A Test of Death Immunity

[cabbagemeister]

In my recent Sheaim game, I hooked up ~6 death nodes and started steamrolling everyone with spectres and wraiths. I was worried when I only had Hyborem left to kill, because I know that the Demon promotions gives immunity to death, and suddenly my absurd Str 15 wraiths would be Str 6. But to my surprise, when I invaded my wraiths started winning again and again, even against fortified Infernal longbows in cities, with <.1% odds (according to the odds calculator). This made me wonder: while the odds calculator seems to understand Death immunity okay, does the combat engine understand it?

To test this, I worldbuildered the following save with patch m:

The Player (Sheaim) has 20 sources of Death mana.
The Player has 25 green spectres

All on flatland, the Barbarians have:
25 Hyborems (Demon promotion, should be immune to death)
25 Acherons (Dragon promotion, should be immune to magic)
25 Clockwork Golems (Golem promotion, should be immune to death)
25 Treants (Elemental promotion, should be immune to death)
and 25 Liches (Undead promotion, should be immune to death)

I attacked each barbarian group in turn with the spectres, counted the number of wins and losses, and reloaded to attack the next group. With a base strength of 3 before death affinity, the spectres should have very low odds against all of these groups--unless death immunity is not working, in which case they would have 23 Str and very good odds.


The results:

25 Spectre vs 25 Hyborem
Combat odds list 3.00 vs 9.00 (Why 9.00? This is very confusing as Hyborem is listed as 7+4 unholy), <.1% chance of victory
Results: 11 Spectres win.

25 Spectre vs 25 Acheron
Combat odds list 3.00 vs 21.00, <.1%
Results: 0 Spectres win, only 5 Acherons are even hurt.

25 Spectre vs 25 Clockwork Golem
Combat odds list 3.00 vs 9.00, <.1%
Results: 10 Spectres win.

25 Spectre vs 25 Treant
Combat odds list 3.00 vs 10.00, <.1%
Results: 8 Spectres win

25 Spectre vs 25 Lich
Combat odds list 3.00 vs 5.00, 3.8%
Results: 24 Spectres win


Conclusions:
The Dragon promotion&#8217;s across-the-board magic immunity seems to be working, but all of the other promotions obviously do not grant death immunity like they should (and like the combat odds calculator thinks they do). Importantly, the results don&#8217;t suggest that death immunity is completely FAILING either, as specters with a true Str of 23 would mop the floor with unpromoted Hyborem, Treants, and Clockwork Golems. Instead, they win less than half the time. Perhaps these promotions&#8217; death &#8220;immunity&#8221; is actually just a 75% or so death resistance? If my spectres&#8217; true strength was around 8 (3 + [20 Death]*.25), that would come much closer to explaining the Hyborem / Treant / Golem combat results.


View attachment Test_DeathImmune.CivBeyondSwordSave

 

[loki1232]

Impressive testing. You're clearly on to something. *Whistles for Kael*

 

[kenken244]

Archeron has stoneskin, which is not displayed in the odds. Hyborem is displayed as a 9 due to the 50% resistance specters have to unholy.

Anyway, This might explain why fireballs still work against magic immune foes.


Edit- Played around with the save- gave myself 107 death mana and pit 40 wraiths against 40 meshabbers of dis. 14 wraiths survived, and quite a few meshabbers were damaged to exactly 4.9 strength. Then testing it with other damage types, gave myself 111 air mana and pit 40 air elementals against 40 air elementals. All 40 survived with combat odds of 50%. it seemed like the damage from affinity just wasn't counting, but the air elemental test seems to disprove this.

 

[zup]

Interesting.

 

[cabbagemeister]

Archeron has stoneskin, which is not displayed in the odds. Hyborem is displayed as a 9 due to the 50% resistance specters have to unholy.

Anyway, This might explain why fireballs still work against magic immune foes.


Edit- Played around with the save- gave myself 107 death mana and pit 40 wraiths against 40 meshabbers of dis. 14 wraiths survived, and quite a few meshabbers were damaged to exactly 4.9 strength. Then testing it with other damage types, gave myself 111 air mana and pit 40 air elementals against 40 air elementals. All 40 survived with combat odds of 50%. it seemed like the damage from affinity just wasn't counting, but the air elemental test seems to disprove this.

Stoneskin was displayed is the odds, which was why Acheron's Str was 21.00 and not 19.00.

The air elemental thing is tricky, because I believe that the elemental promotion does not give you immunity to magic, just resistance to magic (and immunity to death). Therefore the barb air elementals would be resisting the lightning damage from your +111 Str elementals, but not immune to it.

 

[kenken244]

Stoneskin has the effect of blocking the first three hits in combat which is not displayed in the odds. the +2 defense and 3 first strikes are in addition to this. Air elementals are immune due to the immune to lightning promotion. the odds were correctly displayed as 5 to 5 50% should not have resulted in a 40-0 combats.

 

[cabbagemeister]

Ah, you're right, I thought they had the Resist Lightning promotion (the icons look identical).

 

[Keeper_GFA]

Stoneskin has the effect of blocking the first three hits in combat which is not displayed in the odds. the +2 defense and 3 first strikes are in addition to this. Air elementals are immune due to the immune to lightning promotion. the odds were correctly displayed as 5 to 5 50% should not have resulted in a 40-0 combats.

Are you sure? It seems to me that the three first strikes are the method used to block damage in combat.

 

[kenken244]

Yes, I am sure, it even says so on the tooltip, just not in the odds. It has worked this way since it was introduced way back in light. the other effects were added after that when it no longer became a stack effecting spell.

 

[civ_king]

good work, bravo

 

[MaxAstro]

Yes, I am sure, it even says so on the tooltip, just not in the odds. It has worked this way since it was introduced way back in light. the other effects were added after that when it no longer became a stack effecting spell.

I have to question if Keeper isn't right on this one; blocking all damage for the first three hits and giving three first strikes are effectively the same thing. Are you sure the tooltip isn't just clarifying the first strikes?

 

[Thunder_Gr]

I have to question if Keeper isn't right on this one; blocking all damage for the first three hits and giving three first strikes are effectively the same thing. Are you sure the tooltip isn't just clarifying the first strikes?

I do not see how blocking all damage is the same as giving first strikes. Any first strike may miss, and when hit, applies damage to the opponent. Blocking all damage applies only when you get hit and it does not cause any damage to the opponent.

 

[zup]

Stoneskin is kind of bomb. Good thing some heavy hitter (dragons, brigit, yvain) can't have it.

 

[MaxAstro]

I do not see how blocking all damage is the same as giving first strikes. Any first strike may miss, and when hit, applies damage to the opponent. Blocking all damage applies only when you get hit and it does not cause any damage to the opponent.

I guess I am not entirely as familiar with how combat is calculated as I thought, then; I was not aware strikes could miss during combat.

 

[zup]

I thought first strikes always hit and first strike chances may or may not hit.

 

[cabbagemeister]

I think first strikes have a chance to hit (like a strike in normal combat) and first strike chances have a chance to be first strikes.

 

[kenken244]

Each round in combat a unit has a chance of dealing damage to the opponent. Stoneskin blocks the first three hits. First strikes are different because they come before combat starts and don't affect normal combat rounds.

 

[Guale]

I was also under the impression that first strikes always hit and first strike chances follow the normal combat odds for landing and that stoneskin uses those to simulate being immune to three hits. Saying it is in the tooltip isn't appropriate evidence, they can make the tooltip says anything they want.

 

[Zurai]

First strikes don't always hit, but that doesn't matter. Unless you're fighting something immune to first strikes, three first strikes are mechanically identical to zero first strikes and ignoring the damage of the first three attacks an opponent hits with. Think about it.

First Strike round:
You attack and either hit for X damage or miss and deal 0 damage to opponent.
Opponent does not attack and deals 0 damage to you.

Stoneskin round:
You attack and either hit for X damage or miss and deal 0 damage to opponent.
Opponent attacks you and hits and deals 0 damage to you.

 

[kenken244]

...After looking in the xml where there used to be a tag that did that was, I see that it does not do that anymore. But all that is off topic anyway. Back on topic, what could have been the cause of the strange behavior for Archeron is that stoneskin grants +50% death resistance. It seems like it takes more than 100% resistance to grant full immunity.