Why SoDs are a waste of time.

[Tyndmyr]

Aright, our example here will be finger of death, an iconic, easily available SoD. It also has the advantage of being less failure prone than say, phantasmal killer, and having a bit of damage on success. Our example mage will be reasonable, starting 18 int, all upgrades into int, a +6 int enchantment bonus, but no other bonuses to the DC. In short, a relatively average level 15 wizard selecting a level 7 spell to cast.

The DC of FoD will be 25. Failure means death, success means 3d6+15.

In the other corner, we have direct damage. We'll assume a generic d6/caster level spell, with no other effects. Say, an orb of whatever, with no special optimization. Force, lets say, to avoid having to calculate the value of the bonus effects the others have. A 15d6 touch attack. Average value, 52 damge

The CR 15 encounter possibilities are as follows:
Brass Dragon, Mature +18 fort, 253 hp
Bronze Dragon, Adult +17 fort, 241 hp
Inevitable, Marut -Immune to SoDs, orb clearly wins.
Mummy Lord -Immune to SoDs, orb clearly wins.
Red Dragon, Adult +18 fort, 253 hp
Silver Dragon, Adult +18 fort, 253 hp
White Dragon, Old +19 fort, 276

Thus, we can calculate on +18 fort, 253 hp as pretty clearly average for the non-immune. With a 70% pass rating, this gives the SoD 25 damage 70% of the time, and 253 damage 30% of the time, for an average damage per cast of 93.4.

Our generic terrible blaster hits on a 2+, for an average of 49.4 damage. Seems like an easy win, right?

Well, no...not really.

First off, it's easier to pump caster level than save DCs, resulting in far more optimization possible for the blaster.

Secondly, saves outstrip save DC increases. This means that the tougher the fight, the more useless SoDs are. By level 18, your DC will have likely risen by only 1, but your typical targets are up to a +23 fort save, dropping your odds of success dramatically.

Thirdly, damage works on nearly everything. Remember those two immune types? Thats a pretty normal proportion of immune opponents, and it only gets harder as you level. As you approach CR 20, anything describable as a significant challenge is likely to be immune.

Fourthly, you probably have a party. They probably do normal damage. If you do normal damage, your efforts stack with theirs. SoDs do not. If your teammates poured in 120 damage before an SoD works, well, using the SoD was a waste of time, as using a damage spell would acheive the same result, and be much, much more reliable. Thus, late battle SoDs are ineffective.

Fiftly, DMs cheat. As shown in another popular recent thread, a good chunk of them will fudge a roll to avoid bosses going down too early. This makes an early SoD entirely pointless, as you have no chance whatsoever to succeed. Coupled with point 4, these alone entirely invalidate the style of spell.

Sixth, SoDs are good at killing mooks. Lower CR mobs are much more likely to fail saves. However, at any level where you have SoDs, you have a variety of aoes. These are vastly, vastly better at killing mooks.

Therefore, you should never prepare SoDs.

[Kylarra]

SoDs are the epitome of rocket tag anyway, so I generally try to avoid them.

[liquid150]

A level 15 caster has spells that just say "I win, no save" so why should he bother with either of these routes? It's fairly common knowledge that fortitude saves are the worst to target, and that damage is weak.

[Cyclocone]

Polymorph any Object, Glass Strike, etc.

Marut Inevitable, Fort +7, auto win.
Mummy Lord, Fort +13, auto win.

[Ernir]

Well, SoDs work rather well if you acted like a good Schrödinger-Wizard. That is, you used your divinations to find out what you are facing, your maxed knowledge skills to figure out its weak spots, and your limitless library of spells to have the perfect SoD/SoL for just that opponent.

But yes, if you prefer just spamming your favourite spell at everything you encounter (A.K.A. you're most blast-in-the-door Sorcerers), making that spell a SoD might be a bad idea...

[Oslecamo]

A level 15 caster has spells that just say "I win, no save" so why should he bother with either of these routes?

Because most of them aren't so "I delay them 1 turn, no save" when you use them on monsters with SLAs of their own.

Forcecage, solid fog and the like can easily be bypassed by any monster with greater teleport/dimensional door SLA. And there's a lot of such monsters at higher level.

[Malakar]

Yeah... Fort saves are monsters strongest saves. Finger of Death comes in at level 13 anyway. And... You declared save DCs less optimizable than CL... but also didn't optimize DCs. Nevermind that Orb of Force has a CL cap, so Orb of Forcing does... Exactly the same damage at CL 15 as it does at CL 200.

I'll also note that you choose a Core save or die, and one of the widely hailed Power creepy damage spells.

People who are most likely to cast Finger of Death are Necromancers DC specialists.

So they do the following thing:

Int 20 at level 1 +3 from levels +6 item = +9 Int bonus. +7 spell, +2 from feats. So that's already a 28, and that's a boring character.

Now find someone who just casts save or lose spells, (and no save just lose spells) but targets different saves, and that makes things even easier, because you can just use their weakest save.

[Eldariel]

Why the heck are you attacking monsters with good Fort saves with save-or-dies? Dragons in general have good saves; you shouldn't really bother hitting those at all without huge investment. The whole point of SoDs is to hit the weak saves... Yeah, Fort-saves tend to be quite good in general, but what about the NPC caster? Y'know, the particularly dangerous opponent - those tend to have worse Fort-saves. CR 15 NPC Sorc from DMG (since we're using by-the-books foes right now) has Fort-save +5.

And DM fudging...well, if DM fudges rolls, he's gonna fudge HP too so the fight doesn't end a second before he wants it to so it doesn't matter what you do. I'm not gonna plan on my DM cheating since I trust he's put enough thought into the fight that he doesn't have to (especially with all the "Finger of Death dropped the Red Dragon in 3 secs!"-stories; we've had one and it's still the most legendary story in our playgroup).


Let's not forget how common energy resistances are (especially once magic comes into play) and the fact that averages are irrelevant 'cause you always target monsters' weaknesses.

[Tyndmyr]

Yeah... Fort saves are monsters strongest saves. Finger of Death comes in at level 13 anyway. And... You declared save DCs less optimizable than CL... but also didn't optimize DCs. Nevermind that Orb of Force has a CL cap, so Orb of Forcing does... Exactly the same damage at CL 15 as it does at CL 200.

Finding d6/level blasting spells is...ridiculously easy. You either swap out spells, or find a way to uncap the damage. Either way, problem solved. Damage caps on blasting spells are not a big setback for casters.

I didn't optimize DCs OR CLs. Fair is fair. I did give the caster pretty solid starting DCs, though.

I'll also note that you choose a Core save or die, and one of the widely hailed Power creepy damage spells.

Any d6/level touch attack would have worked out the same. The traditional power orb is fire, due to the daze and the SR:No. Those factors were not part of my analysis, so which d6/level touch attack you use is irrelevant to the equation.

Edit: Targetting weaknesses is a valid strategy, but the fact remains, SoDs are overwhelmingly fort saves, and fort saves are generally high. Those are ALL the monster manual CR15 encounters. Yes, your DM might build an NPC via class levels. Odds are pretty solid that he is a class with good fort saves too. Roughly even, at any rate. There is also a chance that he'll be flat out immune to death via death ward from casting, racial immunity, or equipment.

[liquid150]

Because most of them aren't so "I delay them 1 turn, no save" when you use them on monsters with SLAs of their own.

Forcecage, solid fog and the like can easily be bypassed by any monster with greater teleport/dimensional door SLA. And there's a lot of such monsters at higher level.

You're doing it wrong.

[bokodasu]

Yeah... you're not always attacking monsters. Or I guess at least I'm not, maybe you are, I don't know, but that doesn't mean it's useless for every situation.

Just this morning I was thinking how useful it will be when we're both at level 20 and I can hit the rogue (+6 Fort save, 120ish HP) with Finger of Death.

[Tyndmyr]

If you're level 20, and he's only got 120 hp, you have a lot of options to kill him. Why prepare something that's only good against him when you can prepare something good against other targets as well?

[Rad]

A main part of using SoDs well is choosing which save to target. Your example used fort against dragons... not exactly expected to work. Weak save DC progressions progress slower than DCs and are usually only helped by the fact that many monsters will ahve much higher HD than their CR.

That said, if you really optimize damage with a few splatbooks you can get a LOT of direct damage from spells.

[Draz74]

Sixth, SoDs are good at killing mooks. Lower CR mobs are much more likely to fail saves. However, at any level where you have SoDs, you have a variety of aoes. These are vastly, vastly better at killing mooks.

At higher levels, AOE SoDs are also an option. Weird might be just as good as Meteor Swarm at killing off large numbers of mooks. (Not a perfect example, since both of these are weak L9 spells, and Maw of Chaos pwns them both ... but you get the idea. Cloudkill is a decent spell ...)

[Kyeudo]

The OP is doing it wrong.

First, you don't just prepare one SoD over and over again. You prepare several types of SoD or SoL (Save or Lose) so you can target all three save types. All those creatures up there have low Reflex saves, so you hit them with a Reflex save debuff of your choice and cripple their ability to move and attack. I'd have to do a little digging to find which particular spell would be the most appropriate at that level, but its out there. There are that many splat books.

Second, who cares if your party half-killed the monster before you use your Save or Die? The monster is just as dead and probably sooner than trying to grind through its hit points.

Third, whether a GM cheats or not varies from table to table. I let my players one-shot bosses if they can, since I tend to optimize my boss creatures (Hexblade 3/Paladin of Tyranny 3/Rogue 2 can survive alot). The GM's leanings always need to be factored in in your own playstyle, but that doesn't apply to discussions of the rules system as a whole.

[Malakar]

Edit: Targetting weaknesses is a valid strategy, but the fact remains, SoDs are overwhelmingly fort saves, and fort saves are generally high. Those are ALL the monster manual CR15 encounters. Yes, your DM might build an NPC via class levels. Odds are pretty solid that he is a class with good fort saves too. Roughly even, at any rate. There is also a chance that he'll be flat out immune to death via death ward from casting, racial immunity, or equipment.

The fact that there are 200 Fort save or dies, only 100 Will save or dies, and only 10 Reflex save or dies has nothing to do with if you prepare the Reflex and will ones.

And why are you harping on Death immunity. Most of the best ones are not death anyway. Flesh to Stone, Baleful Polymorph core, Endless Slumber, Slipsand, Moonbolt, Call Avalanche, Magic Jar, Fleshshiver, Smoky Confinment, Illusory Pit, Final Rebuke.

Not a lot of death tags there.

EDIT: As regards mooks. You get AoE save or dies for mooks too. See Entomb.

[Thrice Dead Cat]

Sadly, Tyndmyr's info is old news. There's also the consideration that in a party of four, if the wizard pops a SoD and succeeds, then no one else does anything, decreasing the potential fun. If her fails, he wasted a round. This is part of the logic used in treantmonk's guide to being GOD in that you want to help your peons party out rather than just annihilating everything.

[bokodasu]

If you're level 20, and he's only got 120 hp, you have a lot of options to kill him. Why prepare something that's only good against him when you can prepare something good against other targets as well?

In this particular instance, because he's taken a bunch of feats that make him hard to damage in other ways. Anything with a reflex save is right out, for example, which seems to be most of my fun spells. He's also diminutive and invisible, so while anything based on landing an attack is still possible with some helper spells, it's going to be a lot more complicated than pointing at him and having him keel over.

Ok, so most people aren't trying to kill their party's rogue in the most entertaining way possible, but it works for a lot of other human-based targets for the same reasons.

[Saph]

Fourthly, you probably have a party. They probably do normal damage. If you do normal damage, your efforts stack with theirs. SoDs do not. If your teammates poured in 120 damage before an SoD works, well, using the SoD was a waste of time, as using a damage spell would acheive the same result, and be much, much more reliable. Thus, late battle SoDs are ineffective.

This is the biggest reason why I don't bother with SoDs on my wizard's spell list. Instakilling an enemy is great if you're on your own and you're not trying to do damage any other way, because it takes the enemy's HP from 100% to 0%. But the closer the enemy's HP drops to 0%, the more wasteful a SoD becomes.

It's generally a better idea to pick spells that complement what the rest of your party is doing, rather than trying to win the battle on your own. This is why I find buffs, and debuffs more effective than SoDs (and direct damage works pretty well too if you've got four other party members pounding on the enemy target at the same time).

[nightwyrm]

Sadly, Tyndmyr's info is old news. There's also the consideration that in a party of four, if the wizard pops a SoD and succeeds, then no one else does anything, decreasing the potential fun. If her fails, he wasted a round. This is part of the logic used in treantmonk's guide to being GOD in that you want to help your peons party out rather than just annihilating everything.

Unfortunately, people keep misunderstanding what god/batman wizard means. They keep thinking it stands for "wizard kills everything by himself".

[ericgrau]

A level 15 caster has spells that just say "I win, no save" so why should he bother with either of these routes? It's fairly common knowledge that fortitude saves are the worst to target, and that damage is weak.

Average monster will saves are the same as fortitude saves. Reflex is the common low save, and there are a low number of SoD's there.

The OP is doing it wrong.

First, you don't just prepare one SoD over and over again. You prepare several types of SoD or SoL (Save or Lose) so you can target all three save types. All those creatures up there have low Reflex saves, so you hit them with a Reflex save debuff of your choice and cripple their ability to move and attack. I'd have to do a little digging to find which particular spell would be the most appropriate at that level, but its out there. There are that many splat books.

The typical analyses assumes the low save is targetted, and SoD/SoL still suck as they are failure prone. In reality you don't always know the low save, nor the monster's immunities to half the SoDs so in practice SoDs are even worse than theory (which is already bad). And if you didn't prepare the same SoD repeatedly, or you did but you prepared the wrong one, it sucks more b/c the first casting probably failed and now you have a bunch of spells you can't use.

Second, who cares if your party half-killed the monster before you use your Save or Die? The monster is just as dead and probably sooner than trying to grind through its hit points.

B/c now the sucky SoD sucks even more when it fails to kill that monster when you could have killed it with damage more reliably. It's not that the (already bad) SoD got worse, it didn't, it's that in a party your other option just got better.

A lot of SoL's are decent because they are aoe and against enough targets might succeed on a couple mooks. But I'd never bother with a single target SoD. Or single target SoL, but there are less of those.

[liquid150]

Average monster will saves are the same as fortitude saves. Reflex is the common low save, and there are a low number of SoD's there.

It's been a while since I looked at the table, but at least in the MMI I thought that will saves were at least a few points lower than fortitude saves on average. Yes, reflex was the lowest, but will is generally lower than fortitude.

[Kyeudo]

Unfortunately, people keep misunderstanding what god/batman wizard means. They keep thinking it stands for "wizard kills everything by himself".

It's "Wizard cripples encounter to the point where all that remains is mop-up", which some people consider the same thing.

Still, God Wizards are more party friendly than Batman Wizards, since God style play relies on buffs and no-save debuffs along with battlefield control to make the rest of the party steamroll the encounter. It looks like the Wizard isn't doing as much as the Fighter, until you remember that the Fighter is only Hasted, Stoneskined, and Magic Weaponed because the Wizard said so.

[Eldariel]

This image may be helpful:

Spoiler

Show

However, you should remember that ultimately averages are again irrelevant since basically every CR has high- and low-save creatures for each save, with some particularly high saves in some type which inflates the average. That does not make the poor-save creatures any less suspectible to said spells though. That and this only applies to straight-out-of-MM monsters.

EDIT: Here's an SRD-only excel sheet - the above information consists of...basically everything that exists for 3.5. http://rapidshare.com/files/416296833/SRD_Averages.xls

[ericgrau]

EDIT: ninja'd.

[Milskidasith]

Tynd's posts ignores some key points: namely, optimizing saves at all, that you get higher saves due to higher level spells at higher levels, and that, while targeting the weak save may not be an autowin, it is certainly a massive increase.

Let's take a theoretical monster with 200 HP that has a chance of failing a will save 15% more often than a fort save, and it's fort/will save are not so high/low that the 15% increase isn't fully utilized.

.15*200 = 30 extra damage on average, or a little over eight extra caster levels on blast spells.

While that isn't a huge amount, obviously, and optimized blasting is a lot better than unoptimized blasting... it's still more than enough to make SoDs valid again, though due to all the assumptions made in the OP, I don't see any evidence they were ever close to being invalid. "If you cast underelevelled, underoptimized fort SoD spells against creatures with high fort saves, it may deal less damage than if you cast one of the better blasting spells in the game and ignore the touch attack and possible energy resistance" is not a strong premise.

[ericgrau]

Before people start inventing numbers to make false conclusions, here's a link to average monster stats:
http://www.giantitp.com/forums/showp...33&postcount=5
Or use Eldariel's graph.

It works even better if you pick real monsters and real spells, b/c then we get the amusement of saying "Uh, it's immune to all of those" b/c your options are in fact highly limited, etc. and "Ya, that's against its low save, and it still usually passes".

If you optimize using non-core splatbooks tricks then:
a) The DM should likewise make monsters stronger or nothing will be hard
b) There are cheesy ways to meta-magic damage through the roof too, and allowing one without allowing another really doesn't show anything except that maybe splat-books aren't the greatest baseline.

[Tyndmyr]

The OP is doing it wrong.

First, you don't just prepare one SoD over and over again. You prepare several types of SoD or SoL (Save or Lose) so you can target all three save types. All those creatures up there have low Reflex saves, so you hit them with a Reflex save debuff of your choice and cripple their ability to move and attack. I'd have to do a little digging to find which particular spell would be the most appropriate at that level, but its out there. There are that many splat books.

Debuffs are great, but they don't disprove the issue regarding SoDs. Reflex SoDs are extremely rare(I can't think of one off the top of my head), but there certainly are plenty of good debuffs out there. Battlefield control/debuffing and SoDs are different.

Second, who cares if your party half-killed the monster before you use your Save or Die? The monster is just as dead and probably sooner than trying to grind through its hit points.

The point is, the effective value is less. Killing a half dead monster is less valuable that killing an unharmed one. Running out of hitpoints kills things too, it doesn't matter HOW you kill it, only that it's dead. Hitpoint death is no less dead, and it's a very universal method of killing things.

Third, whether a GM cheats or not varies from table to table. I let my players one-shot bosses if they can, since I tend to optimize my boss creatures (Hexblade 3/Paladin of Tyranny 3/Rogue 2 can survive alot). The GM's leanings always need to be factored in in your own playstyle, but that doesn't apply to discussions of the rules system as a whole.

Common attributes shared across many GMs can be. For instance, drowning to heal, while rules legal, is unlikely to ever be seen in practical play, and thus, need not be factored into a practical discussion. Cheating is apparently fairly common, and while yes, if you have a DM that doesn't, it helps make certain tactics more viable, it's something worthy of consideration.

For those who are uncertain, tracking the effectiveness of surprise round SoDs against known mobs is a good way to detect cheating.

[Aran Banks]

. . . optimizing contest time? Level 20 SoL v. Arcane thesis B.S. damage spammer?

[Tyndmyr]

Well, we already have Cindy as a hyper-reliable damage spammer. I believe it was guaranteed *enough damage to kill any printed mob*, unless four attack rolls came up 1's in a given round. Details like SR and antimagic fields were irrelevant.

Given that SoDs still autofail on a 20, I think the ultra-optimized approach is a clear win for straight damage at any point pre-epic.

In epic, all bets are off, because the system goes wonky. However, finding epic challenges that are not immune to SoDs is itself quite challenging. Off the top of my head, I believe every singe diety and elder evil is immune.