Tweaked save/BAB progression

[Kerrick]

I sat down today to figure out saves, in part because I wanted to test out a modification to spell saves - DC = 10 + 1/2 CL + mod (more on this later). First, I used a set of formulae that someone had given me for the BAB calculations:

Item Stat Boost: +1/5 levels (or really 1/6 for non-epic or non focused)

Stat boost (from level progression): +1/8 levels

Note these are modifiers, not the actual stats - at 15th level, frex, the stat bonus you get from items like bracers of Dex is +3.

Now, with those in mind, I devised a series of formulae to account for different saves, classes, and stats - clerics might have a high Fort save, but they're not going to dump a lot of points into Con, most times. Here's what I came up with:

High save, optimized - Clr/Drd (Will), Mnk (Ref/Will), Rog (Ref) = 2 + (level/2 + 2) + (level/5) + (level/8)
High save, mid-range - Bbn/Ftr (Fort), Brd (Ref), Rgr (Fort/Ref) = 2 + (level/2 + 2) + (level/6) + (level/12)
High save, low end - Brd (Will), Clr/Drd/Mnk/Pal (Fort), Sor/Wiz (Will) = 1 + (level/2 + 2)

Low save, optimized - Pal/Rgr (Will) = 2 + (level/3) + (level/8)
Low save, mid range - Bbn/Ftr (Ref) = 2 + (level/3) + (level/9)
Low save, low end - Bbn/Ftr/Rog (Will), Brd/Rog (Fort), Clr/Drd/Pal (Ref), Sor/Wiz (Fort/Ref) = (level/3)

"optimized" means the class' save is modified by a stat that starts high (usually the highest) and is always boosted - Reflex for rogues, for instance.

"Mid-range" is a save that is modified by a stat that is secondary and will probably be boosted (Con for fighters), but not as much as a primary stat.

"Low end" is saves modified by stats that are relatively unimportant to the character - Wisdom for bards, frex. Fort saves for clerics and druids might be mid-range, but paladins suffer from MAD - they'll dump points into Cha, Wis, and Str (in that order) before Con, IMO.

This doesn't take into account save bonus items (rings and cloaks of resistance) or the +2 save bonus feats, because those appear maybe 50% of the time. I assumed the stat bonuses started at +2 (could be +3 for optimized high save). I tested this against the half-dozen or so characters I had, and the results were very close, when I accounted for anomalies like the rogue with boots of Reflex +2 and the cleric with higher Dex than normal.

Oddly, though, the low-end high save is pretty close to the low saves - at 20th level, the high saves range from +20 to +13, while the low saves range from +11 to +7. The obvious problem here, of course, is that boosting the low save will result in them being higher than the low-end high save. This isn't really as much of a problem as it looks, though; paladins are an anomaly anyway, since they get Cha bonus to all saves, and they and rangers are half-casters, so their Wis bonus isn't going to be very high to start with (see below). Only Dex-based fighters will have a decent Reflex save (and they'll likely multiclass with rogue), so we can toss them out as well. That leaves only one class - the barbarian, and guess what - he'll have a high Con score (for his rage), which means his high save (Fort) will be well out of reach of his low saves.


Now, as far as spell DCs are concerned, I used pretty much the same system: DC 10 + 2 (starting stat mod) + 1/2 CL + item stat boost + level stat boost. Since the CL for half-casters (ranger/paladin) is already half level, they use 10 + 1/4 CL instead of 10 + 1/2 CL. I also assumed that since they didn't need as high a Wis score, it would start lower and be boosted less (probably around 2-3 points total, so level/7).

Optimized - Brd, Clr, Drd, Sor/Wiz: DC = 10 + 2 + (CL/2) + (level/6) + (level/8)
Lesser - Pal/Rgr: DC = 10 + 1 + (CL/4) + (level/7)

So, overall, save DCs ended up increasing by about 1.5 every 2 levels for optimized casters, or ~2 points/3 levels for half-casters. High saves keep up with this trend - they increase by about 1 point/level - but low saves increase by 1 point about every 2.5 levels. So, effectively, the high save has about a 60% of success at any given level from 1-20, but a low save starts at 50% and decreases over time to 35% at L10 and 15% at L20.

By L40, a high save against a L9 spell with CL 40 is almost a guaranteed success (DC 43 vs. +37 on an optimized save), but the low save will always fail - the optimized low save is only +20. You can probably safely assume that the PC will have at least 1 booster feat, and maybe two (the epic version grants +4), so that's +26 total - a slim chance, but a chance nontheless. The problem remains, however - the low-end low save is only +14, so adding another +6 still won't give it a chance of success, no matter how small.

Conclusion: High saves maybe need to be reduced a bit - maybe to half character level + 1 - and low saves definitely need to be boosted, possibly to 1/3 level + 1 or 2. This means that the 1st (and possibly 2nd) level save bonuses will be the same for high and low, but PCs are so fragile at those levels that they could use the boost. As far as multiclassing, it helps PCs that keep taking levels in classes with similar saves (base classes and PrCs, frex) - instead of getting a save that's 2-3 points (or more) lower than it should be, their low save would be closer to normal. The fractional save system would help for multiclass PCs, but since low saves are too low overall, it wouldn't solve the problem.

 

[HeavenShallBurn]

I've been looking at this same phenomena as I built my reworked d20 variant. I've come to roughly the same conclusions, instead of a purely linear base it needs to be a level derived "base +x" formula to regularize the range of difference possible at any one level. Like you I decided that a +/- 4 difference was about right as a base figure. Thing is if the goal is a relationship between highest and lowest you can use the same formula as a base for nearly everything.

Apply it to AC and you have an AC that scales just right for attacks of equal level, of course it means you have to shift to an armor as DR system because piling too many bonuses on top breaks the math.

Apply it to saves and they finally scale right within their own level regardless of how high the level goes.

Getting around the Spell-DC issue takes a rethinking. Save DCs can be conceptualized as a form of magic BAB. So what you do is turn it into a parallel of BAB. Each class gets either a good, medium, or poor MAB/DC independent of whether they actually have spellcasting ability. This also solves one of the issues with caster multiclassing, as the unified progression scheme means a multiclass caster won't be as good as a dedicated one, but also won't be totally useless.

 

[Kerrick]

instead of a purely linear base it needs to be a level derived "base +x" formula to regularize the range of difference possible at any one level.

The problem is that a "base +x" formula for saves will eventually outstrip spell DCs (for 0-9th levels, at least). It's a workable solution, as long as you don't play very-high-level (over 50) D&D. By then, though, you're probably getting into demigod status and dealing with divinities, so the normal rules wouldn't apply anyway.

Apply it to AC and you have an AC that scales just right for attacks of equal level...

Already got that one figured out. You can keep the existing BAB progressions; you just don't apply EAB until the BAB reaches +20 (NOT when the PC reaches 20th level).

...of course it means you have to shift to an armor as DR system because piling too many bonuses on top breaks the math.

Not necessarily. I use Upper Krust's AC formula, and it works quite well:

AC = 10 + ((any CR up to 20)*1.5) + (CR-20)

Basically, multiply any CR up to 20 by 1.5; if the CR is over 20, take anything over 20 and add +1 per point. For example, a CR 30 creature should have an AC of around 10 + 30 (20 * 1.5) + 10 (30-20) = 50. Natural armor is based off the creature's HD and skin type, along with its size:

Large +2
Huge +5
Gargantuan +9
Colossal +14
Titanic +20
Macro-Fine +27
etc.

If you're working with dragons, use this instead:

AC = 10 + size mod + NA (3-4 + 3/4 HD) + age category

Except for ages 10, 11, and 12, where NA is +2/3 HD (I figured this out on my own)

Apply it to saves and they finally scale right within their own level regardless of how high the level goes.

The problem here is, how do I determine what a save DC is for a given level? I could try using UK's CR system to compare monster ability saves vs. level, but HD vary wildly even within a given CR bracket (especially when you start getting into creatures with lots of Su and Sp abilities). Still... it could work.

Getting around the Spell-DC issue takes a rethinking. Save DCs can be conceptualized as a form of magic BAB. So what you do is turn it into a parallel of BAB. Each class gets either a good, medium, or poor MAB/DC independent of whether they actually have spellcasting ability. This also solves one of the issues with caster multiclassing, as the unified progression scheme means a multiclass caster won't be as good as a dedicated one, but also won't be totally useless.

What you're suggesting is pretty much the same as UA's multiclassing rules. I really like this system, though I've never gotten to test it out in play.

Another thing I've been worrying about - creatures with SLAs will probably have higher DCs than PC casters, but this might not be a bad thing.

 

[HeavenShallBurn]

I was speaking of rebuilding the base system of BAB/Saves/DCs rather than purely a single status tweak. They're meant to work together as a unified base mechanic that brings variation range down rather than as individual patches.

The problem is that a "base +x" formula for saves will eventually outstrip spell DCs (for 0-9th levels, at least). It's a workable solution, as long as you don't play very-high-level (over 50) D&D. By then, though, you're probably getting into demigod status and dealing with divinities, so the normal rules wouldn't apply anyway.

If you use it in combination with the MAB system for DCs the DCs and Saves match, as intended. It was meant to be a unified solution.

Already got that one figured out. You can keep the existing BAB progressions; you just don't apply EAB until the BAB reaches +20 (NOT when the PC reaches 20th level).

Yes that's one of the most common epic level house rules I've encountered. My "Base+X" formula was designed to get rid of the divide between epic and non-epic as much as unify progression.

Not necessarily. I use Upper Krust's AC formula, and it works quite well:

Yes, UK formula's do work well. I just have a personal preference for armor as DR systems and find it easier to move most of what's now AC over to DR then bump damage a certain amount as compensation.

The problem here is, how do I determine what a save DC is for a given level? I could try using UK's CR system to compare monster ability saves vs. level, but HD vary wildly even within a given CR bracket (especially when you start getting into creatures with lots of Su and Sp abilities). Still... it could work.

Largely the HD gaps are caused by a cascade from the different BAB/Saves/etc of the various monster types diverging so much, and the rules about Y HD per CR. If you're regularizing you might as well regularize the entire system which brings CR back into a much closer relationship to HD.

Another thing I've been worrying about - creatures with SLAs will probably have higher DCs than PC casters, but this might not be a bad thing.

SLAs would have the same DC as that of a caster of equivalent level though, which would actually bring things back into line. You would be divorcing caster levels from DCs to an extent. Caster Levels would still determine how many spells and of what levels were available but DCs would be independently scaled to an appropriate amount for them relative to the total character level. For creatures with SLAs caster level would no longer enter the picture for DCs at least. CL equivalence would be purely a matter of effect calculation and duration. I'm still trying to figure out how to rework effect scaling for a tie into MAB without giving up some of the granularity of CL, haven't managed it yet though.

 

[Kerrick]

I sat down yesterday and went through the MM, noting DCs for the abilities of about 80 monsters. It's not a very large selection, I know, so I had to take CR groupings in twos - 1-2, 3-4, etc. - and compare those to an equivalent PC level range (I took the average number, between the high and low end for each save). Now, given that this is shaky at best (I could only gather enough data for up to CR 14), here's what I came up with:

High saves stay pretty steady over all levels - 55-60% chance of success. Low saves are also fairly steady (40-45%), decline at CR 9 (to 35%), then drop sharply at CR 13 (20-25%).

I tried an alternate progression for the low save - 1/2 level -1. It stays a lot steadier over all levels up to CR 13 (~40%), then drops again (to around 25%). I should note, though, that I had only a few CR 13-14 monsters, so that level could well be suspect too. I also compared it to the spell save DCs (the ones that add CL); the low save still declines over level, but very slowly - at L17, a PC has a 25% of success, vs. 10-15% for the existing low save.

I'm not exactly sure what a good percentage range is for success, so I can't really make any conclusions based on the above, but I think the altered low save is very close to what I want.

If you use it in combination with the MAB system for DCs the DCs and Saves match, as intended. It was meant to be a unified solution.

What is it this MAB system you keep talking about? I mean, I know you explained it, but can you show me how it works in practice?

Yes that's one of the most common epic level house rules I've encountered. My "Base+X" formula was designed to get rid of the divide between epic and non-epic as much as unify progression.

I've never seen it, myself. AFAICT, there is no way to avoid using the EAB - high and medium ABs progress faster than AC, so you must use the EAB after a certain point.

SLAs would have the same DC as that of a caster of equivalent level though, which would actually bring things back into line.

The problem is, there's no rule for setting the CL for monster SLAs. UK uses HD = CL, which I intend to adopt (it's simple and makes sense).

 

[HeavenShallBurn]

What is it this MAB system you keep talking about? I mean, I know you explained it, but can you show me how it works in practice?

It's a BAB for spellcasting, instead of a singular set DC you have an effective Magical Base Attack then roll. It's designed for use with the 'Take 10 Saves' variant from the DMG. But also serves the purpose of un-gimping the saves of multi-classed casters. It's based on the UA magic rating system but forms a mirror image of the tweaked BAB progression to unify the entire BAB/Saves/DC paradigm.

Good MAB/Save/BAB=1/2 Character Level+4
Medium MAB/Save/BAB=1/2 Character Level +2
Poor MAB/Save/BAB=1/2 Character Level

A class/creature/etc has an MAB whether or not it actually gets a caster level with which to cast any magic. When multiclassing into classes with different Progressions you either meet in the middle or if none are poor take one step up, assuming the most recent class is the highest. If your most recent class was a step down you do the same in reverse.

I've never seen it, myself. AFAICT, there is no way to avoid using the EAB - high and medium ABs progress faster than AC, so you must use the EAB after a certain point.

Which is why you then take the unified BAB/Save/MAB progressions and shift the AC curve into parity with a bonus based on character level.

The problem is, there's no rule for setting the CL for monster SLAs. UK uses HD = CL, which I intend to adopt (it's simple and makes sense).

It's the same method I use and it seems to work so far.

 

[Kerrick]

It's a BAB for spellcasting, instead of a singular set DC you have an effective Magical Base Attack then roll.

Sounds a lot like the 4E system - attack roll vs. defense.

Which is why you then take the unified BAB/Save/MAB progressions and shift the AC curve into parity with a bonus based on character level.

That's great for PCs, but what about monsters? The tests I was doing were PCs vs. monster ACs.

 

[HeavenShallBurn]

Sounds a lot like the 4E system - attack roll vs. defense.

The mechanical end is a lot like 4e. I originally came up with the first parts in about 2003. Then began experimenting with various values until I came up with this right around the time SWSE came out. In many ways me and the 4e designers independently arrived at similar basic mechanical systems because we were looking for the same kind of scalability fix.

That's great for PCs, but what about monsters? The tests I was doing were PCs vs. monster ACs.

It's one of the reasons I went to an armor as DR system. You account for monster ACs by giving them a HD related AC progression like the PCs. This keeps the ACs roughly on level with attack bonuses of similar level PCs. Then you transfer most of the factors that used to account for AC (natural armor, deflection, etc) into varying levels of DR. Natural armor for example becomes 1/2 hardness+1/2 natural armor AC bonus.

For example a lion is AC=15 (-1 size, +3 dex, +3 nat)
It becomes AC=14 (10 + 2HD bonus - 1 size + 3 dex) DR=2 (1 hardness + 1 NA)

At 5th level a fighter with 18 str will be swinging AB=10 (6 base(5/2+4) + 4). So he'll hit fairly often (anything over 4). But a wizard with 10 str will be AB= 2 (2 base(5/2) + 0 str) will need to roll a 12 or higher to hit it physically. And even a fighter with a middling str of 15 will hit on a 6 or higher. But even under standard rules a lion isn't hard to hit.

Let's take something that should be a bit more slippery, a succubus.

Standard Rules AC= 20 (+1 dex, +9 nat) DR= 10 cold iron or good
Modified Rules AC=14 (10 + 3HD + 1 dex) DR= 15 cold iron or good (10 pre-existing + 1 hardness + 4NA)

What they loose in AC they gain back in DR. So they're easier to hit but harder to punch through and damage.
[edit]I've been considering giving them a boost based on how slippery they're supposed to be as opponents to bring DR inflation down some. But that would be coming even closer to a monster role style system.

 

[Kerrick]

[edit]I've been considering giving them a boost based on how slippery they're supposed to be as opponents to bring DR inflation down some. But that would be coming even closer to a monster role style system.

If you go by UK's natural armor rules, a succubus would be ~AC 12 (skin = 0 + 1/8 HD). I'm not sure that boosting their DR is a good idea, unless you use the house rule that additional plusses on a weapon reduce the DR by 5 points (frex, attacking a succubus with a +1 weapon only bypasses 5 points of DR instead of all 10).

 

[HeavenShallBurn]

If you go by UK's natural armor rules, a succubus would be ~AC 12 (skin = 0 + 1/8 HD). I'm not sure that boosting their DR is a good idea, unless you use the house rule that additional plusses on a weapon reduce the DR by 5 points (frex, attacking a succubus with a +1 weapon only bypasses 5 points of DR instead of all 10).

I've been giving bonus damage instead. It's kind of a toss up which one would work better.