|
(I did some searching before posting this, but I can't find this already stated... feel free to give me a link to an existing thread if there is one)
I'm running some numbers, here, and I'm just not seeing how Powerful / Deadly Sneak (rogue talents from APG) are ever worth taking. The down-side of the talents is that you get -2 on your attack roll. The upside is that you get to treat 1s as 2s (Powerful) or 1s and 2s as 3s (Deadly).
OK, so that leads us to a fairly simple bit of simulation. Roll 10,000 attacks at a spread of +0 to +20 attack vs. AC 10. If the roll comes up as a success by 0 or 1 point, you add a full sneak-attack plus base damage to the penalty you're taking for the talent (you missed when you would have hit). If the roll comes up a success by 2 or more you roll your d6s and add any bonus you get to your running tally.
For even low-damage weapons and no damage bonuses, my simulations for 1 through 5 sneak-attack dice show that you take an overall penalty even when you're at +9 attack vs AC 10! Even at +10 attack vs AC 10, you need to have 3 SA dice before even Deadly Sneak will give you the equivalent of +1 to damage per roll. So this isn't a talent, it's a handicap. I haven't tried adding in critical hits, but as SA damage isn't multiplied on crits, this would actually make the talents worse, rather than better!
Even more disturbingly, you want higher SA dice pools to get any benefit, but if you use SA against a target that has a higher AC than you thought (and remember, you have to have nearly no chance of missing to make this a benefit), then the increased number of dice work against you, multiplying to penalty that you take.
Am I mis-reading?
Below I'm only considering the case of deadly sneak with a 1d6 weapon with no damage bonus. "Attack delta" is equal to the attack bonus of the rogue minus the AC of the target (so +8, +9 and +10 vs. AC 10).
1 dice:
Attack delta -2: -0.277 per roll
Attack delta -1: 0.084 per roll
Attack delta 0: 0.479 per roll
2 dice:
Attack delta -2: -0.435 per roll
Attack delta -1: 0.195 per roll
Attack delta 0: 0.944 per roll
3 dice:
Attack delta -2: -0.825 per roll
Attack delta -1: 0.327 per roll
Attack delta 0: 1.405 per roll
4 dice:
Attack delta -2: -1.122 per roll
Attack delta -1: 0.384 per roll
Attack delta 0: 1.889 per roll
5 dice:
Attack delta -2: -1.132 per roll
Attack delta -1: 0.509 per roll
Attack delta 0: 2.371 per roll
On the other hand, if you use -1 as the attack penalty, the numbers are all still fairly small, but at least positive for Deadly Sneak and +9 or +10 vs. AC 10 will actually work out as a benefit with Powerful Sneak. I'm still not sure if the talent would be worth taking over anything else, but at least if you buff it in this way as a house rule, it would give you a reason to want to use it, ever.
That's way more math than I needed to decide this rogue talent wasn't really worth it. As combat characters the last thing Rogues need is a penalty to hit for some minor conditional damage boost.
I'm curious if, as a houserule, these talents increased the damage die of the sneak attack would they then be worth it? Powerful Sneak Attack takes them to d8s and Deadly Sneak Attack takes them to d10s. Or would bigger dice become even more of a trap?
|
I'm curious if, as a houserule, these talents increased the damage die of the sneak attack would they then be worth it? Powerful Sneak Attack takes them to d8s and Deadly Sneak Attack takes them to d10s. Or would bigger dice become even more of a trap?
Only worth it without the attack penalty. As others have said, a 3/4 BAB TWF character does not need anything that gives him a -2.
|
|
Hmm... now I'm thinking of a few edge-cases that might help.
My simulation didn't account for 20 being an automatic hit. That helps a bit, but not as much as you might think. +7 against AC 10 is still always a penalty for Deadly Sneak.
What you'd really want to do is take this talent when you have some other way of guaranteeing (or nearly so) a hit, and perhaps especially against creatures with damage reduction. But still, I'm having a hard time justifying it when things like Slow Reactions and Bleeding Attack give damage increases that are so much better with no penalties.
Now... if you removed the attack penalty... that might be something worth taking. In fact, given that it's just raising the low-end damage (and thus never affects how much damage you can do), I would think this would be the way to go.
|
|
From a statistical point of view, Powerful Sneak is worthless. It increases the average damage per sneak die by 1/6, so a level 11 rogue would deal about one point of damage more per attack.
Deadly Sneak is a bit better, increasing the average damage by 1/2 per sneak die but still seems weak.
It's also worth noting that Powerful Sneak and Deadly Sneak prohibit you from using other rogue talents like Bleeding Attack, Crippling Strike or Befuddling Strike which apply no penalty on attack rolls.
|
|
I'm curious if, as a houserule, these talents increased the damage die of the sneak attack would they then be worth it? Powerful Sneak Attack takes them to d8s and Deadly Sneak Attack takes them to d10s. Or would bigger dice become even more of a trap?Only worth it without the attack penalty. As others have said, a 3/4 BAB TWF character does not need anything that gives him a -2.
Well, you'd be insane to pile this on top of TWF, but not all rogues are melee attackers... The Sniper archetype was more along the lines of what I was assuming you'd put this talent on, but even then... I think it makes sense to just house-rule away the penalty entirely. It is supposed to be a benefit, after all.
What about multiclass rogues with a dip in a spellcasting class (heading into Arcane Trickster for example) that used touch attack cantrips to deliver sneak attacks? Would that make any kind of difference? Or would the lower damage amount on a touch attack negate the increased sneak attack numbers?
Thank you for doing the statistics!
However, as was already stated, the math breakdown was not needed to reveal the stink of these Tricks.
These tricks need to at least be useful! Heck, even allowing an extra die, or two dice for Deadly Sneak would be better. But to treat crap as more crap is simply crap.
I'll take Honeyed Words, thank you very much :)
GNOME
|
|
From a statistical point of view, Powerful Sneak is worthless. It increases the average damage per sneak die by 1/6, so a level 11 rogue would deal about one point of damage more per attack.
Deadly Sneak is a bit better, increasing the average damage by 1/2 per sneak die but still seems weak.
It's also worth noting that Powerful Sneak and Deadly Sneak prohibit you from using other rogue talents like Bleeding Attack, Crippling Strike or Befuddling Strike which apply no penalty on attack rolls.
You're being too harsh. I do think that removing the attack penalty makes sense (and without it, this talent is only useful in very, very edge cases), but you're way over-simplifying to say that this is the same result as adding 1/6 to a sneak die. You're preventing a 1 from being rolled. That means that on 3d6 SA damage, you can't roll less than 6. For some applications, that's a big deal (e.g. DR 5).
This change in the shape of the bell curve is profound. 10 and 11 go from being 25% of the results to nearly 30%. Your chance of rolling a 10 or higher goes from 62.4% to 70.9%. Previously you had a 4.7% chance of rolling 3, 4 or 5 on 3d6. With Powerful Sneak, you have a 0% chance of rolling those results. Tell the guy that just got 4 on his SA that this talent isn't worth it, and I think they might not agree...
|
|
You're being too harsh. I do think that removing the attack penalty makes sense (and without it, this talent is only useful in very, very edge cases), but you're way over-simplifying to say that this is the same result as adding 1/6 to a sneak die. You're preventing a 1 from being rolled. That means that on 3d6 SA damage, you can't roll less than 6. For some applications, that's a big deal (e.g. DR 5).This change in the shape of the bell curve is profound. 10 and 11 go from being 25% of the results to nearly 30%. Your chance of rolling a 10 or higher goes from 62.4% to 70.9%. Previously you had a 4.7% chance of rolling 3, 4 or 5 on 3d6. With Powerful Sneak, you have a 0% chance of rolling those results. Tell the guy that just got 4 on his SA that this talent isn't worth it, and I think they might not agree...
Guy that just rolled a 4 on his SA, it's not worth it.
And the change in the bell curve is rather irrelevant.
And the change in the bell curve is rather irrelevant.
Interesting. A thread that started as an exercise in statistics to show how X is subpar now has someone advocating that statistics be ignored in order to support the idea that X is subpar.
IMO, there's one answer the question, "Is X worth taking?" That answer is, "It's worth taking if it fits your character concept and you want to take it."
|
|
Interesting. A thread that started as an exercise in statistics to show how X is subpar now has someone advocating that statistics be ignored in order to support the idea that X is subpar.
IMO, there's one answer the question, "Is X worth taking?" That answer is, "It's worth taking if it fits your character concept and you want to take it."
I'm not saying that statistics are irrelevant. I'm saying that the average damage is much more important than the bell curve.
I calculated how Powerful Sneak and Deadly Sneak affected the average damage dependent on the number needed to hit without the power.
The only case in which Powerful Sneak increased the average damage is when your Attack Bonus is at least one point higher than the targets AC.
Deadly Sneak is a bit better, you 'only' need to hit on a 4 or higher for Deadly Sneak to be of use.
And that neither accounted the other damage a character would be able of dealing nor the reduced chance of confirming critical hits.
If this
IMO, there's one answer the question, "Is X worth taking?" That answer is, "It's worth taking if it fits your character concept and you want to take it."
is your definition of 'worth taking', you shouldn't lecture others about statistics. The power increases your minimum damage on a hit while reducing the chance to hit at all.
So, the correct answer would be 'Only if your character is stupid.' Weapon Specialization (net) is more useful.
In our rp group, one of the players keeps on playing rogues... She plays a diplomatic rogue in one AP, a poisoner in another, a burglar type in another, and a combat rogue in a homebrew campaign.
The combat rogue has two-weapon fighting, and powerful and deadly sneak. Her combat performance is off the charts. We use the standard 20pt buy for attributes for all campaigns. And still, her combat rogue does a lot more damage than her other rogues (with the diplomatic rogue being the least combat-worthy).
We found that deadly sneak really boosts her damage output. So yeah, it was worth taking actually.
|
|
I'm not saying that statistics are irrelevant. I'm saying that the average damage is much more important than the bell curve.
In most circumstances, I'd agree with that statement. However, it's not what you said, previously.
I calculated how Powerful Sneak and Deadly Sneak affected the average damage dependent on the number needed to hit without the power.
The only case in which Powerful Sneak increased the average damage is when your Attack Bonus is at least one point higher than the targets AC.
Deadly Sneak is a bit better, you 'only' need to hit on a 4 or higher for Deadly Sneak to be of use.
And that neither accounted the other damage a character would be able of dealing nor the reduced chance of confirming critical hits.
I think we've beaten the "the attack penalty is a non-starter" thing into the ground. We all agreed on that point a while back. I was saying that you can't just look at the increase in damage as +1/6 per die. It's more complicated than that because you don't get average damage when you roll, you get a discrete result, and when your bell-curve is misshapen, you can't just look at the mean and know much about the results.
This is exactly why we use such even, normal distributions for our random number generation in games. We rely on the intuitive properties of normal distributions. If you remove the normal distribution, much of that intuition goes out the window. This is only compounded by the fact that you just won't make a statistically significant number of SA rolls per session. Even 100 rolls gives you a margin of error that's far too large to discount.
So, the correct answer would be 'Only if your character is stupid.' Weapon Specialization (net) is more useful.
That's true... unless you're going up against a lot of creatures with DR, in which case the picture is very, very different.
On 3d6 SA damage, you have a low, but not insignificant chance of rolling zero damage against a DR 5 target. On the other hand, with PA, you will always do a minimum of 1 damage. That's something to think about that has nothing to do with the mean.
I'd say that, without the attack penalty, the talent would be well worth taking, if situationally beneficial (usually you would probably want to open with something else).
|
|
We found that deadly sneak really boosts her damage output. So yeah, it was worth taking actually.
Well... anecdotal evidence isn't very helpful in this case. You could just as easily have killed a chicken before the game and found that your rogue did a lot more damage. As I said, above, you can't reasonably roll a statistically significant number of SA attacks per session, so you might get very low or very high output during any given session and still be well within the margin of error.
Forgive me for reviving a dead thread but I am currently theory crafting a combat rogue for a Ding-or-Die campaign.
What the hell are you talking about. You are using only THEORETICAL statistics and he is giving you an actual situation. Also, any experienced player who can't "reasonably roll a statistically significant number of SA [which varies depending on what the encounter is] attacks per session" shouldn't be playing a rogue or has a GM that is out to nerf the rogue (or both).
Normally I would agree with you. This board puts way too much stock in math and theory than practical application, however in this instance, I don't believe we have enough information. The story says that the Rogue's damage went up. All we know is that the Combat Rogue has these two Talents. What else is different about them? Isn't it reasonable to assume that the Diplomacy Rogue has a higher CHA leaving less points for combat stats like STR and DEX? What else did the Combat Rogue take that the Diplomat didn't?
We also don't know how often the combat rogue hit. I mean if Combat Rogue does 3 extra points of damage but misses 3 times more because of the penalties, sure the damage per hit went up, giving the appearance that the Talents are useful, but the overall damage is way down because of the miss chance, and really how many of us count our attack to miss ratio and compare it with another character?
|
|
What the hell are you talking about. You are using only THEORETICAL statistics and he is giving you an actual situation. Also, any experienced player who can't "reasonably roll a statistically significant number of SA [which varies depending on what the encounter is] attacks per session" shouldn't be playing a rogue or has a GM that is out to nerf the rogue (or both).
First off, rude.
Secondly, anecdotal evidence is not better than the statistics which have been shown here. These people have proved that the talents are not worth it.
As to your other post (the big jumble of numbers), that proves nothing except that random is random.
Holy thread necromancy, Batman!
Think of it like this: would you take a -2 to hit to do 0.5 more damage per sneak attack die? i wouldn't. Plus you have to spend two rogue talents to boot!
This is not a wine for drinking, this is a wine for laying down and avoiding.
The other thing you're missing is, what if we went Bleeding Attack instead. Then you'd be -2.2, -3.333, -5 damage per hit when compared, and that's not counting the consecutive rounds where that damage is adding up.
Just because it's not mathematically sound, doesn't mean you can't take it and have fun with it. Mechanically speaking the Rage Prophet isn't a great choice, but there are tons of people that take it and love it. Math alone doesn't make a fun game.
|
|
It should be sufficient to remove the penalty on attack rolls. I still wouldn't take it, but at least it would no longer reduce the average damage output.
I'd also list the fact that using Powerful Sneak or Deadly Sneak prevents you from using better Rogue Talents, such as Offensive Defense or Crippling Strike, among its disadvantages.
|
|
Just because it's not mathematically sound, doesn't mean you can't take it and have fun with it. Mechanically speaking the Rage Prophet isn't a great choice, but there are tons of people that take it and love it. Math alone doesn't make a fun game.
First, I want to echo this.
Mathematically, the 1s->2s ability will only ever increase your average damage if you could only hit with a natural 20 with your regular attack. However, the 1s&2s->3s ability will increase your average damage as long as you can hit normally with a 4 or higher (which isn't very often). It breaks even if you can only hit normally with a 5 or higher. Regardless of DR, these statements hold, except when the DR is high enough that the damage is 0 in both cases.
I think the point was to say that unless the dice gods are on your side these rogue talents are not a good idea. How much fun they are was not a factor. It seems that point still stands.
In short someone getting lucky, and rolling 5's and 6's does not mean the abilities don't suck. It means someone got lucky.
Hydra you are also assuming all those attacks hit. That will also be a big factor in how much damage is actually done. Lowering the chance to hit does not make up for rerolling 1's or adding the extra damage. Mathematically it fails, and you can say that I might get lucky in a game and roll well, but by the logic I can dump stats that matter to a character and just hope I roll high when it matters.
Neither is worth it. Powerful is just plain a bad trade, you actually lose damage, on average. Deadly pulls even around level 15, iirc. Not so good for what effectively costs 2 feats, one of them a high level feat.
A normal d6 has average damage of 3.5. With powerful sneak, instead of 123456, you have 223456. That averages to 3.67 damage. In exchange for a -10% chance to hit. This is a VERY BAD trade! It literally never "gets better," the -2 would only ever be worth it if you were only hitting on a 20 or missing on a 1 to begin with. Which are such extreme cases (and ones in which you should be running away or just kicking back and enjoying the squash match, respectively) that no one should really even care about such situatons.
Deadly Sneak has a possible result range of 333456, for an average damage of 4. So for the same -10% chance to hit, you have +0.5 damage per die. That's still a bad trade at level 10, but as I said, at very high levels it eventually pulls even.
Don't waste your feats/talents on these, folks.
(On a side note, if these talents were a way to sack your to hit for more damage on NON-sneak attack situations, I'd be on it like white on rice. It's when the rogue CAN'T get sneak attack that he could use a little help punching through damage reduction and such.)
Just me, but I've always thought anything but Slow Reactions was a waste of sneak attack modifiers. Deny attacks of opportunity? Yes please!
Slow Reactions is rather good, yes. If you're full attacking the same guy, it's kinda pointless to attach to any attack after the first one that hits, though. So having another SA-only talent is useful. And some rounds your party just doesn't need the no-AoOs boon.
On their own, the talents are okay. Nothing special. But when put beside the rest of the talents, and given that the number a rogue can take is so limited, I can never see anyone taking them.
I hope youre not talking about Powerful/Deadly Sneak after math done in this thread has shown that it actually makes your damage output worse.
-2 to hit is a 10% less chance to hit. To simplifiy the math lets say you have a rogue that use a d6 weapon and that does 4d4 SA with no other boni to damage = 5d6 damage (average 17,5). 10% less hits = 10% less damage (a bit simplyfied, with out critical hits, etc). So that is about 1,75 less damage.
d6 average = (1 + 2 + 3 +4 +5 +6)/6 = 3.5
d6 (Powerful Sneak) = (2 + 2 + 3 + 4 + 5 + 6)/6 = 3,66666666666
d6 (Deadly Sneak) = (2 + 2 + 3 + 4 + 5 + 6)/6 = 4
So if you make a touch attack with which you will hit anyway (except on a 1) you will get a damage bonus of 0,16 * 4 (Powerful Sneak) = 0,64 or 0,5 * 4 (Deadly Sneak) = 2.
If you do not autohit the damage "bonus" looks like this:
(Powerful Sneak) 0,64 - 1,75 = -1,11
(Deadly Sneak) 2 - 1,75 = 0,25
The senario gets a bit better with more sneak attack dice and a lot worse with str / magic weapon / etc damage bonis on the attack.
So decide for yourself if a 2 (autohit)/0,25 (normal attack) damage boost is worth two talanets!
Breiti
Just a thought: what about a character making touch sneak attacks? (The case I am thinking of is the underrated Pinpoint Poisoner feat.) If you regularly fight larger enemies, it's quite possible to have touch attacks hit on anything but a 1, even with the -2 penalty to hit.
Indeed, always hitting is one of two situations where Powerful Sneak has a positive effect (the other being always missing). Using Pinpoint Poisoner is actually a nice trick to get that condition more often. But even in that case, the damage increase is pitiful (average +1.67/hit at level 20).
Assuming your only damage came from Sneak Attack, Deadly Sneak "only" needs an 80% hit chance to be better. If you include any sort of actual damage or critical hits it gets worse again. At least the damage is a bit better (+5/hit at 20). Still not worth two talents, but not as actively awful as Powerful Sneak.
Why not just take Deadly Aim if doing that? For -2 to hit, you get +4 damage. With powerful sneak, you get 0.51 extra damage at level 6 (to compare). And Deadly Aim works 24/7, even when you can't get sneak attack.
|
|
I'm currently looking at a build, but I'd assume you could take deadly aim also. Maybe a mysterious stranger dip so you can use your CHA, a stat that is also good for ninjas.
EDIT: Yeah this isn't gonna work very well. As stream of sky posted you're kind of thin on feats and talents. I wasn't really implying that it was gonna be WTF awesome. Just chiming in after the point Mort made with touch attacks. When looking at my gunslinger I quickly realized that anything I could do to drop his to-hit and turn it into something beneficial, was a huge bonus because firing on touch AC means you'll typically be way over your "hit cap". The problem here comes with the sneak attack classes not having full BAB, limited access to feats, and spreading your resources thin with MAD.
2nd EDIT: Actually, someone more clever than me could probably explore it further. I'm not saying that deadly / powerful sneak will be capstone abilities in the build, but I think you might be able to mitigate their negatives to hit by making a character focus on touch AC. Then they are just mediocre, but not putting you into the negatives.
How many feats and talents do you expect to have? I would want Deadly Aim along with the "big 3" (PBS, Precise, and Rapid) all well before Powerful Sneak. If using pistols or revolvers or whatever, TWF tree as well.
Well, a rogue can take it twice via ninja trick. :)
And there are other talents you'll probably want, too. Like Slow Reactions. I just don't see, even if you STILL hit on all but a 1 after TWF and Rapid Shot and Deadly Aim and powerful Sneak penalties (which I frankly find hard to believe, even against touch AC), that the +0.17 damage per SA die will even then be worth the feat/talent cost compared to the other options.
Slow Reactions is rather good, yes. If you're full attacking the same guy, it's kinda pointless to attach to any attack after the first one that hits, though. So having another SA-only talent is useful. And some rounds your party just doesn't need the no-AoOs boon.
Good point, though the damage output of Powerful/Deadly is even more diminished if you're using it on iterative attacks (assuming here your first attack hits with Slow Reactions), given your lower chances to hit.
Slow Reactions is rather good, yes. If you're full attacking the same guy, it's kinda pointless to attach to any attack after the first one that hits, though. So having another SA-only talent is useful. And some rounds your party just doesn't need the no-AoOs boon.Good point, though the damage output of Powerful/Deadly is even more diminished if you're using it on iterative attacks (assuming here your first attack hits with Slow Reactions), given your lower chances to hit.
Oh, I wasn't making that statement to endorse taking Powerful Sneak. I think I've been quite adamant that using it will actually lower your damage output, which means actually paying a feat or class feature for it makes you a sucker.
I was endorsing taking another *good* sneak attack-based talent along w/ slow reactions.
Statistical results are invisible to actual play. Some people are okay with their average being lowered as long as situationally their results are higher.
If you are going into it with eyes open, awesome. The big thing for me, though, is that (based on threads like this one) many people taking them don't actually realize that this is what is occurring. They assume that an ability called "Powerful Sneak" actually makes their Sneaks, in general, more powerful. Cognitive biases can make it "feel" like it is really helping, even if actually checking the data would show otherwise. I guess it gets into a weird theoretical place over whether this is "bad" or not (if a person never knows, are they really being affected by it?), but it certainly doesn't make the talents good, or even okay (in the quality sense, not the "evilbadwrongfun" sense).
Besides, there was a Rogue feat in 3.5 that did the same thing without the full-attack requirement or attack penalty. It wasn't considered particularly good then either. Doesn't the Rogue deserve better than straight worse versions of 3.5 stuff?
|
i had this debate with mike shnider before. as someone who has the awesome "luck" of rolling minimum damage to often for the law of averages to be anything more then B.S. i can tell you that a minus 2 is worth it for me.
now if you look at it on paper it is only a .5 increase in damage over normal with 1d6, but when you get to levels 10+ it could be, and for me it is, an increase of (deadly) 10 points of damage on a successful hit.
now to play this character i would use tripping, grease or any number of spells that use cmb attacks to get my target prone then i would net a +2 on attacks.
i think its worth it over all, and it is one of the better rogue tricks you have to choose from.
Why not just take Deadly Aim if doing that? For -2 to hit, you get +4 damage. With powerful sneak, you get 0.51 extra damage at level 6 (to compare). And Deadly Aim works 24/7, even when you can't get sneak attack.
I don't think you can use Deadly Aim with a gun
Benefit: You can choose to take a –1 penalty on all ranged attack rolls to gain a +2 bonus on all ranged damage rolls. When your base attack bonus reaches +4, and every +4 thereafter, the penalty increases by –1 and the bonus to damage increases by +2. You must choose to use this feat before making an attack roll and its effects last until your next turn. The bonus damage does not apply to touch attacks or effects that do not deal hit point damage.
|
|
Why not just take Deadly Aim if doing that? For -2 to hit, you get +4 damage. With powerful sneak, you get 0.51 extra damage at level 6 (to compare). And Deadly Aim works 24/7, even when you can't get sneak attack.I don't think you can use Deadly Aim with a gun
Benefit: You can choose to take a –1 penalty on all ranged attack rolls to gain a +2 bonus on all ranged damage rolls. When your base attack bonus reaches +4, and every +4 thereafter, the penalty increases by –1 and the bonus to damage increases by +2. You must choose to use this feat before making an attack roll and its effects last until your next turn. The bonus damage does not apply to touch attacks or effects that do not deal hit point damage.
You can:
Early Firearms: When firing an early firearm, the attack resolves against the target's touch AC when the target is within the first range increment of the weapon, but this type of attack is not considered a touch attack for the purposes of feats and abilities such as Deadly Aim. At higher range increments, the attack resolves normally, including taking the normal cumulative –2 penalty for each full range increment. Unlike other projectile weapons, early firearms have a maximum range of five range increments.
i had this debate with mike shnider before. as someone who has the awesome "luck" of rolling minimum damage to often for the law of averages to be anything more then B.S. i can tell you that a minus 2 is worth it for me.
now if you look at it on paper it is only a .5 increase in damage over normal with 1d6, but when you get to levels 10+ it could be, and for me it is, an increase of (deadly) 10 points of damage on a successful hit.
now to play this character i would use tripping, grease or any number of spells that use cmb attacks to get my target prone then i would net a +2 on attacks.
i think its worth it over all, and it is one of the better rogue tricks you have to choose from.
We are not saying the abilities are guaranteed to fail. We are saying the dice gods better be on your side, just like the dice gods have to be on your side anytime you take a lesser choice and expect to get above average results.
Tripping, grease and other tactics only get harder at higher levels, not easier.
The argument is not b.s.
Nobody is saying that it is bad/wrong fun to do it. We are just saying that unless you get some good rolls it is a bad investment.