Most Useless Feat

[Numion]

*Sigh* Some of my basic math teachers are spinning in their graves, and they're not even dead yet...

So would mine .. university math professors, at that, if I made such mistakes as you think

In order for the feat to increase your chances of saving by around 10%, you'd have had to been able to save 90% of the time to begin with, which means saving on a 3 or better.

If you can save on a 15 or better (not uncommon at all for fighters making Will saves, for example), your chance of saving is 30%, so having the feat add +2 actually improves it by 33%. It adds 10% to your chance of saving, it doesn't increase your chance of saving by 10%.

Yeah, but that was not what I was talking about. I didn't mean increase in chances, as thats not really telling us much about the usefulness of the feat. Usefulness (topic of this thread) IMO means actually when the feat made a difference. All your increases aside the feat 'works' only when: a) you would've failed without it AND b) succeeded with it. Thats approximately a band of 2 numbers in the d20 roll, because the feat gives +2.

You still following?

2 / 20 means that in 10% of cases does the feat matter. 90% of rolls you would fail regardless of the feat, or you would've succeeded even without it. Basic probability math. Now, that 10% is indeed an approximation since 1 always fails and 20 succeeds, IIRC. It wouldn't chance the numbers a lot.

It might only come up on 1/10th of d20 rolls, but account for 1/4th of all successful saves

But that 1/4th is the conditional probability for the usefulness of the feat .. with the condition being that the save succeeds in the first place!. Not very intresting, since you've conditioned out an important part (unsuccesfull saves).

 

[Taneel BrightBlade]

random note: did you know that simple weapon proficincy applies to all simple weapons if taken once.

 

[Li Shenron]

random note: did you know that simple weapon proficincy applies to all simple weapons if taken once.

Of course but also the only character who would really benefit from it is the Wizard. However I have seen a couple of Wizard take one Martial Weapon prof instead, it's one weapon only but better.

 

[GravyFingerz]

The biggest problem I had was the nerfing of Improved Critical and Spell Focus/Greater Spell Focus.

The biggest problem I saw with Improved Critical was combining it with the Vorpal weapon enhancement ... they fixed that.

The other problem I saw with Spell Focus and Greater Spell Focus was all the other stuff that increased the DCs of spells - they nerfed all those.

They should refix those and return them to what they once were.

 

[hong]

Yeah, but that was not what I was talking about. I didn't mean increase in chances, as thats not really telling us much about the usefulness of the feat. Usefulness (topic of this thread) IMO means actually when the feat made a difference. All your increases aside the feat 'works' only when: a) you would've failed without it AND b) succeeded with it. Thats approximately a band of 2 numbers in the d20 roll, because the feat gives +2.

You still following?

2 / 20 means that in 10% of cases does the feat matter. 90% of rolls you would fail regardless of the feat, or you would've succeeded even without it. Basic probability math.

No. 2/20 means that the chance of saving increases by roughly 10 percentage points. However, going from a 20% probability of saving to 30% means that you have increased your chance of success by one-half. This is different to going from a 70% probability of saving to 80%, where you increase your chance of success by only 14.3%. Basically, the more in the hole you are to start with, the more beneficial a save bonus is. This is why fighters should get Iron Will, and wizards should get Great Fort.

Well, EVERYONE should get Great Fort, because the consequences of failing a Fort save tend to be worse than for the other types, at high levels. So it never hurts to have as high a Fort save as you can get. But that's another issue.

Now, that 10% is indeed an approximation since 1 always fails and 20 succeeds, IIRC. It wouldn't chance the numbers a lot.

It's also an approximation because at the top end of the scale, when your save bonus is already high enough that you always succeed anyway, it doesn't matter that you get a +2. You might roll a 2 and still make the save.

But that 1/4th is the conditional probability for the usefulness of the feat .. with the condition being that the save succeeds in the first place!. Not very intresting, since you've conditioned out an important part (unsuccesfull saves).

Of course you need to condition! It's necessary to find the conditional expectation, where the expectation is taken over the distribution of save DCs and conditioning is on your save bonus. You can't really obtain a universal expression for the distribution of save DCs (that's dependent on the campaign), but for any individual distribution of save DCs, you can say something about what would happen, given an individual character's save bonus. Basically, you get a different result depending on have a good save or a poor save: a good save means the extra bonus from the feat contributes relatively less, while a poor save means the opposite.

 

[Numion]

No. 2/20 means that the chance of saving increases by roughly 10 percentage points. However, going from a 20% probability of saving to 30% means that you have increased your chance of success by one-half. This is different to going from a 70% probability of saving to 80%, where you increase your chance of success by only 14.3%. Basically, the more in the hole you are to start with, the more beneficial a save bonus is. This is why fighters should get Iron Will, and wizards should get Great Fort.

You keep talking about relative terms while I use absolute terms. Neither is wrong. The feat makes a difference 10% of the time (or 10% of all saving throws), regardless of the (percentage) increase in probability. Relative change isn't intresting because 30% increase of a small chance .. isn't a big improvement, and is different for different save DCs.

Now I can admit too that the 1/10 difference is more precious for those who tend to fail a lot of saving throws. Just like I would value a million dollars (which of I have little) more than bill gates (who has many). But thats subjective.

Of course you need to condition! It's necessary to find the conditional expectation, where the expectation is taken over the distribution of save DCs and conditioning is on your save bonus. You can't really obtain a universal expression for the distribution of save DCs (that's dependent on the campaign), but for any individual distribution of save DCs, you can say something about what would happen, given an individual character's save bonus. Basically, you get a different result depending on have a good save or a poor save: a good save means the extra bonus from the feat contributes relatively less, while a poor save means the opposite.

All probabilities are conditional, so of course you need conditioning (for instance, all d20 rolls are conditional on the dice having 20 sides)

What you said isn't really necessary since we can assume that rolling for save 'matters' in most cases - DCs are in the range where you can make the save with rolls other than 20. Since the distribution is uniform (for d20), the real distribution of the DCs don't matter because I used absolute terms. The feat still comes into play 10% of time. If you use relative terms, then the distribution is important.

My point still is: the feat makes you succeed in one saving throw in ten that you would've failed without the feat. ) 9/10 you would fail or succeed regardless of the feat.

 

[sledged]

Here's a nearly useless feat: Sharp-Shooting from the CW. Compared to Improved Precise Shot, it's only redeeming value is that it doesn't have a Dex prerequisite.

 

[mmu1]

My point still is: the feat makes you succeed in one saving throw in ten that you would've failed without the feat. 9/10 you would fail or succeed regardless of the feat.

Uh... no. The feat most certainly does not make you succeed on 1/10th of the saving throws you'd have failed without it, because that probability is variable and depends on your base save and the spell DC.

What you're trying to say is that, as long as your chance of making the save without the feat is not more than 85%, the feat comes into play on 10% of the saves rolled.

Of course, looking purely at how often something is used (rather than at the magnitude of the benefit it provides, and what the consequences of failure are if you don't have it) is a really bad way of judging feats, since by that estimation, feats that give skill bonuses on very common opposed skill checks (where there's no set DC, so a bigger bonus always helps) are the most useful thing in the game.

 

[apsuman]

Why don´t you use a Cone of Cold, if you have a 5th level slot? The same DC, and potentially more damage (if you are above 10th level)? You have to prepare and cast it as a 5th level spell anyway, so where is the difference (And it`s even the same school as Fireball, so this shouldn`t be a problem for you if you can cast fireball)?
Or, since you want to attack a rogue (and since you don`t want to hassle with the area of cone of cold) , maybe Dominate Person (hey, you didn`t take him out of the fight! He fights for you now) or Hold Monster might be a better choice? How about using a lower level spell and throw an Ice Storm (no (reflex) save at all, and since a high level rogue will take half level anyway, regardless of the saving throw DC, the damage would be the same as your heightened fireball.)

The idea of a heightened sanctuary or Invisibility vs Undead sounds better, but these are two very specific spells, and even then you could probably solve your problem with spells of that level.
I am not saying there are no situations in which it might useful, just there aren`t enough to waste a feat on it. (it`s a bit like toughness for a 10th level fighter - how often will you survive because you have 3 extra hp)

Spell Saving Throw DCs do, by the way, increase at the same rate as good saving throws, if you only count the highest spell level available (this doesn`t apply for half or 3/4 quarters like Rangers or Bards). Good saves just have an intial +2 benefit - on the other hand, especially fortitude saves will increase at a lower rate, since constitution is usually not a primary stat (unlike dexterity and wisdom), but the casting attribute always is.
Heighten spell doesn`t change this formula, since they only allow you to prepare at an higher level then the original spell, not a higher level than you can cast.

Only Arcana Unearthed provides a good use of heightened spells - but there, they mean something very different.

Mustrum Ridcully

Well, I disagree.

heighten is wonderful for sorcerers. Maybe they don't know cone of cold, but they do know fireball. No prep needed, just heighten and fireball away.

To point out the usefullness of heighten, if a sorcerer knows Charm person at level 18 they could cast it in a ninth level slot versus an eighteenth level fighter and has a DC of 10+9+cha bonus for a range of 23(cha 19) to 28 (cha 28 -- +5 from levels +6 item). If the fighter has a wis of 14 for a +2 bonus, his will save is +8 meaning that at level 20 a heightened charm person has 70% to 95% of losing the save. If the fighter has a wisdome of 20 he still loses his will save against this heightened charm person 55% to 80% of the time.

 

[Numion]

Uh... no. The feat most certainly does not make you succeed on 1/10th of the saving throws you'd have failed without it, because that probability is variable and depends on your base save and the spell DC.

It most certainly does. Let me formulate:

Without the feat you'd succeed on X results out of 20 (=x/20), depending on the DC. With the feat, you'd succeed on X+2 results out of 20 (=(x+2)/20). Now the probability for succeeding because of the feat is

P(succeed with feat)-P(succeed without feat)=(X+2)/20-X/20=2/20=1/10

Now X was the only thing dependant of spell DC or save bonus, and those got cancelled out. This is conditional to the DC being within save bonus + d20 roll range, which isn't unreasonable.

I can't understand why it's difficult to understand that hitting the 2 numbers on a d20 is always 1/10 probability, and only when you hit those two numbers does having the feat matter.

Of course, looking purely at how often something is used (rather than at the magnitude of the benefit it provides, and what the consequences of failure are if you don't have it) is a really bad way of judging feats, since by that estimation, feats that give skill bonuses on very common opposed skill checks (where there's no set DC, so a bigger bonus always helps) are the most useful thing in the game.

If you read my first post you'll notice I wrote that my players always take these feats at higher levels. I just wrote some analysis for greater understanding. Most people don't understand how rarely those feats really matter. But when they do, you're golden, in case of some nasty fort saves.