high tier unit abuse with buffs and perks).

I think the various major issues and a lot of minor ones have been pointed out enough on the forums, just a question of if/when they patch them.
 
AI already gets numeric bonuses, look at their gold/mana in an impossible game.

Their deficiencies are a combination of poor tactical/strategic ai, and general game design issues that cause them severe problems (water, terraforming, chokepoints in outplanes mode, efficient monster lair hunting, and especially hero/high tier unit abuse with buffs and perks).

I think the various major issues and a lot of minor ones have been pointed out enough on the forums, just a question of if/when they patch them.

My guess is "soon".
Warlock 1 was laying out gameplay core, they've added diplomacy after some time, campaign system (armageddon) and multiplayer.
Warlock 2 is fleshing out their previous work while at the same time they added ingame store (mtx feature), expanded on campaign system (exile), created modding tools.
 
AI already gets numeric bonuses, look at their gold/mana in an impossible game.

The problem is that current economic bonuses do not translate to increasing AI strength per tile. Essentially AI needs to have better units THEN player on harder difficulties in order for players to get challenged.

Brains over brawns. Add more brawns to AI.

AI goal in Warlock should be to offer greater challenge that spawned mobs. Which is currently not the case.
 
The problem is that current economic bonuses do not translate to increasing AI strength per tile. Essentially AI needs to have better units THEN player on harder difficulties in order for players to get challenged.

Brains over brawns. Add more brawns to AI.

AI goal in Warlock should be to offer greater challenge that spawned mobs. Which is currently not the case.

I remember once in Warlock 1 one great mage got hold of donkey knights on impossible difficulty. It took me around 2 hours to mow down his units and slowly advance to finally capture and burn town that had this resource as it was little behind his lines. At the same time I was at war with 2 others ai and had to cover my defenses on south border (donkey knights were coming from east). It was one of two my best matches ever in this game.

If enemy would by higher tier units than spearmens and archers it would be incredible improvement. I rarely see them use anything else.