Spell Discussions: Level 4

Remove all ads!

It has its uses and is most potent where you really need it (ie against the tougher enemies out there). As with the fireshields it increases in power as your enemies do.

 

Not really. Assuming that you find such such a monster which has a sign "save vs. death 3" on its forehead, no MR and doesn't do anything to disturb your caster for two rounds the calculation is the following (FoD saves at -2 penalty):

- 60 percent chance that the FoD still doesn't work.
- 20 percent chance that you have wasted one round since casting FoD would have worked directly.
- Only 20 percent chance that the additionally invested round you could have used for other things (like casting another FoD with 20 percent success chance) leads to success.

But the usefulness of casting GM in combat is even worse since hardly any enemy has low saves and you don't know enemies' saving throws and immunities.
Enemies with low saves have usually over 50 percent MR which will lower the additional success chance of the GM+FoD combo even more. They are often able to kill party members or interrupt your spellcaster during the addtional round you need for the spell combo.

Yes, Kangaxx, Irenicus and the iron golems are already cowering in fear of your powerful GM spells.

Luckily BG2 was made so cheesy-easy that even funny but inefficient tactics can lead to success.

 

The reasoning given against casting GM gives me pause, although I'm still a little doubtful. But thanks for giving me something to consider. Some questions:

Is the chance of success higher if you cast the same 'hit or die' spell twice, instead of one Greater Malison, followed by a 'hit or die' spell? Let's assume that you have your generic average enemy, with an average save score of 10.

What happens if you intend to follow up with more than one spell? Wouldn't the -ve to saves be additive, in a sense? Ergo. The more spells you intend to cast against a victim, the more the benefits of GM shine through?

And also, perhaps a casting of GM (4th level) would be better than two castings of a 7th level spell (Finger of Death/Prismatic Spray).

 

Success chance vs a monster with save vs death of 10:
FoD: 55 percent
GM+FoD: 75 percent
FoD+FoD: 79,75 percent

Casting without GM is also cheaper since you'll only need one FoD most of the time.

The planning of casting more disabling spells after GM doesn't change the usefulness much since the little save penalty will only come into play in the 3rd round if the first save succeeded.
3+ rounds are long, your warriors might already have killed the enemy then (or the other way around).

 

I'm pretty sure Otiluke's Sphere is bugged. I've been using it forever in IWD2 and BG2 and I've never had a problem with it; failed saves were rare if even at all, I can't remember entirely. I wouldn't recommend it with a sorcerer, though.

On another note, do you receive reduced experienced when you use summons? I can never remember, and if you do then the value of Spider Spawn would go down a bit...

 

Hang on a sec.

From what I understand, you roll a number out of 20, and then determine if the opponent's saving throw beats it, right?

So, whether a FoD will be successful = (10+2/20 * 100%) = 0.60%

Whether a FoD will be successful with a prior casting of GM = 10 + 2 + 4 * 100% = 80%

Whether two castings of FoD will be successful is a binomial. So:
http://www.swogstat.org/stat/public/binomial_calculator.htm

Where N = 2, p = 0.6, and t = 1 and 2.
= 0.48 + 0.36 = 0.84 * 100% = 84%

Sorry if I'm wrong with the binomial. I haven't touched them for about 6 years (which is why I used the calc.)

 

My understanding was that you rolled a D20 and if the number rolled was equal to or higher than your saving throw, then you made the save and were alright (so if a monster had a saving throw of 10 then rolling 1-9 would kill it but between 10 and 20 would be safe)

I think kmonster did his calculations with a saving throw of 12 (rather than the quoted ten) by adding the probability of getting FoD to work first time .55 (11/20, from a saving throw of 12) and adding that to the chance of it working a second time, .45(%chance that the 1st one didn't work) * 0.55 (%chance of it working the second time) =0.2475

so 0.55(1st time) + 0.2475 (2nd time) = 0.7975.

With this example, if the FoD worked the first time it would not be cast the second time.

The Mountain Hare, your calculations are based on casting the FoD twice, at the same time (say in a sequencer) and return the probability of 1 FoD hitting or 2 FoD's hitting (I think...its been a few years since I did them too).

Tough enemies with a saving throw of 3, have a 90% chance of making their save (failing only on rolls of 1 and 2). GM would change the odd in your favour to make you hit 30% of the time (rather than 10). If you really, really need to cast an effect on a monster to win a battle (I'm thinking of some of the tactical mods here) then this is a huge bonus.
[the following is using kmonster's style of calculations, because I don't tend to use too many contingencies...and I'm not too crash hot with binomials]
Eg, you find a monster with a save of 3, you could cast:
FoD+FoD+FoD+FoD
0.1+0.09+0.081+0.0729=0.3439

GM+FoD+FoD (saving throw becomes 7, 30% chance of death)
0.3+0.21=0.51

A much higher chance of success can be achieved with a GM.

Of course this is an extreme example that would only apply to some very tough enemies. With your average enemy it is probably not worth it, but in the tougher fights, where you want to get out as many spells with your mage/sorcerer, priests and multiclasses it can make a big difference.

 

There are other factors with GM which have not been mentioned. For instance in the last few examples FoD has been discussed, but since this is a 7th level spell and GM is only 4th level a simply mathematical analysis of probabilities of success does not tell the whole story - what if your spellcaster only has one level 7 slot? In this case you can't use two FoD even if you wanted to.

Or maybe you have other 7th level spells (Mordy's Sword comes to mind) that are contending the 7th level slots. In this case the use of GM makes more sense.

Also the low level of GM allows it to be placed in Spell Sequencer/Spell Trigger. This fact makes the spell even more useful.

 

Stu:

Yeah, I know I messed up, because as you pointed out, if the damage roll = the opponent's saving throw, then it makes its save. So it's actually 11/20. Whoopsie. My bad.

The binomial calculator agrees with you both! The probability of killing a foe with two fingers of death = 0.7975.

Probability of killing a foe with Greater Malison = 15/20 = 0.75

Sorry for the repetition, I'm just typing this stuff on a screen so it makes sense to me.

I always used to think that 'Greater Malison + Save spell' = Always good. But apparently it's not so black and white.

You'd probably be better off just ignoring GM for the lower level spells (Glitterdust, Chromatic Orb, etc.). But with high level spells, it seems more prudent to cast GM beforehand.

In a sense, GM is a 'substitute' spell for the higher level spells. One GM + One Prismatic Spray has only a slightly lower probability of failing to overwhelm the opponent's save than two prismatic sprays.

And as has been pointed out, it's excellent against enemies with tough saves.

Edit: Does this help clear things up, folks?

http://i5.photobucket.com/albums/y188/Red_Wizard/greatermalison.jpg

 

@Stu: My calculation was based on a save vs death of 10 and therefore on a save vs FoD (-2 penalty) of 12.


It's of course no problem constructing theoretical situations where GM can be very useful, but (especially in an unspoiled normal no-reload game) it's not worth blocking a memory slot which could have been used for stoneskin or other level 4 spells.

@MH: Great link ! I always dreamt of a success probality of 1.15

 

kmonster:

Yeah, there really is no such thing as a probability of 1.15. Just round it down to 1. The rest of the graph is fine.

The graph demonstrates that two glitterdusts is approx equivalent to one GM + glitterdust if the enemy has good-average-poor saves. However, GM shines if the enemy has really ****ty, or awesome, saves.

So I guess if you're taking on your generic opponent, you must ask yourself.

1. Which would I rather lose? One 4th level spell, or one X level spell.

2. How important is the initiative (first strike with a spell)?

 

Ah, the manual hints at a -2 penalty, but there is no mention of it in the in-game description. I took it as a manual error, but correct me if I'm wrong.

Also of consideration is how long you intend the battle to last and how many spells you want to cast, you will usually be better off casting GM if you intend on casting more than 2 or 3 spells involving a save.

 

Some sources mention the -2 penalty with FoD, some don't. I guess it's more likely that the -2 was erroneously omitted than added.

Are you serious ?

 

From what I remember from university stats, probabilities can only be between 0 and 1.

So 1.15 means it will happen 115% of the time, so you may as well round it down.

Had he said you should round a probability of 0.115 down to 0.1, the entire statistical community would have died from shock.

 

The "are you serious ?" was directed at the "Just round it down to 1. The rest of the graph is fine." Rounding down the 1.15 to 1 doesn't make the graph error-free.

 

kmonster:

Why not?

 

A natural 1 always fails, a natural 20 always succeeds.

 

^^^All 3 of these lines should be curved because they should all start at zero (and end at twenty), which would throw off the probability a slight amount. But even so, the graph still shows the basic proportions in terms of saves between the three effects rather well.

 

The graph was correct enough for leading MH to a wrong "GM shines if the enemy has really ****ty saves" conclusion.
The graph is only correct from 2 to 16. The purple line should stay at 0.95 and fall below the yellow curve after 16. The yellow and blue curve shouldn't get lower left of 2.
A complete graph would go from -2 to 20, lower or higher base saving throws have the same success probability as -2 or 20.
The graph is still interesting if you know what is right and wrong but keep in mind that it only compares some probabilities, it doesn't show if there's a situation in the game where GM is useful.

 

I had a post typed up, but lost it.

So, to summarize:

- I didn't take saves down to -2, because opponents can't have a -2 save when targeted with Glitterdust. If the spell had been Chromatic Orb, I would have altered the saves accordingly.

- The chance of an opponent failing a save with a saving throw of 20 = 0.95. This is because if the caster makes a roll of 20, the opponent can still save.

- If an opponent has their saves pushed ABOVE 20 (eg. to 21), they will ALWAYS fail their saves, because the caster will always roll a number less than 20.