.com.unity Forums

.com.unity Forums (http://forum.shrapnelgames.com/index.php)
-   Dominions 3: The Awakening (http://forum.shrapnelgames.com/forumdisplay.php?f=138)
-   -   Vengeance of the dead, what the hell (http://forum.shrapnelgames.com/showthread.php?t=35670)

llamabeast August 13th, 2007 06:18 AM

Re: Vengeance of the dead, what the hell
 
Oh I see where he's coming from now - I hadn't understood he was summing them up like that. Got it.

Wyatt Hebert August 13th, 2007 08:04 AM

Re: Vengeance of the dead, what the hell
 
thejeff: By holistically, I mean the problem of MA Ermor being one of only a few nations that have recruitable mages easily able to cast VotD combined with the double-bless Shadow Vestal rush. The problem seems to be that he couldn't move around due to the rush, and he got lots of kills (and didn't have time for an efficient anti-Undead capability) due to the rush, then got dropped by VotD. Without either one, VotD wouldn't have been as big of a deal.

Honestly, without the capability to spam it, there are much better answers, even Overland answers, to a SC (from what I read).


llamabeast: http://forum.shrapnelgames.com/images/smilies/laugh.gif I do hope I didn't make a mistake. It would have been bad if I had, but I did check it the inverse way, too. The probability of NOT getting a success in the first 3 is 0.96^3 (starting with three failures)*, ergo 1-0.96^3 is the probability of getting a success in the first 3, which is still 11.52%.

Let's see if I understand something about the rules. 1) All units are set to autoroute after 50 turns. 2) Mindless units without a commander or with a routing commander dissolve instead of routing. 3) You cannot retreat in Assassination attempts. 4) VotD is an assassination.

Are any of these incorrect?

This looks like a problem induced by the bug-fix to complete shut off retreat in an assassination attempt. Without substantial changes to the code or changes to the gameplay (allow retreats from Assassinations, which will kill you), or both (allow retreats, but you don't die), I fail to see how this problem can be resolved. The simplest solution might be to allow mindless units to be tagged to retreat in Assassination attempts; this will, of course, have its own share of consequences.

Thoughts?

Oh, and VotD is hardly needed against most combat casters. Earth Attack is probably superior for that purpose (I might be wrong, but that's one tough Elemental, and it tramples quite well). It takes a single path (typically easier to get), costs a bit more, but has no MR check. The only other issue is that it's easier to 'hide' from, but enough races can cast and spam it that it's not a huge problem?

Wyatt Hebert

* I realize that the actual probability will be the summation of the probabilities higher than 3 successes (using that formula), so that yields Sum{x=4->Inf}(0.96^(x-1)*0.04). However, without going through it, it comes close, if not exactly, to the 11.5% chance, so I just left it at that. http://forum.shrapnelgames.com/images/smilies/laugh.gif I might be wrong; my statistics class was quite a while ago, but I think I'm remembering it clearly.

thejeff August 13th, 2007 08:26 AM

Re: Vengeance of the dead, what the hell
 
About the rules:
1) yes
2) yes, normally
3) I believe you can retreat in assassination attempts, you just don't survive?
4) yes.

Point 2 is actually where the problem comes in, I think. There is no commander with VotD, so some special rules are in play. Obviously, despite the lack of a commander, the dead don't dissolve immediately during a VotD. It may be as simple as the check to start dissolving/routing happens when a commander routs or dies. Since there aren't any the check is never made. Adding a check to the 50 turn auto-rout mechanism to start mindless dissolving would be a better fix.

Wyatt Hebert August 13th, 2007 08:34 AM

Re: Vengeance of the dead, what the hell
 
Problem being is that due to my understanding of the way the code is implemented, NO units rout during an assassination. I think that's just a rule implementation. The simplest (meaning, most likely) way to fix it is to remove that, but then you have the problem of any unit routing in an assassination, including assassins themselves.

I guess my point is that #3 implies #2 is never checked, yielding the situation given. Is the issue severe enough that we get a codefix for it (which may be quite complex)? Is it severe enough if the only way we can fix it (or that it will be fixed) is if all units can rout during an assassination?

Resource limitations begin to dominate the thinking at this point. I agree that having them start dissolving would be nice, but the non-rout override in Assassination seems to prevent it. otoh, they have no commander in the first place, so I'm not sure if it applies (or they all may be a commander chassis). I'm assuming they have no commander, otherwise, a few particular kills would start them all dissolving.

Wyatt Hebert

Jazzepi August 13th, 2007 08:37 AM

Re: Vengeance of the dead, what the hell
 
>Problem being is that due to my understanding of the way the >code is implemented, NO units rout during an assassination.

I'm fairly certain this is untrue. Retreating from an assassination attempt just means you die.

Jazzepi

llamabeast August 13th, 2007 08:38 AM

Re: Vengeance of the dead, what the hell
 
Units definitely do rout during assassinations, including the assassin - I've seen it quite often.

And yep, I think your maths is right Wyatt - sorry, I just misunderstood what you'd done the first time.

Wyatt Hebert August 13th, 2007 10:01 AM

Re: Vengeance of the dead, what the hell
 
Hmm, I thought they had fixed that issue completely; hence, I mentioned the removal of the capability. I'll look into verifying that, too, if and when I have spare time.

It's fine. You're right, in general, on the binomial expansion, but afair, the first term would still be 1x^3y^0, or x^3. http://forum.shrapnelgames.com/images/smilies/laugh.gif I had to do the addition anyways just to be sure I was doing it right. It was the long way around. http://forum.shrapnelgames.com/images/smilies/laugh.gif

Wyatt Hebert

Fate August 17th, 2007 11:08 PM

Re: Vengeance of the dead, what the hell
 
This thread is descending into a question of VotD, which doesn't seem like the point.

The Turn Limit kills the defender of an assassination. I consider this obviously bugged, and it should be fixed. Either the defender lives, or the attacker gets to face the entire army (I mean, how long can you fight a guy in the middle of an army before someone notices?).

If this makes VotD unbalanced, balance it without a bug.

HoneyBadger August 18th, 2007 09:54 AM

Re: Vengeance of the dead, what the hell
 
Make it so that if the undead army fails to kill their target in 75 rounds, then they all dogpile on the one who cast the spell in the first place http://forum.shrapnelgames.com/images/smilies/happy.gif That would make quite a few people a little more wary of casting it on an SC god, I'd say...But, that seems overly complicated, so I just vote 3. I hate to see a unit, even a very tough unit, fight the good fight, only to be betrayed by the game engine, of all things.

Another possible solution, though, would be to strengthen the *quality* of the undead, based on the unit they're attacking's kills, instead of the numbers. So that at 100 kills or so, you get 100 longdead, 250 you get 100 shadows, 500 you get 100 ghosts, 1000 you get 100 banes, all the way up to Tartarians or whatever-keeping in mind that I really didn't research the exact quality of one undead over another, except in the most general of ways, but I think the point's clear enough for comfort.

Aethyr August 18th, 2007 10:30 AM

Re: Vengeance of the dead, what the hell
 
I would be in favor of changing it from a "nightmare" to a "real time" event (SP's bodyguards would then be able to help out).

Then, and I agree with Fate and with a previous post from GP, if the furn limit is reached the conflict is thrown into a "traditional" battle with the rest of the troops (including, but not limited to PD) coming to help out the assaulted party. Not a panacea, but I think this solves a lot of concerns wihtout over correcting.


All times are GMT -4. The time now is 05:05 AM.

Powered by vBulletin® Version 3.8.1
Copyright ©2000 - 2025, Jelsoft Enterprises Ltd.
Copyright ©1999 - 2025, Shrapnel Games, Inc. - All Rights Reserved.