.com.unity Forums - AI DEATH MATCH 2

.com.unity Forums (http://forum.shrapnelgames.com/index.php)

- Space Empires: IV & V (http://forum.shrapnelgames.com/forumdisplay.php?f=20)

- - AI DEATH MATCH 2 (http://forum.shrapnelgames.com/showthread.php?t=8617)

Re: AI DEATH MATCH 2

Quote:

Originally posted by Dralasite:
Just curious, how many AIs did we have that lost all matches? Unless you have the info handy, I can look through the old Posts and figure it out.

It was in the first round.
Here are the AIs:

Group A1 - Cylon
Group A2 - Romulan
Group A3 - Ekhazan
Group A4 - Sergetti
Group A5 - Corrian
Group A6 - Toltayan
Group A7 - Shadows
Group A8 - XiChung

Upon once time, the Sergetti was the most strong AI in the market!!
Really I think that the AIs scripts are a lot more strong now.

Re: AI DEATH MATCH 2

Two unexpected results! http://forum.shrapnelgames.com/image...s/rolleyes.gif

United Flora vs Space Vikings

The Space Vikings defeated the United Flora at the turn 160. They expanded fast and the UF had not time to develop their optimal level.

Here the link: UFlora_vs_Vikings_F2.zip

Tessellate vs Aquilaeian.

The Aquilaeian defeated the Tessellate at the turn 500 by points. They won 7.5M vs 7.4M! Think the Tessellate would have win if the game would be played for some more turns.

Here the link: Tessellate_vs_Aquilaeian_F1.zip

Re: AI DEATH MATCH 2

Wow both top contenders fall to underdogs! Just when I thought I had it all figured out http://forum.shrapnelgames.com/images/icons/icon7.gif

Re: AI DEATH MATCH 2

Quote:

Originally posted by Dralasite:
Wow both top contenders fall to underdogs! Just when I thought I had it all figured out http://forum.shrapnelgames.com/images/icons/icon7.gif

http://forum.shrapnelgames.com/images/icons/icon7.gif I thought like you...

Re: AI DEATH MATCH 2

Quote:

Originally posted by Unknown_Enemy:
Hi people,

Would you be so kind to help me ?
I am trying to update an AI. So I run several tests games, and my AI regularly reverse to a very early ship design.

For instance, in place of a nice and mean battle cruiser it reverses to a puny destroyer. Which is really a pain.
http://forum.shrapnelgames.com/images/icons/icon9.gif

I did read on this thread about this bug, so please could someone explain me why does it happens, and most of all tell me how to avoid that ?

Think this could help you:

Quote:

Originally posted by Rexxx:
Well, after checking the Gron savegames I have to say again how helpful for me this contest is.
From time to time (every second or third turn) the Gron think of a LC equipped with CSMs as their best short range attack ship (I guess due to a lousy oversight in the design-file). Not the smartest thing a talisman race can do in mid- and endgame when they can have a DN coming with TKPs and the Talisman.

Quote:

Originally posted by Master Belisarius:
I saw this problem with most of the modded AI races...

In other of my Posts on this topic, I wrote the work-around that in my view fix this problem (talking about the problem with the Sergetti's ships).
When does exist several ship designes with different sizes, at some point, the AI start to rebuild their ships using small designes (old designes!).
This start to occur (at least in the cases that I knew), after some time the AI reach the DN size. But think in some way is also related to the weapon that currently is using for the DN.
An example, the Tessellate attack ships.
As most the modders, I did different attack ships, depending of the ship size. Here the DN main design.

Name := Attack Ship
Design Type := Attack Ship
Vehicle Type := Ship
Default Strategy := Maximum Weapons Range
Size Minimum Tonnage := 801
Size Maximum Tonnage := 5000
Num Must Have At Least 1 Ability := 2
Must Have Ability 1 := Weapons Always Hit
Must Have Ability 2 := Quantum Reactor
Minimum Speed := 4
Desired Speed := 6
Majority Weapon Family Pick 1 := 22
Majority Weapon Family Pick 2 := 0
Majority Weapon Family Pick 3 := 0
Majority Weapon Family Pick 4 := 0
Majority Weapon Family Pick 5 := 0
Secondary Weapon Family Pick 1 := 30
Secondary Weapon Family Pick 2 := 0
Secondary Weapon Family Pick 3 := 0
Secondary Weapon Family Pick 4 := 0
Secondary Weapon Family Pick 5 := 0
Shields Spaces Per One := 250
Armor Spaces Per One := 10000
Majority Comp Spaces Per One := 200
Majority Comp Ability := Weapon
Secondary Comp Spaces Per One := 200
Secondary Comp Ability := Weapon
Num Misc Abilities := 7
Misc Ability 1 Name := Cloak Level
Misc Ability 1 Spaces Per One := 10000
Misc Ability 2 Name := Combat To Hit Defense Plus
Misc Ability 2 Spaces Per One := 10000
Misc Ability 3 Name := Point-Defense
Misc Ability 3 Spaces Per One := 250
Misc Ability 4 Name := Self-Destruct
Misc Ability 4 Spaces Per One := 10000
Misc Ability 5 Name := Multiplex Tracking
Misc Ability 5 Spaces Per One := 10000
Misc Ability 6 Name := Extra Movement Generation
Misc Ability 6 Spaces Per One := 10000
Misc Ability 7 Name := Point-Defense
Misc Ability 7 Spaces Per One := 10000

The weapon "30" is the Shard Cannon.
Well, time in time, after start to build these designes, the AI decided to re-create their attack ships... and had not better idea than build an old CR design, that used their Crystalline Armors and Napal Bombs as weapon... Here is:

Name := Attack Ship
Design Type := Attack Ship
Vehicle Type := Ship
Default Strategy := Ram
Size Minimum Tonnage := 401
Size Maximum Tonnage := 500
Num Must Have At Least 1 Ability := 1
Must Have Ability 1 := Shield Generation From Damage
Minimum Speed := 4
Desired Speed := 6
Majority Weapon Family Pick 1 := 0
Majority Weapon Family Pick 2 := 0
Majority Weapon Family Pick 3 := 0
Majority Weapon Family Pick 4 := 0
Majority Weapon Family Pick 5 := 0
Secondary Weapon Family Pick 1 := 13
Secondary Weapon Family Pick 2 := 0
Secondary Weapon Family Pick 3 := 0
Secondary Weapon Family Pick 4 := 0
Secondary Weapon Family Pick 5 := 0
Shields Spaces Per One := 250
Armor Spaces Per One := 0
Majority Comp Spaces Per One := 80
Majority Comp Ability := Shield Generation From Damage
Secondary Comp Spaces Per One := 10000
Secondary Comp Ability := Weapon
Num Misc Abilities := 5
Misc Ability 1 Name := Quantum Reactor
Misc Ability 1 Spaces Per One := 10000
Misc Ability 2 Name := Combat To Hit Defense Plus
Misc Ability 2 Spaces Per One := 10000
Misc Ability 3 Name := Supply Storage
Misc Ability 3 Spaces Per One := 10000
Misc Ability 4 Name := Point-Defense
Misc Ability 4 Spaces Per One := 250
Misc Ability 5 Name := Self-Destruct
Misc Ability 5 Spaces Per One := 10000

Why use this old CR design and at least, don't use the other BC or BB availabe ship designes????

I had an idea: did a paste of the original DN two times, with the idea that maybe the AI could use the second design (that was the same than the first of course!), instead the CR. I was wrong...
Then, remembered that the Pyrochette never had this problem... and remembered that them used 2 kind of DN ships with different weapons.
With this idea, I did a new try: designed a new DN that use the other Crystalline Weapon (can't remember the name... I don't like so much this weapon!!), their weapon family is 31.
This thing fixed the problem!!!!!! The Tessellate started to rotate their ship designes, using the original DN with shard cannon during some time, and time in time, designed the DN with the other weapon as main attack ship.

I had the same problem with the auxiliary ships: Support, Offensive sweepers, troop transport, and even with units. Using the same idea (designes with different weapons for the DN ships), never had again this problem.
IT WORKED TO ME... and think should work to everybody!!!
Hope it will help to create more strong AIs!

Think this post should help you.

Re: AI DEATH MATCH 2

United Flora vs Narn Regime.

The United Flora needed some time, but finally defeated the Narn at the turn 410.

Here the link: UFlora_vs_Narn_F2.zip

[ May 08, 2003, 00:48: Message edited by: Master Belisarius ]

Re: AI DEATH MATCH 2

Hi people,

Would you be so kind to help me ?
I am trying to update an AI. So I run several tests games, and my AI regularly reverse to a very early ship design.

For instance, in place of a nice and mean battle cruiser it reverses to a puny destroyer. Which is really a pain.
http://forum.shrapnelgames.com/images/icons/icon9.gif

I did read on this thread about this bug, so please could someone explain me why does it happens, and most of all tell me how to avoid that ?

What is this mess about "calling names" ?

Re: AI DEATH MATCH 2

Quote:

Originally posted by Dralasite:
Wow both top contenders fall to underdogs! Just when I thought I had it all figured out http://forum.shrapnelgames.com/images/icons/icon7.gif

Hey, who's calling the Vikings underdogs?
*Rollo sharpens his axe http://forum.shrapnelgames.com/images/icons/icon12.gif

Yay!, for the Aquilaeian and Vikings.
Two venerable AIs that gave those young upstarts a good and deserved kicking http://forum.shrapnelgames.com/images/icons/tongue.gif .
(funny that these duels were both between races of the same modders)

Rollo

Re: AI DEATH MATCH 2

Quote:

Two venerable AIs that gave those young upstarts a good and deserved kicking

I agree! UF was undefeated until then. It was a nice suprise. Now I'm back on the edge of my seat wondering if UF will be #1 or not. This thread is better than spectator sports http://forum.shrapnelgames.com/images/icons/icon12.gif

Re: AI DEATH MATCH 2

Quote:

Originally posted by Rollo:
(funny that these duels were both between races of the same modders)
Rollo

http://forum.shrapnelgames.com/images/icons/icon7.gif And more funny will be the next round, when 3 AIs of the same modder will be there! http://forum.shrapnelgames.com/images/icons/shock.gif

Re: AI DEATH MATCH 2

CueCappa vs Aquilaeian.

The CueCappa won at the turn 360 (as I expected), and then, after the first round of games here are the standings in the group F1.

GROUP F1 Points
Tessellate 1
CueCappa 1
Aquilaeian 1

Link to the game: CueCappa_vs_Aquilaeian_F1.zip

Re: AI DEATH MATCH 2

Vikings vs Narn

Another unexpected result... at least to me. The Narn won by points at the turn 500.

Here the link:Vikings_Vs_Narn_F2.zip

Then, the standings after the first round of games are:

GROUP F2 Points
United Flora 1
Space Vikings 1
Narn Regime 1

Re: AI DEATH MATCH 2

De-lurking...

MB - I just want to say that I have read all of this and the Last AI Death Match and they are excellent. Great undertaking for the AI Designers and the players. Thanks.

Rexxx - I noticed the Khrel made it to the top 12 AI's. They and the Tessellate are the only two non-TDM AI's there. So when (or where if already done) are you going to release the Khrel AI/shipset for download inside the forum?

Re: AI DEATH MATCH 2

Indeed, your solution is working.

Thanks for it.

Re: AI DEATH MATCH 2

Quote:

Originally posted by cybersol:
De-lurking...

MB - I just want to say that I have read all of this and the Last AI Death Match and they are excellent. Great undertaking for the AI Designers and the players. Thanks.

Rexxx - I noticed the Khrel made it to the top 12 AI's. They and the Tessellate are the only two non-TDM AI's there. So when (or where if already done) are you going to release the Khrel AI/shipset for download inside the forum?

I'm always glad to know a Lurker! http://forum.shrapnelgames.com/images/icons/icon7.gif
Welcome on board!
About the Khrel files, already I have posted all the scripts involved in the contest, on this forum (but without the images).
Try this: http://forum.shrapnelgames.com/newup...047347357.ZIP.

If you want the Khrel images, drop me a mail to jlperyr@adinet.com.uy, and I'll send it to you (don't think Rexxx or Richard Meic would have problems with this).

Re: AI DEATH MATCH 2

Quote:

Originally posted by Unknown_Enemy:
Indeed, your solution is working.
Thanks for it.

http://forum.shrapnelgames.com/images/icons/icon7.gif Cool!

Re: AI DEATH MATCH 2

CueCappa vs Tessellate.

The Tessellate defeated the CueCappa, at the turn 500 by points (again).

Here the link: CueCappa_vs_Tessellate_F1.zip

Re: AI DEATH MATCH 2

MB - It is funny that you suggest that because Last night I was looking through some of the old favorite shipsets threads and ran across the Khrel. So I had just finished doing exactly what you suggested. My only question is whether the original empire files (2k, 3k, 5) are equivalent to new one (2k)?

Just so it is easier for others to get the shipset for the fine AI, here is the link.

Khrel Shipset Link

Re: AI DEATH MATCH 2

Quote:

Originally posted by cybersol:
MB - It is funny that you suggest that because Last night I was looking through some of the old favorite shipsets threads and ran across the Khrel. So I had just finished doing exactly what you suggested. My only question is whether the original empire files (2k, 3k, 5) are equivalent to new one (2k)?

Just so it is easier for others to get the shipset for the fine AI, here is the link.

Khrel Shipset Link

Here all the Emp. files that Rexxx sent me, were done at the 21/02/03.
Khrel_Emp_Files.zip

Strongly suggest to use the Khrel scripts that I have posted with the AI other scripts, because Rexxx sent me several updates for the Khrel.

Re: AI DEATH MATCH 2

Space Vikings vs United Flora.

The UF took revenge, and defeated the Vikings fast, at the turn 170.

Here the link: Vikings_vs_UFlora_F2.zip

Re: AI DEATH MATCH 2

Yes, I assembled the shipset from the link I gave with the AI from the death match download and added the empire files that you recently uploaded. So now I have an new AI opponent that was in the top twelve. Thanks.

EDIT: As long as you are providing .emp files, the Tessellate that Gyphin has in his signature does not include these files. Nevermind, I just found these files inside the death match scripts download that you already provided.

The Deathmatch looks exciting six AI's tied after 1 round.

[ May 10, 2003, 21:09: Message edited by: cybersol ]

Re: AI DEATH MATCH 2

Quote:

Originally posted by cybersol:
Yes, I assembled the shipset from the link I gave with the AI from the death match download and added the empire files that you recently uploaded. So now I have an new AI opponent that was in the top twelve. Thanks.

EDIT: As long as you are providing .emp files, the Tessellate that Gyphin has in his signature does not include these files. Nevermind, I just found these files inside the death match scripts download that you already provided.

The Deathmatch looks exciting six AI's tied after 1 round.

http://forum.shrapnelgames.com/images/icons/icon7.gif You was more fast than me!
Anyway, here are the *.emp files Tessellate_Emp_Files.zip

Re: AI DEATH MATCH 2

Aquilaeian vs Tessellate.

The Tessellate defeated the Aquilaeian without big problems, at the turn 360.

Here the link: Aquilaeian_vs_Tessellate_F1.zip

Re: AI DEATH MATCH 2

Narn Regime vs United Flora.

The UF had not problems to defeat the Narn very fast. They won at the turn 160.

Here the link: Narn_vs_UFlora_F2.zip

Re: AI DEATH MATCH 2

Quote:

The UF took revenge, and defeated the Vikings fast, at the turn 170.

Was this different from when the Vikings won because of starting position?

[ May 12, 2003, 18:47: Message edited by: Dralasite ]

Re: AI DEATH MATCH 2

Quote:

Originally posted by Dralasite:
<blockquote>quote:<hr /> The UF took revenge, and defeated the Vikings fast, at the turn 170.

Was this different from when the Vikings won because of starting position?<hr /></blockquote>Was very similiar. The Vikings won at the turn 160.
The Rollo's races expand very fast, and more when they start in the down side of the map. The UF crushed the Vikings without problems, this time.
But also, remember that the "luck" is very important too, specially when 2 strong AIs are fighting.

Re: AI DEATH MATCH 2

Aquilaeian vs CueCappa.

Another unexpected result... the Aquilaeian won at the turn 500 by points (the CueCappa only kept one system).
During most part of the game the CueCappa was ahead in the numbers and had the inititative, but arround the turns 380-430 the things changed to the Aquilaeian side.

I'm somewhat embarrassed... The Aquilaeian had a lot of luck playing against the Tessellate and winning 7.5M vs 7.4M. And now, they won this game against the CueCappa when I'm sure that the CueCappa is clearly more strong than the Aquilaeian (specially with this map/settings).
I know, I know, the luck is always a factor in any game, and more when 2 AIs are playing... but I'm not sure to be happy that my AI defeated the CueCappa, because at the end I want to see the most strong AIs in the highest places of this contest.

Here the link: Aquilaeian_vs_CueCappa_F1.zip

This was the Last game for the group F1. Here the standings:

Tessellate 3 Points
Aquilaeian 2 Points
CueCappa 1 Point

Then, the Tessellate and the Aquilaeian got places to the Semifinal.

Re: AI DEATH MATCH 2

Quote:

Originally posted by Rollo:
My poor Cappa, such bad luck http://forum.shrapnelgames.com/images/icons/icon9.gif . But I suppose they have proven their worth http://forum.shrapnelgames.com/images/icons/icon7.gif .

Agree 100% with you.

Re: AI DEATH MATCH 2

Narn Regime vs Space Vikings.

Staring in the down side, the Vikings had not problems to crush the Narn, at the turn 150. Then, they got the second place in the group.

Here the link: Narn_vs_Vikings_G2.zip

The standings in the group G2 are:

United Flora 3 Points
Space Vikings 2 Points
Narn Regime 1 Points.

The two semifinals are:

G1
Tessellate
Space vikings

G2
United Flora
Aquilaeian

I will run a max of 6 games x every AI (3 games staring in the upper side, and 3 starting in the down side). But, if I got a result like 4-0, I will not run all the games.

Re: AI DEATH MATCH 2

My poor Cappa, such bad luck http://forum.shrapnelgames.com/images/icons/icon9.gif . But I suppose they have proven their worth http://forum.shrapnelgames.com/images/icons/icon7.gif .

Re: AI DEATH MATCH 2

Tessellate vs Vikings.

The Vikings won at the turn 230.
This time was not unexpected... because the Space Vikings defeated the Tessellate in the first round.

Here the link: Tessellate_vs_Vikings_G1.zip

Re: AI DEATH MATCH 2

could you post up the standings of all races that has participated in the match and sort them by highest point to the lowest?

Oh, and I'd like fries with that also http://forum.shrapnelgames.com/images/icons/icon10.gif

Re: AI DEATH MATCH 2

Quote:

Originally posted by TerranC:
could you post up the standings of all races that has participated in the match and sort them by highest point to the lowest?

Oh, and I'd like fries with that also http://forum.shrapnelgames.com/images/icons/icon10.gif

If you have excel, you would like this: AI_DeathMatch_2_Fixture.zip

Re: AI DEATH MATCH 2

MB,

could it be possible that you have filled out the score wrong in the excel for the Tessellate/Space Vikings?
The Tessellate have 1 point but they lost to the Space Vikings.

Anyway, interesting file http://forum.shrapnelgames.com/images/icons/icon7.gif

I mostly play TDM/FQM solo and for my next game i am going to add some of these races as they are not all in TDM. I suppose i just have to copy the appropriate files into the TDM/FQM mod directory and these races should then be available?

Re: AI DEATH MATCH 2

As long as we are finding errors in the fixture spreadsheet. I believe the Orks won third round C5 Orks vs Gron.

Looking over the results so far, it really strikes me how the AI field is broadly competitive. The top 12 AI's were all extremely good, but somehow the top six take it to an even higher level. I'm just amazed at the great efforts all the AI designers have put into their creations.

Re: AI DEATH MATCH 2

United Flora vs Aquilaeian.

The UF defeated the Aquilaeian without problems, at the turn 270

Here the link: UFlora_vs_Aquilaeian_G2.zip

Re: AI DEATH MATCH 2

Thanks for the fixes Cybersol/Minipol!
(In my discharge must say that the spreadsheet was supposed to be released at the end!)

Re: AI DEATH MATCH 2

Quote:

Originally posted by Master Belisarius:
Thanks for the fixes Cybersol/Minipol!
(In my discharge must say that the spreadsheet was supposed to be released at the end!)

whoops... http://forum.shrapnelgames.com/images/icons/icon7.gif

Re: AI DEATH MATCH 2

Space Vikings vs Tessellate.

This time, the Tessellate won a fast game at the turn 190.
In the early game, the Vikings expanded fastest, but the Tessellate were able to remove the Vikings colonies of their systems.

Here the link: Vikings_vs_Tessellate_G1.zip

Re: AI DEATH MATCH 2

So far, it looks like Vikings vs Tessellate are evenly matched, and United Flora vs Tessellate is an even match (I believe both Groups are 50%-50% so far for wins losses?)?

[ May 15, 2003, 16:25: Message edited by: Dralasite ]

Re: AI DEATH MATCH 2

Quote:

Originally posted by Dralasite:
So far, it looks like Vikings vs Tessellate are evenly matched, and United Flora vs Tessellate is an even match (I believe both Groups are 50%-50% so far for wins losses?)?

Yes Vikings/Tessellate is a very balanced match... If the Tessellate can reach the mid-game with a decent possition, they can win, but the Vikings can defeat them because they research and expand more fastest.

About UFlora vs Tessellate, don't know. Still they have not played during this contest... but think the UF have advantage.

Re: AI DEATH MATCH 2

Aquilaeian vs United Flora.

Starting at the upper side, the Aquilaeian have even lesser opportunities to win... The UF won at the turn 270.

Here the link: Aquilaeian_vs_UFlora_G2.zip

Re: AI DEATH MATCH 2

Quote:

About UFlora vs Tessellate, don't know

Oops, I meant to say United Flora vs Vikings, that is also even losses and wins.

Re: AI DEATH MATCH 2

Tessellate vs Vikings.

This time the Tessellate offered a decent battle, but the Vikings won at the turn 360.

Here the link: Tessellate_vs_Vikings_G1B.zip

Re: AI DEATH MATCH 2

Quote:

Originally posted by Dralasite:
<blockquote>quote:<hr /> About UFlora vs Tessellate, don't know

Oops, I meant to say United Flora vs Vikings, that is also even losses and wins.<hr /></blockquote>http://forum.shrapnelgames.com/images/icons/icon7.gif I can see now! http://forum.shrapnelgames.com/images/icons/icon7.gif And agree with you, of course.

Re: AI DEATH MATCH 2

Quote:

Originally posted by Master Belisarius:
<blockquote>quote:<hr />Originally posted by Unknown_Enemy:
Thanks to the devious mind of Nodachi http://forum.shrapnelgames.com/images/icons/icon12.gif , the SE4 batch works on windows XP. But I could not get the installer to work.

So I put all the files from setup1.cab in a directory, and reinstalled SE4 to its default loacation.

Then indeed if you don't try to update anything from the SE4 Batch configuration, it is working.

It is probably the same for windows 2000.

Great!
To change the SEBatch configuration, you only need to change the values into fields of the file "paramete.dbf", using something that could change these files.<hr /></blockquote>Hi Master Belisarius,
It took me a while to get SE4 Batch running, but finally it did.

When running your programming i saw in Task Manager that it automatically executes SE4.exe.
Does your program make a command line voor SE4 based on the variabels you provide in the mainscreen and in the configuration screen ??

Could you explain if it so what commandlines you are using ?

Furthermore i hope that you will take the time to make SE4 batch again and available for all Versions of Windows. It will most definitly encourage a lot of new AI programmers to join the club.

I for sure am very interested !

Thanks in advance.

gr. Tymy

Re: AI DEATH MATCH 2

What was se batch written it? Perhaps it can be decompiled? It may be easier to rewrite, I dunno.

Re: AI DEATH MATCH 2

United Flora vs Aquilaeian.

Again, the UF won... this time at the turn 210.
Then, if in the next game they win again (what I can bet 1000 to 1!), they will reach the final.

Here the link: UFlora_vs_Aquilaeian_G2B.zip

Re: AI DEATH MATCH 2

MB, what races are in the finals so far?

[ May 16, 2003, 23:20: Message edited by: JLS ]

Re: AI DEATH MATCH 2

Quote:

Originally posted by Dralasite:
What was se batch written it? Perhaps it can be decompiled? It may be easier to rewrite, I dunno.

Was done in Visual Fox Pro, and yes, could be decompiled, but don't think worth it...
The code is VERY SIMPLE, and could be re-done in other better language, without problems.