Fixing the 3e Math in a simple and easy way

[Runestar]

I don't want to get into the whole psion thing but suffice to say there are those that believe that they are overpowered. One of the chief things is their capacity to maximize, heighten and scale their spells to their exact need.

Well, suffice to say that I am most certainly not one of those people.

 

[Sadrik]

I'm assuming that a 20th level fighter is hitting an opponent with four iterative attacks. Then I'm seeing how well he does with the iterative fix (Wulf's system) and then how it works with your proposal.

The percentages are the chance of hitting a particular target with a flurry vs an iterative attack. For example, Wulf admits that his fix is a little off when the fighter misses only on a 1. We see this by considering the odds of a fighter hitting someone with a full attack consisting of 4 iterative attacks: the first one hits 95% of the time, the second 70% of the time and so on. The mean damage is 230% of a single attack, which I've expressed as 2.3. Two attacks at no penalty (95% chance of hitting each time) gives a mean damage of 190%, which I've expressed as 1.9. So at this edge case, Wulf's fix would result in the fighter dealing less damage to an opponent- the equivalent of 1.9 hits rather than 2.3 hits. That's 82.6% of the normal damage.

If we were using your proposal for BAB ("Sadrik BAB") then all the percentages are 30% lower (for the -6 virtual penalty). Two attacks at 65% is only 1.3 hits total. With your latest suggestion it could become 4 attacks, with a total expected damage of 2.6. Now it's better than the standard case, which may be OK. (13% more damage than the standard isn't huge).

Wulf's system is designed to give similar mean damages when there's about a 70% chance of hitting the opponent on the first attack. I used 75%. The dual flurry gives 1.5 hits (75% twice) and the iterative attack gives 1.55. Very close.

The 30% penalty for Sadrik BAB means that the attacks are at 45%. Two would be 0.9, three would be 1.35, and four would be 1.8. The last would be 16% more damage than the standard 1.55, which, again, is probably fine.

The place where the 30% virtual penalty causes the most trouble is when the chance to hit (for a standard fighter) is low. Wulf says (I think rightly) that for half-way decently designed monsters they will never happen at upper levels. But just for kicks, let's see what the numbers would be.

Suppose there's a 40% chance of hitting the creature. For a dual flurry (Wulf's system) that's 0.8 hits. For a standard iterative attack it's (.4 + .15 + .05 + .05) = 0.65 hits. A 23% discrepancy.

But for Sadrik BAB it's 0.4 hits if you flurry four times. Each attack has only 10% chance of hitting. That's only 61.5% as much as the standard. A bigger discrepancy. But hopefully this won't happen too often; monsters shouldn't have that good an AC.

Long story short, it seems that adding more iterative attacks helps make up the virtual penalty at higher levels. There might still be the tendency for them to be a "flurry of misses" though.

The exact details of the progression is unclear. I'd guess single attack, dual attack at -2, triple attack at -1, and quadruple attack at no penalty.

But even then... suppose the fighter has to move and attack? He's got a 45% chance of hitting vs 75%. That's a lot less damage on average. Only 60% of what Wulf's fighter or the PHB fighter would do.

Ok, I think I have a handle on what you are doing with the numbers now.
So if a 16th level warrior type (+16/+11/+6/+1) is attacking a raw AC of 18 - ignore bonuses for stats and magic on both sides. This would give him a 95%/70%/45%/20%.

It seems that you would need more scenarios than just this one to determine the complete outcome of a system. Because you might have more or less change depending on a low AC or a high AC. *Going back re-reading your original it appears you have considered this.*

Ok so assuming the same raw AC of 18 vs. +12/+12/+12/+12 you wind up with 75%/75%/75%/75%. Under this it winds up 3.0 compared to 2.3.

Raw level 16 BAB vs. raw AC 23 --> 5 points higher
2.0 vs. 1.4
Raw level 16 BAB vs. raw AC 13 --> 5 points lower
3.8 vs. 3.0

It looks like this is significantly more powerful than the standard rules.

Perhaps the key is a simple -2 to do all of your attacks, similar to rapid shot. Using this assumption we get:

Raw level 16 BAB vs. raw AC 18
2.6 vs. 2.3
Raw level 16 BAB vs. raw AC 23 --> 5 points higher
1.6 vs. 1.4
Raw level 16 BAB vs. raw AC 13 --> 5 points lower
3.6 vs. 3.0

It looks marginally better. But considering that you are rolling vs the AC of 23 there is more of a whiff factor (miss with all) and vs. the 13 their is a auto hit factor with all of your attacks. These seem to be acceptable accommodations for the simplicity of rolling all at the same value.

So the rule with this BAB system would be full attacks give you bonus attacks but they are all at -2 rather than higher BAB max with -0/-5/-10/-15.

Figuring out the other break points:
6th level vs AC raw AC of 8/13/18
+7 for standard action or +5/+5 for full attack
11th level vs raw AC of 10/15/20
+9 for standard action or +7/+7/+7 for full attack
16the level vs raw AC of 13/18/23
+12 for standard action or +10/+10/+10/+10 for full attack

 

[Voadam]

One of the criticisms of high level 3e math is that the discrepancy between 1/1 BAB and 3/4 BAB becomes more evident and turns rogues and monks whose combat schtick is physically hitting things well into characters into specialists against soft targets only as they miss a lot against CR-Appropriate opponents with decent ACs and on top of that have less iteratives.

At high levels this proposal brings fighters and other warriors down to that level and pushes rogues and monks further down. (Clerics and druids have so many spell options they can still be competitive with wizards and sorcerers).

 

[Sadrik]

1) If you want to go this way, and you want to follow the same DC formula that everything else in the game uses, then use 10 + 1/2 caster level + ability mod.

This is true, however, the DCs seems to outstrip the save bonuses in some cases. Classes that have a good will save and utilize wisdom, or dex and ref or fort and con seem to have a really good shot at succeeding against maximized DCs at the upper levels, but anyone else falls short and not by just a little they fall way short. Standardizing the saves is the beginning of the solution and leveling the DCs is the completion of that loop. So, to continue I think that the 10 + 1/2 CL is too good for the number spread that we are looking at. It also gives 1st level spells up to a +9 bonus and that might be a bit high.

2) But I wouldn't. You're generally correct that "low level spells are not as useful" but unfortunately that is not always true-- it's only a tendency, not a rule. Tasha's Hideous Laughter is just one example.

I don't buy the argument. If it is too good at upper level with a level appropriate DC then it is too good at lower level too then.


Another proposal I had for 3e and fixing the top end of spellcasters. I can't find the thread now... Mind you this is very sweeping and requires new charts and what not but I thought that the spells could be broken into every 3 levels you get a new spell level (for full casters) rather than every 2. In this way levels 1-6 spells are in play within 20 levels and levels 7-9 are pushed out into the epic levels, making many of the wonky SoDs and other really ground breaking magic epic. To make this work, you want to give the classes the same amount of spells they had previously so if a level 5 wizard had 3/2/1 the level 5 wizard under this would still have 6 spell slots.

This change would go well with the above every three levels you add +1 DC because in this case it would be: "you get a DC bonus equal to the highest spell level you can cast".

The point of that thread was to limit spellcasters and make them more comparable to non-spellcasters.

 

[Sadrik]

One of the criticisms of high level 3e math is that the discrepancy between 1/1 BAB and 3/4 BAB becomes more evident and turns rogues and monks whose combat schtick is physically hitting things well into characters into specialists against soft targets only as they miss a lot against CR-Appropriate opponents with decent ACs and on top of that have less iteratives.

At high levels this proposal brings fighters and other warriors down to that level and pushes rogues and monks further down. (Clerics and druids have so many spell options they can still be competitive with wizards and sorcerers).

See the post above yours to see how it affects warrior types, you will see that they actually by the numbers fair slight better than the rules as written.

As far as the average BAB it is two points lower than the good BAB and running those numbers against the same AC 18 for a 16th level guy they are on par with the warrior type if they make a standard action, and have -10% to each of their four attacks if they do a full attack. This translates to -.4 on all their numbers so to copy-paste-modify:

Raw level 16 BAB vs. raw AC 18
2.2 vs. standard 16th level warrior 2.3 vs. standard 16th level rogue 1.5
Raw level 16 BAB vs. raw AC 23 --> 5 points higher
1.2 vs. standard 16th level warrior 1.4 vs. standard 16th level rogue 0.8
Raw level 16 BAB vs. raw AC 13 --> 5 points lower
3.2 vs. standard 16th level warrior 3.0 vs. standard 16th level rogue 2.2

These numbers represent the number of hits you can expect against the AC given. As you can see they are higher than the standard rogue. In fact these numbers are only marginally lower than the warriors.

 

[Cheiromancer]

The nice thing about Wulf's fix for iterative attacks is that it is modular. It can be introduced or removed without changing anything else. Sadrik BAB, in contrast, is not modular. You need to change the iterative attack rules to keep it balanced at high levels, and even then you don't address how (at high levels) single attacks have become a lot weaker. Frankly, I don't think Sadrik's BAB proposal will work. A BAB progression of 1/2 just isn't fast enough to keep pace with the improvements in monster AC. Either you get low level fighters who are too strong, or high level fighters who are too weak, or both!

As for the save proposal - this might be worth looking at more. I'm still not clear how multi-classing should work.

Fiddling with Spell DC's... you know, all these changes shouldn't be proposed in the same thread. The discussion should be about either BAB or iterative attacks or saves or spell DC's, not all of them at the same time.

 

[Sadrik]

Fiddling with Spell DC's... you know, all these changes shouldn't be proposed in the same thread. The discussion should be about either BAB or iterative attacks or saves or spell DC's, not all of them at the same time.

This is fair lets do a clean up post:
There are three proposals to clean up and more balance the underlining math of 3e:

1. BAB is equal to half your level. Classes with a good BAB add +4, classes with an average BAB add +2 and classes with a poor BAB add +0. If you are multi-classed you simply average all of them and round down to get your total BAB. You gain iterative attacks when taking a full attack. You get -2 to all of your attacks when taking a full attack. You gain an additional attack at 6th/11th/16th.
2. Saves are equal to half your level. You get a non-stacking bonus of +2 for saves that were good for your class.
3. Related to #2, spell DCs need to be leveled out for a variety of reasons: balance with other non-core classes, predictability in the system, and most importantly ease of use. Two ideas have been floated: the obvious one 10 + 1/2CL + stat and the other 10 + 1/3CL + stat. Both need objective vetting, the first may cause DCs to be too high, the second seems more obscure and does not balance against the non-core classes.