Evolution RTS, Forb vs Shard - Game 1

[Forboding Angel]

I'm trying to re-familiarize myself with shard behavior in light of semi-recent events. Anyway, I figured it would be a good idea for me to re-learn it's behaviors. I thought I would record the games and show them on youtube.

I would normally play better, but I had capped fraps fps to 15, I didn't realize that capping frps fps also caps my real fps. Loloops. So playing at 15 fps was kinda hard at times ;p

Here it is in all it's 1080p glory

http://youtu.be/MCeJeCuM4N8

[AF]

ps: pull from the master repo if you're modifying any behaviours, there are stability/performance changes in the lua code, though no major new features ( that I remember )

[Forboding Angel]

Well I'm using not-so-ballsey. Should I be using newer than that?

I forgot how entertaining playing vs shard can be. It does help that I have some gameside lua trickery that helps shard out a bit. In fact, that is one of the things I have been adjusting. I think I have it just about perfect atm.

That reminds me though, I think I need to impose a limit upon how many buildings shard is allowed to build. I have recently capped (by default, of course it can be raised by a host) unit limit to 250 units per player, and in a long game, shard could conceivably hit that with just buildings.

[BaNa]

I'm sorry for going offtopic here guize, but I'm happy to see you two back together.

[Das Bruce]

That reminds me though, I think I need to impose a limit upon how many buildings shard is allowed to build. I have recently capped (by default, of course it can be raised by a host) unit limit to 250 units per player, and in a long game, shard could conceivably hit that with just buildings.

Isn't that more a problem of the AI not prioritising resources appropriately?

[AF]

Its more that no measures for treating the unit limit as something to be wary of have been put in place, especially when in the past, rapid expansion of base construction was better than nothing.

The lua code in the git repository is newer than not so ballsey, though the vast majority of changes to it are sanity checks, validation, and other stability and memory changes. Also some refactors to rely more on hooks and other architectural improvements and streamline things.

[knorke]

video did not tell me anything about how the AI is now different.
Why do you not just write what you tweaked and then show it changed the AI's behaviour?
I think for AI videos a "dotwar" is more suitable to see its tactics etc.

[Forboding Angel]

It's not a video of the AI, it's of me just playing the game. The fact that it happens to be me vs an ai is coincidental for the purposes of this.

[AF]

video did not tell me anything about how the AI is now different.
Why do you not just write what you tweaked and then show it changed the AI's behaviour?
I think for AI videos a "dotwar" is more suitable to see its tactics etc.

This may not be a good way of doing it as a one off, since NTai experimented with multiple attack systems to throw people off, and Shard is more than capable of that, not to mention them all running at the same time if at all. It's just that few have done it other than me for some reason.

Besides I like this video, it shows pretty much what anybody who wants to buy a game wants to know, what is it actually like to play, complete with UI/chrome etc..

[1v0ry_k1ng]

NTAIs attack system was lamentable from what I can see from this video Shard seems to be continuing this legacy of the 'attack the player at his strongest point'

[smoth]

depends on the point of the AI, some ai are meant to give a fun match others are made to win.

[AF]

NTAIs attack system was lamentable from what I can see from this video Shard seems to be continuing this legacy of the 'attack the player at his strongest point'

Shard attacks the player at the weakest point, or at least the point on the grid that is occupied but has the fewest units present. I cannot speak for any modifications Forb may have made, and the attacking system was built to be extensible, not undefeatable.

There're thousands of ways of improving Shards attacking behaviour, many from NTai, many that haven't been done yet.

NTai also suffered quite badly from being a mishmash evolutionary design, and certain experimental attack systems that were somewhat disastrous, such as the unit herding behaviours.

[AF]

As a sidenote, change line 74 of attackhandler.lua from:

Code: Select all

cell.count = cell.count + 1

to:

Code: Select all

cell.count = cell.count + 1 + e:WeaponCount()

This may bring the behaviour even closer to what IK originally suggested ( or not, depends how the 0 damage weapons on the extractors are handled ).

[Forboding Angel]

NTAIs attack system was lamentable from what I can see from this video Shard seems to be continuing this legacy of the 'attack the player at his strongest point'

You must be joking... It attacked me at my weakest point, very early on. If it was anyone else playing (besides a few people like google waldo, or cucumber), they would have lost. It was pure luck on my part that me EMPs worked as well as they did and that I was able to get that turret up in time. In fact, at that point I was sure I had lost. When it comes right down to it, that attack did cost me the game, even tho it didn't kill me. Shard expands very aggressively, and if you get bogged down like I did you can pretty well say goodnight.

@af, I didn't modify shard code at all, just added little helpers modside (I actually did this a very long time ago, but I've been tweaking it lately).

Hmm, I guess it's time for game 2 eh? I'll do it tonight if I have time.

[Vadi]

Like the vid, but the lack of links to evorts/spring doesn't make it that nice for sharing - since people (obviously) won't google to look it up themselves.

[Forboding Angel]

Maek clicky for game 2

[knorke]

ok, so it is not a this-is-new-in-evo-ai video but a gameplay video.
Though now you have 3 of those videos and I do not see the differences between them.
To me they are not engaging/interessting/funny enough/informative to watch complety: 30 minutes of unedited fraps footage is too long and boring for my taste.

So maybe you aimed at potential new players to show "what anybody who wants to buy a game wants to know" but these people are the most likely to close the video after 3 minutes.