Vengeance of the dead, what the hell

[LoloMo]

How about in a draw, you survive but is feebleminded.

[Wyatt Hebert]

Having read all of this thread, I would tend to agree with Edi. However, a few points I have thought of:

1) VotD, Dreams of Rl'yeh. Do these work like Arena fights? (If so, this would explain why Immortals still die.) If not, then auto-destroy at turn limit killing Immortals seems like a bug.

2) Assuming a 4% chance to work (and this may be high), odds are it will take 16 castings for VotD to affect the target (0.96^x=0.5; x*log 0.96=log 0.5; x=(log 0.96)/(log 0.5)). This, therefore, requires 16 mage-turns of, what, a SSSD or DDDS or whichever it is, which is, in itself, 96 Research, as well as the upkeep of those mages. Assuming it's not even a threat to a SC with less than 100 kills, well, it seems rather fair. That's 16 times the cost of the spell (2 pearls, right?), 16 mage-turns of a decent mage type, and to have already killed 100+ troops. And, yes, it's a God in this case, so he can be called back. This by Caelum, in this case, which doesn't even have to worry about losing Mage-turns to Call God.

These are the numbers to look at. Is _this_ considered fair? (And, of course, understand that _this_ particular case involves someone without a weapon that can hurt Undead. It is conceivable that with a better Weapon (which I understand the OP didn't have access to) he could have lived through this attack, and caused the number of castings to double.

Any error in the numerical analysis?

Wyatt Hebert

[thejeff]

Part of the problem with this discussion is that the two sides of the debate seem to be talking past each other.

One side, largely, is saying that dying because the attackers don't auto-rout when they should is wrong.

The other side is saying the overall chance of killing an SC is low, so the spell is balanced.

These are different arguments. They're not even opposed. The spell can be balanced, but achieve its affect in an unfair way. I'd say that's where I stand. The balance is likely fine, it may even be underpowered, but losing an SC due to time limits when the attacker should rout first is a lousy way to achieve that balance.

I'd suggest either strengthening or cheapening the spell, but allowing the dead to rout, by giving them a few non-mindless commanders if necessary.

Are there any other ways to get an attack force (either assassination or province attack type spell) without a commander? Phantasmal attack, maybe? (I think that's what it's called. The one that sends phantasmal warriors to attack a province?) Are they mindless?

[Kuritza]

>> These are the numbers to look at.

The numbers lie. I am not sure why, but astral spells with some penetration items work FAR better than beforementioned 4%. In my case it took exactly 3 castings to pass through 25 MR, supposedly without any penetration items on casting mages. Heck, I routinely have my 25-MR SCs paralyzed with the very first casting for 40+ turns, and I have given up hope of ever using a SC against R'lyeh.
Your 16 attempts number looks really, really overestimated.

And in case of Caelum you can count the turns you have to spend recruiting priests for calling back your God instead of recruiting mages and the losses your army will suffer without him now that your God doesnt stand between your mages and enemy sacreds. Probably you will simply lose the game.
Ah, and the mage-turns for equipping your god again.
Oh, and why bother calling him, he still has all these kills on him and will eat VotD again.

And one more thing to consider. This is not just a highly situational SC-removal spell. Most mages dont stand a chance against 40-50 zombies they will likely have to face alone, which is a very good result for 3 astral pearls.

[Reverend Zombie]

How about...removing the turn limit entirely for VotD attacks?

[thejeff]

Highly agreed with Kuritza's last point.
I'd always considered it much more useful against battle mages who've racked up a decent number of kills, than against SCs. I just assumed most SCs would survive it. Hadn't considered the turn limit.

[Kuritza]

Thats the ideal solution to me (there might not be a stalemate in VotD, but default), but all exceptions require the devs to change game code usually, which they hate.
I am not speaking about the Dominions devs now, of course, but I guess there are serious reasons for such trend.

[Wyatt Hebert]

You say the numbers lie. Very well, they may. However, do you even understand the information assumed by the numbers? For example, what is the probability that it will take as few as 3 castings (Given it has a 4% chance each time)? Exactly 3 castings is .96*.96*.04. Exactly 2 castings would be .96*.04, and Exactly 1 casting is .04, by definition. This leads to a probability of 11.52%.

Now, _assuming that 3 castings will penetrate 50% of the time_, the actual probability of failure is x^3=0.5, or ~79%. This gives the chance of success at 21%. If this is accurately the chance of success, then the game code does have issues.

I guess the other consideration is the path combination required to cast VotD. I _think_ that only MA Ermor has the requisite paths innately (i.e. giving the cost information is correct), though I could easily be mistaken on that score. The other check is that, possibly, it's overpowered in Ermor's case. Granted, from what is apparently a subtext of this message, Ermor has the capability to not only use double blessed Shadow Vestals AND VotD. Now, is THIS overpowered, and, if so, what could be done to fix the combo without changing VotD.

Oh, and please note that I _estimated_ nothing. I simply applied binomial statistics. _Given_ a 4% success rate, there is a probability slightly higher than 50% that the first 16 trials will fail. Stated slightly differently, what is the probability of rolling <49 on a d50 16 times in a row? ~50%. If the success rate is wrong, of course the result will be wrong. That's why I did the secondary calculation assuming your case was an average one. I doubt your case was average, however. I think the Ermorian player got lucky with a roll at some point. ~21% chance of success is to hit someone with, I believe, only 2-3 points higher than you. That would, I believe, imply something on the order of +12 Penetration, which is nigh impossible that early in a game.

Let's figure out how the numbers fall out, and then we'll start worrying. If astral spells have a bonus to penetration overland, that needs to be verified and fixed, unless that is working as intended. Honestly, VotD might require one higher in one of its two paths, and that probably would help. On the other hand, I think some of the stuff in the next patch is Shadow Vestal being reduced, so that might help this kind of situation, too.

Let's look at the issue holistically, rather than the specific cause of death. Can we agree on that?

Wyatt Hebert

P.S. I am planning, time-allowing, to run a test scenario with VotD. Ermor vs. Ulm, and I will be paying close attention to Penetration vs. MR score each time. Combat occurs before dominion change, and spells happen before Combat, so the numbers I read on the turn should be the numbers used in the calculation, correct?

[thejeff]

Let's look at the issue holistically, rather than the specific cause of death. Can we agree on that?

No, we can't. That's the point I was trying to make earlier.

By holistically, you seem to mean from a balance point of view, not a thematic one. But I'm not arguing a balance issue.

I'm not arguing that the spell is overpowered. I'm arguing that auto kills because the undead don't rout or die when the attacking side is supposed to are a problem.
If that gets fixed, then VotD may need a cost or path decrease , or a boost in power to compensate. Or not if the auto kill was unintended.
The spell would remain useful taking out evocation type battle mages, regardless.