AI WARS! » Forum Post by Frogboy

Frogboy

Join Date 03/2001

+2417

Quoting Wintersong,

reply 48

Quoting Frogboy, reply 42There is a lot of internal dislike of the current way ships are handled.There is not much to dislike. Boat that explores. End.

No.

You have to find a beach. Then you have to build a city. Then you have to build a harbor. Then you build a ship. Then you have to load units onto the ship. Then you have to find another beach to unload the units.

That's a lot of work.

In Civ V, units just walk across the water (they turn into boats at the water's edge).

Successfully updated karma reason!

It's a lot of work but has a huge pay off: being able avoid entire kingdoms or wilds of enemies.

Turning units into boats once they touch water is ok if you want to make a casual game. Please don't take the easy way out.

edit: though I do agree having to build an entire city next to water is rather ridiculous. Perhaps allow for pioneers to construct small harbors? City harbors could offer x2 build speed or something to give players an incentive to also make a coastal city if they wish.

Successfully updated karma reason!

Heavenfall

Join Date 07/2008

+441

For shits and giggles I tried introducing a mod mechanic that would give the AI comparable levels of metal to that of their materials. The AI still produced lower quality gear, even when higher quality was available.

So I assumed a theory that the AI would always produce units when it could, with the highest quality. This would lead to it perpetually being out of one resource, as there would always be a limiter. But this was not so. Even when I made sure the AI had over 20 000 metal and materials, and 100k+ gildar, it still continued building units with leather armor instead of master heavy plate, which I know it had access to. Indeed, not even with such abundant resources did the AI ever construct any other unit with more than one warrior (squad/group/party/raid), except diplomacy units.

There seem to be a disconnect here between what the AI can do and what it does do. Even when in a position to be able to construct the right units, it continues to produce the wrong units. I am oversimplifying, of course.

The AI, it seems, need to be able to "save" resources until a suitable unit can be produced, especially middle or late game. The AI badly needs to avoid wasting resources on poor unit management (single units with only a weapon and leather gear).

Furthermore, not once have I seen AI make use of training units with higher quality. This is also a huge waste on behalf of the AI, as it is one of the best ways to save resources.

Successfully updated karma reason!

Frogboy

Join Date 03/2001

+2417

Quoting Heavenfall,

reply 53

For shits and giggles I tried introducing a mod mechanic that would give the AI comparable levels of metal to that of their materials. The AI still produced lower quality gear, even when higher quality was available.

So I assumed a theory that the AI would always produce units when it could, with the highest quality. This would lead to it perpetually being out of one resource, as there would always be a limiter. But this was not so. Even when I made sure the AI had over 20 000 metal and materials, and 100k+ gildar, it still continued building units with leather armor instead of master heavy plate, which I know it had access to. Indeed, not even with such abundant resources did the AI ever construct any other unit with more than one warrior (squad/group/party/raid), except diplomacy units.

There seem to be a disconnect here between what the AI can do and what it does do. Even when in a position to be able to construct the right units, it continues to produce the wrong units. I am oversimplifying, of course.

The AI, it seems, need to be able to "save" resources until a suitable unit can be produced, especially middle or late game. The AI badly needs to avoid wasting resources on poor unit management (single units with only a weapon and leather gear).

I've seen companies of doom hammer, plate mailed units plenty of times.

Did they have the tech to use that stuff? Had they built the prereq buildings?

Successfully updated karma reason!

Heavenfall

Join Date 07/2008

+441

Yeah, I made sure they had the tech. I see the enemy use strong weapons, but they are always limiting themselves to the weaker versions of armor.

Edit: The building, yes. Well, if they didn't, they should have.

Edit2: So as to not sound all negative, I've also noticed two awesome things the AI does - it uses Death Ward to protect significant assets - and it seeds new towns with pioneers next to free resources. That last bit is actually quite excellent, a strategy I'll have to adopt.

Successfully updated karma reason!

Heavenfall

Join Date 07/2008

+441

Example: Turn 355. AI Relias has conquered the map. He has trained some master heavy plate units, so there is no question he has the tech. His resources are 46k materials, 106k metal, 1.7mil gildar. About half the cities have built a command post. There are NO units with more than one warrior on the map. Also, no AI-designed unit is wearing a helmet.

Edit: 90% of the units trained are wearing light plate armor. I'd say another 9% are diplomacy units, and the remaining 1% are the well-designed (solo) units.

Successfully updated karma reason!

You only want the Ai to build companies until it does.

Successfully updated karma reason!

Heavenfall

Join Date 07/2008

+441

Hah. No, I really am concerned that the AI needs x8 times the health in order for me to face difficulties fighting it (twice that of ridiculous).

Successfully updated karma reason!

Quoting Frogboy,

reply 7

Quoting Ratatosk7,
reply 1
Stupid question here. Are AI changes added each patch like crashfixes or are they only added when specifically mentioned in changelog?

I'm so glad Stardock has such great post-release support. I wish I could go a few months in the future and play an Elemental I won't even recognize, while a lot of developer/publishers out there simply drop all support post-release as if they never were involved in the game in the first place

I do lots of tweaking to the AI with each build but we don't publicize the specifics since AI strategies are highly confidential. For instance, in last night's build, the AI unit work should have been noticeably better.

In tonight's build, players should notice the AI isn't quite as quick to go to war with you and are much more willing to target each other.

Well "a lot of developer/publishers" also tend to release finished games with concrete systems and a semblance of balance. As far as I am concerned all of these post-release patches we are getting are simply an extension of development. I'm getting a little bit annoyed by posts saying how great post-release support is...yeah it's great because it's easy to support a game that was never finished in the first place. I'd wager one of the expansions (perhaps one that introduces naval warfare) would be an accurate reflection of post-release support.

What are you hoping to accomplish with that statement?

Elemental was finished on release. That's why we released it. We thought it was finished. We weren't "rushed' to get it out there.

You get annoyed by posts saying how great post release support is? I get a little bit annoyed by users who say Elemental wasn't "finished" on release. How do you even define that? You can make the argument that the game was horrendously buggy due to poor QA or that you didn't like the game when it first came out but to say it wasn't finished is not an accurate statement. We felt the game was finished and released it. Most non-developers have no concept of how insane it's gotten in 2010 to make a stable game with a new game engine. Heh, I've been making software for twenty years and I had no clue on how ridiculous it had gotten until Elemental got out there into the wild.

Our customers weren't happy with the game and so we endeavored to do what was required to make sure our customers were happy. That's what we do.

I like you Marlowwe, you're a good guy and all. I just want people to make the distinction between releasing something badly and releasing something unfinished. I think I've been very open about us messing up on the launch of the game. But it wasn't an issue of not having integrity.

I have always wanted to ask a dev the following questions when a game is released but does not play like it is finished which happens most of the time these days unfortunatly. Now please don't take offense at these questions. I love Gal Civ games. And I can see the potential in this game.

1. How did you come to the conclusion that the game was finished?

Keep in mind that a LOT of people had the same conclusion when playing the game at launch that it felt and looked unfinished/unpolished. Of coarse you not going to stop all bugs and such but this was not the main prolmes with the game. Many of the game mechanics were not finished or they did not play like they were finished. Basically the game did look and play like it was rushed out the door to a lot of people playing the game at launch.

2. In Beta you had said several times that you had some build that you weren't showing us that would take care of a lot of issues on Launch day. However this was not the case. What happened?

I'm one of those that support Stardock but think the game was not finished upon launch. To this day it still plays like a beta but it is 10 times better now than it was at launch. And by March it should be the game it was suppose to be at launch. this is not a FLAME so please don't be offended there is no need to take offense.

I have stayed away from this game for the most part, checking it from time to time to see if it has improved to the point where I will have fun playing it. Been playing CIV V and Age of Wonders: SM in the mean time.

Successfully updated karma reason!

Heavenfall

Join Date 07/2008

+441

I just played a great game where I limited myself in the following ways:

1) I didn't build any unit with more than one soldier in it

and

2) I didn't train any unit with a higher quality level than the default.

And suddenly the AI was extremely hard to fight against (in fact I lost). I would consider this "proof" that the AI is capable of so much more, but is simply limited in how it is built.

Successfully updated karma reason!

I've been having fun. Quite anxious for extra time with it in the holidays!

Successfully updated karma reason!

Quoting Frogboy,

reply 7

Elemental was finished on release. That's why we released it. We thought it was finished. We weren't "rushed' to get it out there.

You get annoyed by posts saying how great post release support is? I get a little bit annoyed by users who say Elemental wasn't "finished" on release. How do you even define that? You can make the argument that the game was horrendously buggy due to poor QA or that you didn't like the game when it first came out but to say it wasn't finished is not an accurate statement. We felt the game was finished and released it. Most non-developers have no concept of how insane it's gotten in 2010 to make a stable game with a new game engine. Heh, I've been making software for twenty years and I had no clue on how ridiculous it had gotten until Elemental got out there into the wild.

Our customers weren't happy with the game and so we endeavored to do what was required to make sure our customers were happy. That's what we do.

Its great that you support Elemental after the release and your customer support is great, too, but if you look at the low reviews and at other games in the genre you should have seen what Elemental is missing:

- Many different spells which are useful and balanced (Age of Wonders: Shadow Magic)

- Tactical battles with cover and city walls (Age of Wonders: Shadow Magic)

- Many different special abilities (Age of Wonders: Shadow Magic)

- Ability to upgrade units (Civilization V)

- Experience for units with different abilities (Civilization V)

Successfully updated karma reason!

Asks nicely for all the above before we get into expansion territory, pleeeaaaase

Successfully updated karma reason!

Quoting Heavenfall,

reply 60

I just played a great game where I limited myself in the following ways:

1) I didn't build any unit with more than one soldier in it

and

2) I didn't train any unit with a higher quality level than the default.

And suddenly the AI was extremely hard to fight against (in fact I lost). I would consider this "proof" that the AI is capable of so much more, but is simply limited in how it is built.

Perhaps it would be a solution to remove groups as the 6th stat and instead of increasing the stats units could get abilities as in Civilization V.

Successfully updated karma reason!

Wintersong

Join Date 07/2006

+235

Quoting Frogboy,

reply 51

No.

You have to find a beach. Then you have to build a city. Then you have to build a harbor. Then you build a ship. Then you have to load units onto the ship. Then you have to find another beach to unload the units.

That's a lot of work.

In Civ V, units just walk across the water (they turn into boats at the water's edge).

I thought you were talking just about the boats.

Successfully updated karma reason!

Quoting Frogboy,

reply 41

Quoting AlLanMandragoran,
reply 28

Quoting Frogboy, reply 25Depends who the mod specialist is. Most of you have probably heard of him.

Ooh, who is this person? When will we know? I wonder if I've seen his posts here...

I've gotta hand it to you, Brad. I'm starting to see where this thing is goin'. Building a solid infrastructure to allow the mod community to content the crap out of this will be a real future money maker for you guys. If I had some extra gildar laying around I'd invest beyond my $50. If Derek and crew can patch my real life income daily like they do the beta builds I'd be a happy camper and could throw some Tarthian gildar your way. Derek did tell Toby to do something about my gas in another post so I'm optimistic this will happen.

Despite the initial bumpy road you guys are walking the talk in my humble opinion. Kudos!

Indeed.

I said this back in September but it's important even after the success we've been seeing in the past few weeks:

EVEN if Elemental had been delayed to February it still would have been a disaster.

This is important: The problem with Elemental's launch was not an issue of not having enough time. It was a problem of Stardock still operating like a tiny studio while trying to make a AAA title. Our "logistics" failed. If it had been delayed until February, it would have still been a disaster.

As sucky as it is, Stardock needed this shake up. If you're going to make AAA titles, you can't make them like a hobby shop. Most long time people know that Stardock's main business is enterprise software. That's what pays the bills. Our software. But with Elemental, we decided to make a "real game".

The very idea that Elemental was seen and designed to content with games like Civilization (which had over 10X the budget of Elemental) should have put the company on a different track. It should have been making its games the same way it makes its software. Full time project managers, organized QA, and lots of other basic processes that just weren't done because our games were usually just a few people working on them.

So after Elemental launched, the best way for Stardock to build the studio to being where it needs to be is to go out and begin attracting the world's best talent. Kael was the first person I turned to (not just because we're friends but also of his qualifications). Kael's day job was as an ENTERPRISE project manager working with Fortune 500 companies.

The next key position is to bring someone on board who can design games and knows what modding is inside and out.

Do you guys remember me talking last year about ME learning Python so I could begin putting that into Elemental? Think about how misguided that is. The CEO of Stardock picking up python and trying to learn "on the job" on how to integrate deep modding like that? No. Wrong. The correct path is to find people who are already gods at this.

Not to mention, if we're going to move back to having a second studio team, we will need to bring on an experienced designer/PM from the game industry who is familiar with our genre.

I get what your saying and agree but you have already put out an outstanding game that even rivals and surpasses the best game in it's genra Masters of Orion 2 and that is Galciv2.

Successfully updated karma reason!

I got a little bored waiting for 1.1 last night so I decided to sneak into Stardock HQ and I found the following pictures of Stardock's new employees. You may recognize them due to their effect on the industry.

Project Manager

Python/mod God

Successfully updated karma reason!

"I'm so glad Stardock has such great post-release support."

Huh!? Stardock has released Elemental in early Beta form, in case this has eluded everyone. They've been using sales of this unfinished product to complete their game.

The state of graphics alone looks like concept drawings rather than a finished concept.

Successfully updated karma reason!

Quoting Mithelemir,

reply 68

"I'm so glad Stardock has such great post-release support."

Huh!? Stardock has released Elemental in early Beta form, in case this has eluded everyone. They've been using sales of this unfinished product to complete their game.

The state of graphics alone looks like concept drawings rather than a finished concept.

...in your opinion.

I do see it differently... , in fact, I can't think of another game where the developers, and CEO for that matter, join-in on the forum discussions, taking in comments/suggestions/feedback, giving us UPDATES on their progress, informing US of the PROCESS, etc., -- and not to mention the occasional left-left-right-upper-cut combo that, frequently, get sent towards them.

One other Stardock game I have is GalCiv I, so I visit the GalCiv forum regularly. And even now, in fact a couple days ago, I saw in the forums a developer giving live troubleshooting advice to someone who had just recently purchased GalCiv II -- which has been out for years. That is a perfect and real example of post-release support. Further, Stardock has announced a visit to GalCiv II for an update.

So I guess before I went into all that, I should have just asked: "Which planet are you from?"

Successfully updated karma reason!