The Failure of Scaling Bonuses and LA Buyoff Edit
Ghostwheel 18:40, September 23, 2009 (UTC) Edit
I've noticed a few instances where people have tried to add scaling bonuses to things like AC, ability scores, attack, and other things that require a d20 rolled as part or against them, and I'd like to explain why this doesn't work.
Let's say that a character has a scaling bonus to AC; we'll take an extreme example, say they get a +10 natural armor bonus to AC (not an enhancement to natural armor as given by the amulet, so it stacks with it). Character X has said ability, and at level 20 they have A+10 AC, where A is the AC they would have with only items/ability boosts/etc to AC. On the other hand, character Y in the same party buys exactly the same stuff, but doesn't get the +10, and thus has an AC of A. Now, let's say the DM brings in monsters that should be a match for the party, and are able to hit an "average" character (Y, who has buffed up their AC with items, but doesn't get the +10 from scaling) on an 11. This monster will damage character Y about half the time, and miss the other half of the time. However, it will only hit character X on a natural 20. On the other hand, if the DM gets tired of not being able to hit character X and brings in a monster who hits him on a natural 11 or higher, that means that character Y is always going to get hit. In effect, the scaling bonus pushed character X off of the RNG (random number generator) table as far as AC goes, since the roll isn't an opposed check.
Scaling bonuses don't usually grant this extreme of a bonus to the stat, but at the same time they continually make the difference between the die needed to roll ever larger, rather than keeping it static, and keeping the benefit of the ability the same at all levels. (+2 to AC at level 5 gives the same amount of defense against attacks as +2 to AC at level 20--people need to roll 10% higher on average to hit you, or ~20% if you assume that 1d20 rolls around a 10 on average, and needs around a 10 to succeed on the check.)
The same goes to bonuses to Attack, DCs, and other things that rely on the RNG to function--that is, use a d20. Opposed checks are given more allowance here, since the gap is wider, allowing someone to have up to around a +40 to their check before being pushed off the RNG.
On a similar note, LA Buyoff just doesn't work; in fact, it essentially makes LA meaningless, especially the low LA of +1 (and +2 if you start at a high enough level). The difference in XP is so miniscule after a few levels (and by the book, the person at the lower level should receive more XP) that characters reap all the benefits of their LA without any of the penalties. Now, some people might say that the benefits of LA disappear at higher levels, or are subsumed by stronger abilities. This is true for some racially-granted abilities, but not for ones that stack, or for ability modifiers. For example, an ability that would be subsumed might be a race's ability to fly, since one could pick up Winged Boots at higher levels, and the benefit of the race would be gone, effectively worth the cost of Winged Boots, and LA Buyoff would be viable, and perhaps even correct to use.
On the other hand, +2 racial to strength for example, is going to give you +1 to hit and +1 damage no matter what level you are when compared to someone of equal level with the same items. Now, if races gave instead Enhancement bonuses to stats, then it would be true; +2 to a stat would be subsumed at higher levels, effectively being worth 4k gp, but when it stacks with items then saying that it doesn't matter at higher ACs is wrong.
Another example would be +2 to a casting stat. The +2 gives a higher DC and more bonus spells regardless of what level you're playing at as long as LA has been bought off and the two characters have the same items. Thus, one should either change racial bonuses to ability scores to become enhancement, or get rid of LA Buyoff.
Scaling bonuses are totally fine if they are named and replace what would otherwise be an item you buy or something. (In other words, there's precedent for them.)
Ghostwheel 21:15, September 23, 2009 (UTC) Edit
I agree--on one condition, that the equivalent bonus is one that you could buy/receive normally at the level you'd get the enhancement. So, for example, a character that got 1/4th their level as an enhancement bonus to natural armor would be completely normal--they're not over the predicted AC for a creature at their level. But, on the other hand, if they get 1/2 their level as an enhancement bonus to natural armor, they would be over the predicted AC for a creature of their level by 1 at level 4 (when others should have +1 and they have +2, by 2 at level 8 (when they have +4 and other creatures of their level should have +2), and the gap keeps getting wider. So yes, it's alright--if you stay within the acceptable limits of what an equivalent character could buy, but I've seen a number of things that go above and beyond this, contributing to character imbalance as far as the RNG goes.
And if the bonus were 1/3 character level, round up, would that be acceptable?
Ghostwheel 05:07, September 24, 2009 (UTC) Edit
I personally feel that it should be 1/4 levels (see my fixed CWBL thing), but I think 1/3 is fairly close to acceptable *nod* Edit: That's as long as you keep to pre-epic gameplay. If you go epic, the numbers drift too far apart to be completely usable IMO, though I'd need to check CWBL of epic levels to be sure.
Heh. I don't know anyone who takes Epic seriously. It's just too...out there. I suppose one day someone might find a way to codify it into something sane (making it use the same system of advancement as pre-epic would be the ideal starting point.) But I do, however, feel that 1/4 scaling is, well, actually too slow for most stuff. 1/2 would indeed be too fast, but, somehow, 1/3 works well. It makes you care about any magic item you have (If you're going full Book of Gears, you're maxing out at eight items providing a constant bonus, so you pick up stuff you actually care about.)
On the subject of LA buyoff, the best solution was one I got from a DM. If your DM is open to discussion, don't bother with the XP buyoffs or all that bookkeeping. Just assess your DM and pick a number that you think won't make him turn white and crap a brick (Like 3. Learn just how useful that number is. Three wishes, three wise men, three tries...). Suggest that every time the character's ECL hits a multiple of that number, one point of LA disappears and the character picks up a class level.
Andrew Arnott (talk, email) 14:49, September 24, 2009 (UTC) Edit
In the end, a bought off LA PC will be strictly stronger than a non-LA PC. I'm not a fan of that. As a DM, however, I have realized the best way is to just compromise with everything. If Joe really wants to play an Ogre, we can figure out a way to make it balanced. I give minor tweaks to the Vow of Poverty character in the group I'm DMing for that very reason. Later on, I will be giving larger tweaks to the big dumb barbarian.
I think that the whole 3rd edition race system is horribly broken outside LA +0 and even then, not much compares to humans and dwarves. I don't mean broken as in overpowered, I mean broken as in "doesn't work right". I think the "talk to your DM" method is the best solution without a total re-write of the system.
Ghostwheel 15:03, September 24, 2009 (UTC) Edit
A couple of ways to do the ogre, just throwing out ideas here:
- Rather than LA or HD, make the player waste feats to get certain things. Ogres are strong and big and dumb, right? Start them off at +2 str, -2 int. At levels 1-6 they lose feats, and every time they would have gained one they gain +2 str the first two times, and become Large the third (without the Tome shenanigans about +8 str, +4 con, etc). There, effectively an ogre.
- Use the LA + HD as given, but change the bonus to stats to an enhancement bonus and allow for LA buyoff. Since the bonus to stats will be subsumed at higher levels, it allows characters to buy off that LA without taking any sort of penalty.
- Use the LA + HD as given, but allow no LA buyoff since the bonuses always stay with you, even at higher levels.
- Trade each LA for another 2 HD, effectively gaining no HD but spreading the abilities over more levels.
- There's also the F&K way of reducing it to its CR+1 of HD with no LA, and while that might work in wizard-level games, I think it might be overpowered for rogue-level games.
Andrew Arnott (talk, email) 18:49, September 24, 2009 (UTC) Edit
They all are options that work for the specific Ogre case but my point is that the system, in general, does not work as it was intended to. If you want to use something resembling the base rules for races, then you will have to come up with LA on a case-by-case basis. For some groups (like mine) that is perfectly fine.
Ghostwheel 05:52, September 25, 2009 (UTC) Edit
-nod- I think a semi-viable fix to LA (or at least a step in the right direction) is allowing LA buyoff, but making all bonuses from race (to AC, ability scores, etc) an enhancement one. Completely do away with racial bonuses to abilities. (This has the advantage to not allowing players to stack crazy templates one atop another for insane bonuses to scores at relatively little cost/LA, since they wouldn't stack.) Or, on the other hand, not to use LA buyoff, if you keep it the same so that the races retain their advantage throughout their whole progression, but also count as having invested resources for said benefits. Whether the investment is worth the advantages is up to the play and DM of the game.
Ghostwheel 05:34, September 27, 2009 (UTC) Edit
Had a quick question for ya guys; I'm pretty sure that it matters that characters follow about the middle of the RNG in rogue-level games, but should that also be the case in wizard-level games, or should they be able to be all over it? And if they don't keep to the ~middle of the RNG, why shouldn't they?
Ghostwheel 07:01, October 2, 2009 (UTC) Edit
Some people don't seem to understand just how vital +1 to attack can be. Thus, I thought I'd illustration here.
Let's say we have a rogue. He's got 34 Dex (18 + 6 (item) + 5 (level) + 5 (tome)), a +15 base attack bonus, two shortswords (1d6) that have +5d6 misc. enchantments on them, and is twf (-2 to attack.
In short, this is his attack bonus: +12 (Dex) + 15 (BAB) + 5 (Magic Weapon) - 2 (Two-Weapon Fighting) = +30/30/+25/+25/+20/+20
His damage comes out to 1d6 (Shortsword) + 5d6 (magic) + 10d6 (Sneak Attack) + 5 (Magic), which is 16d6+5 or 61 on average.
Thus, against an AC of 40, we have 61*11/20 + 61*11/20 + 61*6/20 + 61*6/20 + 61*1/20 + 61*1/20, which comes out to 109.8 on average. (We'll ignore things like DR and energy resistance for this example, as well as crits and the like)
Now, let's say he gets +1 to attack (from Weapon Focus, perhaps). His attack sequence becomes +31/+31/+26/+26/+21/+21, and damage stays the same.
Now we have 61*12/20 + 61*12/20 + 61*7/20 + 61*7/20 + 61*2/20 + 61*2/20, which comes out to 128.1 on average.
This is a damage increase almost 20 points on average. That's an increase of 16.66% over the previous damage. Now, a damage increase of over 15% from a single feat? That's damn decent in my book. And that increase doesn't have to come from Weapon Focus, but can come from a race. People seem to underestimate the difference racial mods or Weapon Focus can make. But when it's an increase of damage by more than 15%, it's hard not to see the use of it. Sure, it's a boring feat. But that doesn't make it the same as underpowered.
TarkisFlux 07:18, November 15, 2009 (UTC) Edit
I seem to have missed this earlier. You're getting hung up in a single point of numerical analysis and missing the bigger picture. You're right that it's average damage * chance to hit, but you lose some perspective when you jam it all together like that. So I'm going to split it up into bonus damage per hit based on what you need to roll to hit without the bonus.
Need to roll Chance to hit regular Chance to hit with +1 % Original Damage 21 5% 5% 100% 20 5% 10% 200% 19 10% 15% 150% 18 15% 20% 133.3% 17 20% 25% 125% 16 25% 30% 120% 15 30% 35% 116.7% 14 35% 40% 114.3% 13 40% 45% 112.5% 12 45% 50% 111.1% 11 50% 55% 110% 10 55% 60% 109.1% 9 60% 65% 108.3% 8 65% 70% 107.7% 7 70% 75% 107.1% 6 75% 80% 106.7% 5 80% 85% 106.3% 4 85% 90% 105.9% 3 90% 95% 105.6% 2 95% 95% 100%
And we'll leave off 1 because it's an automiss and looks like 2. Unfortunately you can't use the % original damage to get an easy number like your damage increase, it's just there to indicate that get the most substantial bonus if you wouldn't expect your attacks to hit someone. And in those cases the bonus is huge, as you have demonstrated. Hell, the example character's unlikely attack is contributing twice as much as it was previously by just that +1 bonus. But let's take your example and drop the AC for a while and see what happens. If we change the AC to 39, then your damage boost is only 14.1%. If his AC is 38? 12.5% boost. And it will keep going down as his AC goes down, until there is no meaningful difference at all. If we go the other way, and bring his AC up to 41? Damage bonus drops to 12.5% already because his in both cases his tertiary attacks only hit on a 20. It goes up a bit as the bonus to the secondary attack begins contributing more to the damage percentages, and reaches a local maximum of 25% at AC 45, before dropping again to 14.3% at AC 46. It actually gets up to a 33% bonus at AC 49 (and that is the highest it goes), but since that's on top of 1/20 of your attacks even hitting I don't find that substantial at all.
My point with all of this is that you're off base with the impact of a +1 bonus. It really means different shit at different times, and it means a lot more when you can't hit people effectively without it. You can't even determine an average increase of damage because it's entirely dependent upon opposition. What I can say is that against easier opposition it doesn't provide a significant increase in damage, and against harder opposition it provides a more significant increase in damage that doesn't even matter because of the increased amount of wasted rolls you're making.
As a side note, that +1 bonus matters way less with tome iterative attacks, because you're less likely to have a set of attacks that suck against your foe. And if you do have such a set of attacks, you're going to be missing a lot anyway.
ElfsMaster 00:43, February 10, 2010 (UTC) Edit
Sorry for commenting on this "dead" discussion, I just find it interesting to point some things off:
A +1 bonus to attack is, in most cases, more significant than a +1 bonus on damage (except on low level, when they're about the same). This is because, on average, you should be able to hit an enemy with a roll of 8~13, depending on your build and BAB progression (and DM, of course... but the core books state that if the group need rolls close to 20, the monster is too strong, if they hit on a 2 or 3, it's probably too weak, so I'm averaging on 8~13). That leaves the average contribution of a +1 bonus in about 10% (a little more, a little less). So, if you can do more than 10 points of damage in an attack (which is a very low damage actually), +1 damage gives you less than 10% average DPR. This means that, if you have an average damage above 20, you will benefit more from Weapon Focus than from Weapon Specialization.
- Wait. Everything I've seen is that a +1 on a d20 roll is +5%. Do you have a different way of looking at it that makes it a +10%? --Be well 11:14, March 24, 2010 (UTC)
TarkisFlux 15:35, March 24, 2010 (UTC) Edit
He's talking increase percentage, not absolute percentage. Which is fine in the range that he's in, probably, but starts doing weird things at the edges. If you need a 20 to hit that +1 is closer to a 100% increase, for example. And if you hit on a 2, that +1 is a 0% increase. While it sounds fairly awesome, I've come to the conclusion that it just does more to obfuscate the issue than anything else.