Twilight Forums

Idle chat for wannabe heroes
It is currently Thu Jul 31, 2014 6:31 am

All times are UTC - 7 hours




Post new topic Reply to topic  [ 32 posts ] 
Author Message
PostPosted: Sat Jul 26, 2008 2:28 pm 
Offline
Site Admin
User avatar

Joined: Thu Apr 05, 2007 2:04 pm
Posts: 4262
Sometime soon I'm going to release a VERY early PvP scenario. I've got a lot of ideas and think it might work, but I don't consider myself much of a PvP expert to begin with, so the plan is to start small, test some basic principles, and then add to it slowly but steadily until we've got a good balance of strategy, complexity, and fun.

This earliest version is going to be not much more advanced than a simulation of PvP at its most basic and random (almost entirely chance, few options, little strategy). Frankly, I expect it will be boring, and not particularly rewarding. But if a few brave souls are willing to try it out for a bit, I can use that data to start adding improvements.

I'll post more when it's ready, but keep in mind I'm looking for bug testing and constructive suggestions.


Top
 Profile  
 
 Post subject:
PostPosted: Sat Jul 26, 2008 10:17 pm 
Offline
User avatar

Joined: Thu Sep 06, 2007 10:56 pm
Posts: 1466
Location: Berkeley when at school, San Diego when at home
Meh. You peoples and your newb gloves. :P

_________________
Image
Vote for TH on global rankings!


Top
 Profile  
 
 Post subject:
PostPosted: Sun Jul 27, 2008 2:57 pm 
Offline
Site Admin
User avatar

Joined: Thu Apr 05, 2007 2:04 pm
Posts: 4262
It's now public, and in the casinos. Adding more details here momentarily.


Top
 Profile  
 
 Post subject:
PostPosted: Sun Jul 27, 2008 3:04 pm 
Offline
Site Admin
User avatar

Joined: Thu Apr 05, 2007 2:04 pm
Posts: 4262
Right now it's you against a set of predefined characters, in a limited combat setting. In this first incarnation, I just want to test how the relative powers affect the likelihood of winning or losing, and just kind of make sure things are working at this level.

Combat assumes you always hit, and the opponent always hits. The amount of damage done hinges on your choice of the type of attack, versus their chosen defense. It's a lot like rock-paper-scissors in that certain attacks do better or worse against certain defenses, in a circular sort of fashion. Right now the defensive choice is 100% random, so it's all luck. Later, when real people are behind the characters, they'll have the ability to predefine choices for offense and defense.

Also right now the game assumes your choice for defense is random, as is the opponent's choice for offense. Both of these will be able to be adjusted later, too.

Right now there isn't any risk to playing, other than your time; so it's not a complete waste, I do give out some minor amount of XP, but mostly I'm assuming at least a few folks will want to mess around with it out of curiosity. There will be actual rewards for winning later.

Hit me with questions, bug reports, and the like.


Top
 Profile  
 
 Post subject:
PostPosted: Sun Jul 27, 2008 5:55 pm 
Offline
User avatar

Joined: Thu Sep 06, 2007 10:56 pm
Posts: 1466
Location: Berkeley when at school, San Diego when at home
Wait, what if I want to change my defenses mid combat with another player? How would that work?

_________________
Image
Vote for TH on global rankings!


Top
 Profile  
 
 Post subject:
PostPosted: Sun Jul 27, 2008 7:03 pm 
Offline
Site Admin
User avatar

Joined: Thu Apr 05, 2007 2:04 pm
Posts: 4262
Dunno. I'd imagined that as a strategy you'd have to pick before hand. Mostly to cut down on tons of clicking. Presumably it's a strategic element: you'd pick based on past encounters with the person (fight them once, see what they do, adjust). Or based on other benefits (equipment, say) that makes one defensive choice better than the others.


Top
 Profile  
 
 Post subject:
PostPosted: Sun Jul 27, 2008 8:13 pm 
Offline
User avatar

Joined: Thu Sep 06, 2007 10:56 pm
Posts: 1466
Location: Berkeley when at school, San Diego when at home
I get the feeling that that'd be way too predictable as well as, well, boring. Especially since stats play into this too.

_________________
Image
Vote for TH on global rankings!


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jul 30, 2008 12:58 am 
Offline
User avatar

Joined: Thu Sep 06, 2007 10:56 pm
Posts: 1466
Location: Berkeley when at school, San Diego when at home
I still feel shady about this kind of defense. If I want to have confidence in something like this, I want a system that I can morph very fluidly. This is a point Marshall brought up in some random debate that I thought was awesome. In KoL context, speed ascensions (or similarly diamondy activities) require rigid optimizations against a foe that really doesn't change, while in PvP, everything and anything can change, especially since you have a live opponent who can change his/her defenses literally in a snap of a second. There are many more elements to manipulate in PvP as well, and there is always a compromise in PvP that you must go with (consumption mini vs. white canadians). While I'd say that something that involves collections like the flower mini or the WC mini is inherently flawed, some of the minis are well balanced against each other.

[nonsense]And that's the story of how I beat up Dr. Horrible.[/nonsense]

_________________
Image
Vote for TH on global rankings!


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jul 30, 2008 6:34 am 
Offline
Site Admin
User avatar

Joined: Thu Apr 05, 2007 2:04 pm
Posts: 4262
Socket wrote:


1) Instead of your attack/defenses being random, allow the user to set % chances. I.E., you might set to 50% agressive, 25% standard, 25% cautious. Sure, an opponent might spade it out but you could always change it around tomorrow. Not to mention that


This is 100% my intention. Think I may have even mentioned it earlier in the thread.

#2 is also very likely, in some fashion or another.


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jul 30, 2008 5:09 pm 
Offline
Site Admin
User avatar

Joined: Thu Apr 05, 2007 2:04 pm
Posts: 4262
Yeah, I can see that quick and simple can be convenient, but I've never felt a win/lose button was especially satisfying on its own. Since fighting takes some game time, I want to make sure players are getting some entertainment for their money (which is time, as we all know). I don't want things to drag out that much -- ideally a tough fight will take maybe 8-10 turns once things are working properly, so there's room for some strategy and some excitement, without it taking too much clicking or time.

Ultimately, though, I know I can't make everyone happy, so I'm picking what I think I'd most enjoy to start, and then I'll work with it as I get feedback.


Top
 Profile  
 
 Post subject:
PostPosted: Thu Jul 31, 2008 4:48 pm 
Offline
Site Admin
User avatar

Joined: Thu Apr 05, 2007 2:04 pm
Posts: 4262
That's part of what I'm testing. Thus far:

Ryme won: 45/565 = 8%
Ivan won: 15/191 = 8%
#3 won: 55/149 = 37%
#4 won: 81/112 = 72%
The Mick won: 236/358 = 66%

I wanted to see how drastic the range in power would affect the range in odds. What I've concluded thus far is either I don't have nearly enough samples, or I've got bugs in my code. I would have expected a fairly linear difference between the three, and I definitely would have expected the middle slot to win very close to 50% of the time, and the #4 slot to win less than the Mick.

I do think I want to adjust the calculations so that this spread can be a bit wider, but manipulating your PvP power at the cost of other benefits is something I intend to be part of the game. If 6 points was inconsequential, my plan would have serious problems, so I'm glad that's not true.

If anyone feels like spending some time fighting the middle three, it would be much appreciated.

The three attacks are just rock/paper/scissors right now. Later there will be things that change them some. For instance, gear that boosts one of the attack types.


Top
 Profile  
 
 Post subject:
PostPosted: Sat Aug 02, 2008 10:23 am 
Offline
Site Admin
User avatar

Joined: Thu Apr 05, 2007 2:04 pm
Posts: 4262
Wow. Just realized I'd forgotten all about ties, which really skews the numbers. Here's the newest results.

Against Ryme (640 fights):
ryme won 49 times = 8%
tie 140 times = 22%
players won = 70%

Against #2 (218 fights)
#2 won 20 times = 9%
tie 51 = 23%
players won = 68%

Against #3 (174 fights):
#3 won 63 = 36%
tie 56 = 32 %
players won 32%

Against #4 (177 fights):
#4 won 126 = 71%
tie 37 = 21%
players won 8%

Against #5 (405 fights):
#5 won 273 = 67%
tie 29 = 7%
players won 26%

Still kind of skewed, but not quite as weird as they seemed before. Certainly the middle slot makes more sense, as it's split about evenly. I'd like to lower the chances of tieieng there, though. I'm gonna redo some of the calculations, and try a new set shortly.


Top
 Profile  
 
 Post subject:
PostPosted: Sat Aug 02, 2008 10:28 am 
Offline
Site Admin
User avatar

Joined: Thu Apr 05, 2007 2:04 pm
Posts: 4262
Also, a special thanks to the top 10 testers of the initial PvP test:

Code:
Name               rounds fought
xXRevaNXx              49
Matt the vigilante     32
Armarant              28
Socket                    24
Hariwald               22
Arthur Dent            19
Xarth                    18
chuwenhsuan            18
Brew                    17
lazyfire                17


Top
 Profile  
 
 Post subject:
PostPosted: Sat Aug 02, 2008 11:40 am 
Offline
Site Admin
User avatar

Joined: Thu Apr 05, 2007 2:04 pm
Posts: 4262
Revised the PvP calculations a bit. Also put in tiebreakers, so that if you tie but your margin of victory is still pretty decent, the judges might select you as a winner. (Or loser, if the margin goes the other way.) Ties are boring, right?

Also bumped up the XP gain slightly, to encourage further testing, because I realize this is still probably pretty unexciting, since it's mostly chance. But I really do want to know how this stuff will pan out before making wild and crazy assumptions.


Top
 Profile  
 
 Post subject:
PostPosted: Sat Aug 02, 2008 2:03 pm 
Offline
User avatar

Joined: Thu Sep 06, 2007 10:56 pm
Posts: 1466
Location: Berkeley when at school, San Diego when at home
Quote:
Your moderate attack far overmatches your opponent's attempt to dodge. Ryme loses 27 points.

Ryme proceeds with a moderate attack, which far overmatches your attempt to dodge. You lose 22 points.

You won the duel, so you'd probably get something here. An item or reward or something. But for now .... nothing.

Normally you wouldn't get this here, but as thanks for testing the system, and to prevent it from being a complete waste of your time, here's 175 XP. Thanks!


Wouldn't the match stop if I won? You still get a crafty shot at me. :P

_________________
Image
Vote for TH on global rankings!


Top
 Profile  
 
 Post subject:
PostPosted: Sat Aug 02, 2008 2:33 pm 
Offline
User avatar

Joined: Sat Apr 07, 2007 6:48 pm
Posts: 477
Location: Also Through The Dimensional Gate
Well, otherwise the attacker gets an extra advantage. On another note, Ryme, I've been running a bunch of pvp's on a few multis, for now using only using fierce.

_________________
Image


Top
 Profile  
 
 Post subject:
PostPosted: Sat Aug 02, 2008 2:54 pm 
Offline
User avatar

Joined: Thu Sep 06, 2007 10:56 pm
Posts: 1466
Location: Berkeley when at school, San Diego when at home
Ah, yeah, I just realized that.

Any more specifics on how the formulas will be calculated? (HP, power, etc.)

_________________
Image
Vote for TH on global rankings!


Top
 Profile  
 
 Post subject:
PostPosted: Sat Aug 02, 2008 9:29 pm 
Offline
User avatar

Joined: Sun Apr 08, 2007 1:01 am
Posts: 4530
Location: the Conservatory with the lead pipe
Quote:
Normally you wouldn't get this here, but as thanks for testing the system, and to prevent it from being a complete waste of your time, here's 14 XP. Thanks!
Your watch beeps with a suggestion from the Penny Pincher software. The reminder gets you a nice discount earning you 30 chips.
Is software supposed to trigger while using the arena?

_________________
Image
The churches are empty / The priest has gone home / And we are left standing / Together alone
--October Project: "Dark Time"


Top
 Profile  
 
 Post subject:
PostPosted: Sat Aug 02, 2008 10:04 pm 
Offline
Site Admin
User avatar

Joined: Thu Apr 05, 2007 2:04 pm
Posts: 4262
Yeah, software can trigger. I think I need to adjust it for what could be shortened fight time, but that's not applying yet.

Equipment isn't going to work right now (VR helm, for instance), because arena equipment is going to have some restrictions.


Top
 Profile  
 
 Post subject:
PostPosted: Mon Aug 04, 2008 5:18 pm 
Offline
Site Admin
User avatar

Joined: Thu Apr 05, 2007 2:04 pm
Posts: 4262
Socket wrote:
too small a sample to be meaningful by itself.


Yep ...

509 duels so far since the changes
147 wins for schmoe
221 ties
141 losses for schmoe

That's too many ties. By quite a bit. I'll have to adjust that some.


Top
 Profile  
 
 Post subject:
PostPosted: Wed Aug 06, 2008 9:35 am 
Offline
Site Admin
User avatar

Joined: Thu Apr 05, 2007 2:04 pm
Posts: 4262
New results with 0.2

Against Ryme (323 fights):
ryme won 3 times = 1%
tie 40 times = 12%
players won = 87%

Against #2 (783 fights)
#2 won 46 times = 6%
tie 232 = 30%
players won = 64%

Against #3 (847 fights):
#3 won 243 = 29%
tie 370 = 44%
players won 27%

Against #4 (724 fights):
#4 won 428 = 59%
tie 239 = 33%
players won 8%

Against #5 (217 fights):
#5 won 185 = 85%
tie 31 = 14%
players won 1%


Top
 Profile  
 
 Post subject:
PostPosted: Wed Aug 06, 2008 9:41 am 
Offline

Joined: Mon Nov 19, 2007 6:29 pm
Posts: 1811
Location: Florida
I'd be curious to know the stats of that other one :p


Top
 Profile  
 
 Post subject:
PostPosted: Wed Aug 06, 2008 9:59 am 
Offline
Site Admin
User avatar

Joined: Thu Apr 05, 2007 2:04 pm
Posts: 4262
Oh. 105 offense, 95 defense. Results were about the same as #3-- pretty evenly matched.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 32 posts ] 

All times are UTC - 7 hours


Who is online

Users browsing this forum: No registered users and 3 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group