Jump to content


Photo

1.70 Messing with the Original Balance


  • This topic is locked This topic is locked
36 replies to this topic

#21 ThetaOrion

ThetaOrion

    title available

  • Members
  • 676 posts

Posted 11 February 2006 - 09:49 PM

I have a tip for you Theta. Condense your posts and provide a replay pointing out obvious and blatant flaws and I'm sure you will make more impact than writing lengthy posts with vague feelings.


Yes, I know that Thudo only reads 1/4 of what I write and ignores half of that.

But my words or feelings were all I had, because most of these replays are too large to upload or too long to watch. Thudo is not going to watch my replays. I know that too.

I also know that Excedrin's tests were meant to prove how stupid ThetaOrion's wordy claims really are.

But, I'm very happy that you guys took me seriously even if the original purpose was to debunk me.

LarkinVB: Just watched Ork vs Ork but had to quit at minute 7 as beta7 orks were so stupid that it hurts. Just watch through all of these replays and you will certainly find enough material for betas 9+.


Zenoth: Just watched the three first replays ... I must say there is definitely space for improvements before the release of 1.7 final.


You guys got Thudo's attention, whereas, I never would. Thudo wanted me to shut up and go away, which I can do of course. But, the real geniuses finally got his attention, I do believe.

--
||
--

Regarding this thread, I can buy into Excedrin's and your claim that the Space Marines shouldn't be artificially restricted or returned to 'normal' in order to give the Orks a chance.

I can buy into, "There should be only one agenda for the mod. Will a certain decision make it play better (=more successfull) ??? This decision should be based on a match versus a good human player."

I can live with that.

AS LONG AS,

Excedrin keeps making more of these Beta 8 and Beta 9 tests against 1.41 AI or against 1.51 AI, and you LarkinVB keep looking at the results and commenting.

If they keep tweaking the 1.70 Orks until they are winning 19/20 against the Plain Vanilla 1.41 and/or winning 9/10 against 1.51, then I won't even notice the inherent 1.41 imbalance creeping through into the Orks. And, then the 1.70 Orks won't be an embarrassment when they are released to the world.

--
||
--

Regarding this thread and the Cage Match direction it is taking. I would like to see 2 Beta 8 AI Orks face off against 2 Plain Vanilla WA 1.41 AI Orks on Thudo's Mountain Trails Map. I want to see if the Beta 8 faction in team play can stand up against the old 1.41 PV or the old 1.51 AI Mod in team play.

And, in the Final 1.70 Beta, I would like to see 3 Final Beta AI Orks square off against 3 PV 1.41 Orks (or 3 1.51 AI Mod Orks) in Arkhan's favorite Testing Grounds Map, to see which faction is getting the Squiggoth the fastest and which AI carries the day in team play.

But, I don't know whether these Cage Matches can be done 2vs2 or 3vs3. And, if such matches requires online time and two machines of comparable abilities, I'm not sure that I could figure out a way to run them myself. We are going to have to rely upon Excedrin to make the tests, and to occasionally make them creative by introducing team play into the mix.

#22 Klementh

Klementh
  • Members
  • 16 posts

Posted 12 February 2006 - 01:38 AM

LOL ThetaOrion...

Malkor, thanks a lot, but does it have a home page or forum thread where I can read more about it and download future versions?

#23 Malkor

Malkor

    Eternity

  • Members
  • 375 posts
  • Projects:Loladins of Legend.
  •  Inanely inane inanities!

Posted 12 February 2006 - 04:17 AM

I don't post anywhere about it often enough to warrant a site or thread dedicated to it.

#24 Excedrin

Excedrin
  • Project Team
  • 154 posts

Posted 12 February 2006 - 07:18 AM

I also know that Excedrin's tests were meant to prove how stupid ThetaOrion's wordy claims really are.

Woah dood, I was just showing that it can be done... If anything it was to disprove Finaldeath's comment that it would cause a problem.

I'm honestly surprised that people are reading this much into AI vs AI. It's not that useful in my opinion...

#25 ThetaOrion

ThetaOrion

    title available

  • Members
  • 676 posts

Posted 12 February 2006 - 08:02 AM

To me, what was significant about the Cage Fights was the insights that it triggered and brought forth from LarkinVB.

It finally gave the developers some specific things they could focus on, something this betatester wasn't able to do.

#26 Malkor

Malkor

    Eternity

  • Members
  • 375 posts
  • Projects:Loladins of Legend.
  •  Inanely inane inanities!

Posted 12 February 2006 - 08:55 PM

Most of my starcraft AI tests are ai vs ai, actually, as I have very few willing human testers and I can't test it logically as I know exactly how to abuse it and what to expect.

#27 Klementh

Klementh
  • Members
  • 16 posts

Posted 12 February 2006 - 11:48 PM

Malkor, do you know the Broodwar AI Project at: http://www.entropyze...BroodwarAI.html ?

Is it any good?

#28 ThetaOrion

ThetaOrion

    title available

  • Members
  • 676 posts

Posted 13 February 2006 - 01:43 AM

The team as a whole has made some major breakthroughs during the past couple of days, the -dev discovery and the Cage Matches.

But . . .

we might have lost Arkhan, which is not a good thing if that has happened.

Now that I have tried the -dev switch, I can now say that with -dev the game really is playing just like Arkhan and Thudo have been telling us all along that it should be playing.

It's just that I and others were playing without -dev, and thus we were playing with a completely different game. It wasn't even the same.

I hope we didn't lose Arkhan because of it -- because both sides of the aisle were completely right in their observations. From the developer side, it was right and ready, and from the point of view of some of us betatesters it needed a lot more work so it wouldn't be an embarrassment when it is released.

#29 Malkor

Malkor

    Eternity

  • Members
  • 375 posts
  • Projects:Loladins of Legend.
  •  Inanely inane inanities!

Posted 13 February 2006 - 02:44 AM

Malkor, do you know the Broodwar AI Project at: http://www.entropyze...BroodwarAI.html ?

Is it any good?


I've known of it's existence for a while. As for being good, I never took the time into looking at it, as no real evidence as to what it actually does is given (and the coders are russian...Can't really ask them about stuff(Also the fact I see tanks and vultures and ghosts leads me to believe it's probably easier to abuse than RD)). The fact they're also using some kind of launcher and not the flawless mpqdraft makes me question what exactly they're doing.

You can check it if you want to.

Why the hell are they testing in single player mode with cheats, too? That REALLY changes how the AI works...

Edited by Malkor, 13 February 2006 - 02:48 AM.


#30 ThetaOrion

ThetaOrion

    title available

  • Members
  • 676 posts

Posted 13 February 2006 - 03:17 AM

Why the hell are they testing in single player mode with cheats, too? That REALLY changes how the AI works...


Does -dev automatically activate the cheats?

I never played with -dev until today, because I like to play under the same circumstances that I know the end user will be playing under, aka no -dev and no WXP\DATA folder. If you can get your betatests as close to the same circumstances that the end users will be using, then your betatesting input is more valid and useful, with all other things being the same.

Edited by ThetaOrion, 13 February 2006 - 03:17 AM.


#31 Malkor

Malkor

    Eternity

  • Members
  • 375 posts
  • Projects:Loladins of Legend.
  •  Inanely inane inanities!

Posted 13 February 2006 - 03:18 AM

Referring to the BW AI project that Klementh linked to.

#32 ThetaOrion

ThetaOrion

    title available

  • Members
  • 676 posts

Posted 13 February 2006 - 04:51 AM

Oh, okay, then back to the original topic.

In both Beta 7 and Beta 8 without -dev, the Space Marines and the Orks were grossly imbalanced, and it happened repeatedly over and over again in the tests I made.

Beta 8 with -dev feels balanced -- I mean I was able to win and my Ork AI ally actually attacked the Space Marines. It was a completely different game with -dev, even though it was the exact same challenging scenario that I have been playing since around Beta 2.

I know it made Thudo mad that I fought so hard for that belief of mine, but that was what I was seeing.

ThudMeizer: And negativism is in Theta's reports but its largely emotion. We don't need that whatsoever. Its childish and must not continue.


Anyway, my childish emotion and my final determination to fight for what I was seeing and fight for what I believe payed off in the end with the Cage Matches and maybe even the -dev discovery. Thudo wouldn't have made the -dev discovery if some of us wouldn't have passionately held to our guns. It's important to honestly report what one is seeing, even if the developers don't like the reports.

Of course, it created frustration all across the board. We betatesters weren't being believed, or told to shut up, or being implicitly called liars, and the developers were taking a lot more heat than they deserved when we fought back, all because of the -dev thing.

Without -dev, there is a feeling of imbalance in the game, whether we are talking about Beta 7 or Beta 8. And, I believe it even more now that I know about the -dev thing.

#33 Klementh

Klementh
  • Members
  • 16 posts

Posted 14 February 2006 - 01:04 AM

And now back to the BroodWar topic. How do you test AI vs AI on it?

#34 LarkinVB

LarkinVB

    title available

  • Members
  • 1,488 posts

Posted 14 February 2006 - 08:13 AM

And now back to the BroodWar topic. How do you test AI vs AI on it?


Not ! Do me a favor and use some user forum for this.

#35 Finaldeath

Finaldeath
  • Project Team
  • 188 posts
  • Location:UK

Posted 14 February 2006 - 12:36 PM

Yeah, thud should just lock this, its gone completely off topic. This isn't the starcraft forum!

#36 Malkor

Malkor

    Eternity

  • Members
  • 375 posts
  • Projects:Loladins of Legend.
  •  Inanely inane inanities!

Posted 14 February 2006 - 09:26 PM

And now back to the BroodWar topic. How do you test AI vs AI on it?



I'll answer this for him before it gets locked.

Run a direct cable connection or a LAN game, and make it 4v4, 2v2, whatever TVB options. Set an ally on your team, and an enemy. Choose Terran. When the game starts, lift off and float to your ally's base, and kill all your workers. Bring any survivors to your ally's base. You have now effectively and easily created an observation situation. The second option is to make an UMS where both comps give you vision and you have no units.

And, yeah, if you got any more questions, please use AIM or MSN for it.

msn: icsw_lurker@hotmail.com
aim: Ascherzon
e-mail: Malkor@gmail.com

Edited by Malkor, 14 February 2006 - 09:26 PM.


#37 thudo

thudo

    Wacko AI Guy!

  • Division Leaders
  • 12,164 posts
  • Location:Lemonville North, Canada
  • Projects:DoW AI Scripting Project
  • Division:DoW
  • Job:Division Leader

Posted 14 February 2006 - 09:36 PM

heheh.. Yeaaa I will lock this Mr. FD! hehe.. Now I have a moderator moderating an external forum. However... I do respect your experience in this field.

Btw, nifty way to initiate AI.. Gawd.. Starcraft.. (circa April'98!). Them were the days the cybernetic dinosaurs roamed the landscape.. <sigh>
Advanced Skirmish AI Team Lead for the coolest Warhammer40k PC RTS out there:

Dawn of War Advanced AI Headquarters

Latest DoW Advanced AI Download!




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users