
hogarth |

It's silly that a Rog1/Brd1 has a lower BaB than a Sor2. Can we get a couple lines added about being able to sum fractions before rounding down?
I doubt it'll happen. Especially considering that Jason made his own (non-backwards-compatible) system for prestige class saves rather than using the perfectly serviceable fractional save system.

hogarth |

hogarth wrote:I doubt it'll happen. Especially considering that Jason made his own (non-backwards-compatible) system for prestige class saves rather than using the perfectly serviceable fractional save system.That's the beauty of it: This is a more elegant solution than a patch of that nature.
And my point is that Jason seems to prefer patches of that nature to elegant solutions. (just kidding, kinda...) :-)

Dorje Sylas |

I know this additional suggestion won't be made official because it interferes a bit more with backwards compatibility, but from my experience running gestalt games I would also recommend that the +2 for a good save only applies once.
Using the example from UA Cleric5/Fighter 2, Fort +7 (rounded down from 7-1/2), would instead be +5-1/2. Or something crazy like the super adventurer Fighter 3 /Cleric 5 /Rogue 3 /Wizard 3 /Mystic Theurge 2 /Eldritch Knight 2/ Arcane Trickster 2:
With +2 for all good saves (Fort +14-1/3, Ref +11-1/2, Will +16-2/3)
+2 applied only once (Fort +10 1/3, +Ref +9-1/2, +10-2/3)

Ross Byers RPG Superstar 2008 Top 32 |

I know this additional suggestion won't be made official because it interferes a bit more with backwards compatibility, but from my experience running gestalt games I would also recommend that the +2 for a good save only applies once.
I don't think this would be a problem with backward compatibility because most characters are monoclass and won't be affected. Those that do multiclass will still see at most a +1 untill you get into the really weird five-different classes things, and they were going to break anyway.
And PrCs already lost backward compatibility for saves. (Fractional bonuses could already fix the +2 thing, since you'd have one total for god saves and one total for bad saves.)

![]() |
1 person marked this as a favorite. |

Thanks for the vote of confidence Hogarth.
Anyway, fractional saves have been something that I have been considering for a while, but I am not 100% sure that the extra math is worth the effort. It might be relatively simple for you or I, but I am not convinced that everyone else would see it that way (hence the slightly modified system for Pclass saves)
Jason Bulmahn
Lead Designer
Paizo Publishing

KaeYoss |

I have no problem with math, and still don't want to bother with fractional saves. I'd say they're unsuitable and unelegant as the standard. Easy enough to house rule to boot.
I like what has happened with PrC save progressions, though, and think that could be the way to go with multiclassing in general: Just say that the +2 is granted only once for each save type, and that in multiclassing, you get to use the saves (and BAB?) of one level above what you have, at least for all classes beyond 1st (or every other class or something, but that would make it too complicated I think)

Davelozzi |

I for one like the idea of fractional saves but if I am remembering correctly, the explanation in Unearthed Arcana seemed way more confusing than it needed to be.

Ross Byers RPG Superstar 2008 Top 32 |

Thanks for the vote of confidence Hogarth.
Anyway, fractional saves have been something that I have been considering for a while, but I am not 100% sure that the extra math is worth the effort. It might be relatively simple for you or I, but I am not convinced that everyone else would see it that way (hence the slightly modified system for Pclass saves)
Jason Bulmahn
Lead Designer
Paizo Publishing
I'd consider multiclassing an 'advanced' option. New players should stick to a single class, which would still have everything spelled out on the table, so nothing would change for them.
I understand if it doesn't make the cut, but food for thought.

Epic Meepo RPG Superstar 2009 Top 16, 2012 Top 32 |
Another solution to the problem - one that doesn't rely on fractions, which some people despise - is to simply set a minimum BAB and minimum save bonus for each character level.
Just add a few columns next to the feat and ability score increase progressions. Each level, you add up your bonuses as normal, but use the minimal value in every case where the minimum value is higher. (This also helps to solve the problem of huge disparities between the BAB and saves of epic-level characters.)
Say, for example, you create a Brd 1/Rog 1. You calculate your BAB and saves normally, then check the minimums. Your saves are okay for 2nd level, but your minimum BAB is +1, so you use BAB +1 instead of BAB +0. No fractional bonuses needed, but you still keep pace with the BAB of a single-classed sorcerer.

Freesword |
Count me as against fractional BAB and saves...at least the method as implemented.
I like the end result well enough, but the rounding of fractions of a bonus has to go.
I prefer something that has been called like stacking.
You add up the number of levels that give you the same progression (full BAB, 3/4 BAB, 1/2 BAB, Good Save, or Poor Save) and reference on the base advancement chart for what the appropriate BAB or Save would be for a character of that level. When you have 2 or more different progressions from multiple classes you add the results from the chart for each together. In short, add levels where the progressions match before referencing the chart, add results from the chart for progressions that don't match.
The result is that the highest possible result is the best progression for single class and the lowest possible result is the worst possible progression for single class.
This would give Ross's Rog1/Brd1 the BAB of a character with 2 levels of 3/4 BAB. Nearly the same result as fractional is achieved with only adding and referring to the chart.
For the record, I like math and don't mind doing it. I think keeping it to simple add and look up is better than rounding for the game in general.

hogarth |

Another solution to the problem - one that doesn't rely on fractions, which some people despise - is to simply set a minimum BAB and minimum save bonus for each character level.
Just add a few columns next to the feat and ability score increase progressions. Each level, you add up your bonuses as normal, but use the minimal value in every case where the minimum value is higher. (This also helps to solve the problem of huge disparities between the BAB and saves of epic-level characters.)
Say, for example, you create a Brd 1/Rog 1. You calculate your BAB and saves normally, then check the minimums. Your saves are okay for 2nd level, but your minimum BAB is +1, so you use BAB +1 instead of BAB +0. No fractional bonuses needed, but you still keep pace with the BAB of a single-classed sorcerer.
Hmmm...that's an interesting thought. By the same token, you could have a maximum base save per level as well, if you were worried about huge saves from multi-classing.

Dorje Sylas |

seekerofshadowlight, what about a minimum/maximum value? Setting a minimum to the poor BAB/Save value will do basically the same as fractional increases since you really can't have a score lower then that calculating saves that way. A maximum cap could actually be useful for rebuilding 21+ level rules, as it could slowly be capped to prevent totally out of control growth.
The monkey in the middle will be people who keep swapping between Average BAB classes.

Freesword |
Freesword wrote:This would give Ross's Rog1/Brd1 the BAB of a character with 2 levels of 3/4 BAB. Nearly the same result as fractional is achieved with only adding and referring to the chart.What about an aspiring Arcane Trickster, then? A Rog1/Wiz1 would still have a BaB of 0 in your example.
That's why I said "Nearly the same result". The fact that at low level a 1/2 BAB and 3/4 BAB mix can end up below a 1/2 BAB progression is an edge case anomaly. My suggestion may not be mathematically perfect, but does work better than the status quo of 3.x with simpler math than fractional.

Quandary |

Exactly, Freesword:
"Like Stacking" (BAB&Saves) matches "fractional saves" ~95% accurately,
while presenting a simple picture to players.
In fact, because the Class BAB/Save Charts can be kept *AS-IS* (perhaps merely labelling them Good/Average Saves, High/Medium/Low BAB for clarity), until a player decides to multi-class and look up how that works, THEY DON'T EVEN NEED TO KNOW that there's anything beyond their Class Advancement Chart. If people are worried about those wierd corner cases (Rogue 1/Wizard 1 = BAB 0), there can even be a line saying that you can treat a level of a higher BAB Class as if it were of a lower BAB progression (if it would be advantageous).

Majuba |

I'm better with math than anyone I know, and I whole-heartedly oppose having fractional BAB and Saves. decimals are so much better!... j/k
Just displaying those on the leveling charts (and they would *have* to be displayed) would cause far too many questions.
I could get behind minimum BAB (1/2 level round down) and Saves (1/3 level round down). But even this is fairly unnecessary. Until you hit at least 3 classes with the same weakness, this wouldn't result in more than a +1 to any single element.
Even a Fighter 5/Cleric 5 only has one lower Reflex save than a Fighter 10. A Rog1/Brd1 has +0 BAB? So? That loss of +1 is marginal, and does not increase over time.
Is it a big deal right then? Maybe. Is it vital to change? Not in the least.
I will say one thing in favor of this idea though. It would show a progression at every level, and that *could* make the charts make a little more sense to newbie druids, trying to figure out how to change their BAB at 5th level.

Ross Byers RPG Superstar 2008 Top 32 |

I don't think that the actual fractions would need to be displayed, because you're right, that would look terrible and be confusing to new players. Instead, in the multiclassing section of the rules, the formulas for each could be displayed (1/2 level, 3/4 level, and level are not complex formulas).

![]() |

I don't think that the actual fractions would need to be displayed, because you're right, that would look terrible and be confusing to new players. Instead, in the multiclassing section of the rules, the formulas for each could be displayed (1/2 level, 3/4 level, and level are not complex formulas).
Yep. It's a brief calculation, and it fixes a lot of problems. I would love to see stackable favored saves gone for base classes too.

Majuba |

(1/2 level, 3/4 level, and level are not complex formulas).
For you and I, no. But there are players out there (and some of mine are some of them), who will stare at you for at least a minute or three when you ask them to add 1/2 and 3/4.
Really. It's not uncommon. I *hate* it from the depths of my soul, but it's not uncommon.
It also hardly "solves a lot of problems". It gives typically a +1 bonus at most. That's not a problem.

Roman |

Thanks for the vote of confidence Hogarth.
Anyway, fractional saves have been something that I have been considering for a while, but I am not 100% sure that the extra math is worth the effort. It might be relatively simple for you or I, but I am not convinced that everyone else would see it that way (hence the slightly modified system for Pclass saves)
Jason Bulmahn
Lead Designer
Paizo Publishing
I have to agree with you on this one. The current system for saving throws is nice and simple - you simply add up all the bonuses. I have no problem doing fractions if I am a player, though it could be annoying for me as a DM if I try to stat up many multiclassed NPCs, but fractions are something that would doubtlessly bother my players - in fact, I would probably have to calculate their saving throw progressions for them even as a DM.
The fractional systems are elegant, but fractions are also confusing to a considerable number of players. I think it would be better to simply remove the +2 bonus from the good save progression, thus making it start from zero. The extra +2 for good saving throws could instead be made into an enhancement bonus (or resistance bonus, if that one is kept as a separate bonus type) granted by the class, which means no stacking. This would preserve the simplicity - you simply add up all the bonuses from all the tables for all classes to get the resultant base saving throw for the character and to that you add the various other bonuses (for ability scores, enhancement/resistance bonuses [granted by classes, items, spells, etc. but no stacking here], etc.).

![]() |

What about a simple rule that says that if you have 2 or more +0s in your BAB from multiple classes, you add 1 to your total BAB? Similarly, if you have 3 or more +0s in a saving throw from multiple classes, you add 1 to that save? There's probably a more elegant way to word this, but I think it would work.

![]() |

I for one am NOT in favor of fractional saves at all. If we go that route it makes more sense to just have one attack chart and you find your BAB by looking at your highest level class and comparing your total level to the chart.
You want to play a 2nd level rogue with +1 to hit, then play a 2nd level rogue and don't go picking up a new class.
There's nothing wrong with the system as is, unless you just want to munchkin the system for every single possible +1 you can squeeze out which of course eventually breaks the system.
This should be a house rule, not part of the game. Changing this means recalculating the power levels of every single monster in the game, as suddenly level 10 parties become far more powerful compared to the poor weak CR 10 monsters. 3.5 monsters are already too weak compared to Pathfinder classes, why make it worse?

![]() |

What about a simple rule that says that if you have 2 or more +0s in your BAB from multiple classes, you add 1 to your total BAB? Similarly, if you have 3 or more +0s in a saving throw from multiple classes, you add 1 to that save? There's probably a more elegant way to word this, but I think it would work.
+0* in the table, and a note at the bottom saying that 'A player with a multi-class character gets an extra +1 to the save for every two *'s they collect'
I'm also an advanced maths geek and I don't use fractional saves because I know that asking some players to add 1/2 and 3/4 just isn't going to work - there are a lot of people who just don't 'do maths'.
However, an add up the stars and divide by 2 system would probably work.

Lehmuska |

There's nothing wrong with the system as is, unless you just want to munchkin the system for every single possible +1 you can squeeze out which of course eventually breaks the system.
Munchkining, like, limiting the fortitude save of a fighter/barbarian to what a single classed fighter or a barbarian could have? Fractional BAB and saves do not make characters into powerhouses. They make characters have more average numbers instead of extremely low or extremely high.
I'm also an advanced maths geek and I don't use fractional saves because I know that asking some players to add 1/2 and 3/4 just isn't going to work - there are a lot of people who just don't 'do maths'.
We could just call them 3/4 and 2/4 for BAB and 3/6 and 2/6 for saves. Suddenly they're much easier to use even if you're not into math.

seekerofshadowlight |

I'm also an advanced maths geek and I don't use fractional saves because I know that asking some players to add 1/2 and 3/4 just isn't going to work - there are a lot of people who just don't 'do maths'.
Got it in one right there. Not to count the chart there will be and then the double, triple checking and the "well i thought it worked this way...no,no,no you dont add it like that!"
We could just call them 3/4 and 2/4 for BAB and 3/6 and 2/6 for saves. Suddenly they're much easier to use even if you're not into math.
Umm what...huh...where...I have no clue where we are at now.
I called it over complicated up in the thread as I can not do it on my hands. And you all know the chart will cause people to do it wrong and bring much confusion.
An elegant system would be one most gamers, not just the math guys can do fast easy and not mess it up
Fractional is none of that.
It does work, just not elegant or easy .

Jack Townsend |

Is it really a punishment to have a +0+0 BAB progression due to multiclassing in a 1/2 | 3/4 mix? I say no. Because at level 3 and 4 this problem becomes irrelevant since these levels gain a +1+1 to BAB, which is like having become a fighter.
The only real problem occurs, when a character has two 1/2 BAB classes, but that's a corner to be never approached, occuring only when you mix two arcane full casting classes and starting with a PrC at an odd characterlevel. That would be real Min/Minimizing...

![]() |

Thanks for the vote of confidence Hogarth.
Anyway, fractional saves have been something that I have been considering for a while, but I am not 100% sure that the extra math is worth the effort. It might be relatively simple for you or I, but I am not convinced that everyone else would see it that way (hence the slightly modified system for Pclass saves)
Jason Bulmahn
Lead Designer
Paizo Publishing
The problem with saves being a PrC thing is that even if there's a problem to be solved (about which I'm not convinced) doing it for PrCs and not for base classes just seems weird, given that you can multiclass into them as well. To me, it looks like a partial solution to a problem that I'm not even convinced exists, at least not in terms of abuses (ie, save-dipping) although the sort of thing Ross is talking about for BAB is a problem. Also, doesn't doing this for PrCs now break the old fractional save option to some extent (in terms of backward compatibility)?

Dorje Sylas |

Is it really a punishment to have a +0+0 BAB progression due to multiclassing in a 1/2 | 3/4 mix? I say no. Because at level 3 and 4 this problem becomes irrelevant since these levels gain a +1+1 to BAB, which is like having become a fighter.
The only real problem occurs, when a character has two 1/2 BAB classes, but that's a corner to be never approached, occuring only when you mix two arcane full casting classes and starting with a PrC at an odd characterlevel. That would be real Min/Minimizing...
Druid/Rogue, +0 BAB until 3rd level. Very possible class combination. Depending on the level splits you can end up with 20th BAB of +14 or +15. The saves look a little better since they aren't adding on good saves.
Cleric/Wizard going toward Mystic Theurge (without new save progression), again to a corner case and a good example of the save issue on the upper and lower ends. You get a good deal of variance depending on how it is actually played. Assuming an even split (5 Cleric/5 Wizard/Theurge 10) by 20th level the will save will be +15. Which when you think about it is better then spending a feat on Iron Will. Ref ends one point lower at +5 not much of an issue end game, but getting their can be painful. With an even split Ref gets stuck at +0 for the first 4 levels of play. While not a large difference it has the same mental impact as dead level... and we all hate dead levels. Never mind that during those same levels the Will save is 4 levels high then it should.
Both case are that 1/2 + 3/4 or 3/4 + 3/4.

Ross Byers RPG Superstar 2008 Top 32 |

How about this, then:
If forcing players to add 1/2 and 1/4 is too much to ask, how about we just make levels with the same progression stack on the tables, but you still just add the table values?
That way the difference will be at most +1. The Wiz/Rog going arcane trickster might not get fixed, but a Rog/Brd would be.
This would also fix the +2 save bump problem, because if you had (for instance) a Ftr2/Brb3/Clr1, they have six levels of good fort saves, instead of three +2s at the bottom.
Edit: I think Freesword brought this up before I did in this thread...don't want to look like I'm stealing anyone's ideas.

Roman |

The thing is that this solution conflates two 'problems', only one of which I consider to be an actual problem.
I do see the problem with multiclassing leading to very high saving throws through the stacking of the initial +2 bonuses. Making the +2 a class-given resistance bonus (or ehnancement bonus if resistance bonuses are removed as a separate class) would solve this problem.
The fact that stacking of the initial +0 bonuses can lead to low saving throws is not a problem for me - I consider it a feature. It provides some incentive to stick with the original class rather than multiclassing willy-nilly and it does so in a manner without increasing the power of the initial class (and power creep is a major issue for me). I would hate to see this feature disappear.

![]() |

I have maybe a point that will be relvent to this, other than i don't use the multiclassing rules (personal homebrew thing that has no baring on this discussion).
At what grade do children learn to add fractions? This game has to be usable by everyone. That is why we use simple math. younger audiences have to be able to follow the system. This is why i think your fractional idea will not work.

Ross Byers RPG Superstar 2008 Top 32 |

I have maybe a point that will be relvent to this, other than i don't use the multiclassing rules (personal homebrew thing that has no baring on this discussion).
At what grade do children learn to add fractions? This game has to be usable by everyone. That is why we use simple math. younger audiences have to be able to follow the system. This is why i think your fractional idea will not work.
I learned fractions in the 3rd or 4th grade.
And as I already pointed out: Mono-class characters will not be affected by these rules, and I consider multi-classing (or applying class levels to monsters) to be 'advanced' rules. I'm okay with a 10-year old player having a bit of trouble with them.

Quandary |

Ross, you're right: "Like Stacking" involves NO fractions (or decimals),
and simply requires Multi-Classed Characters ONLY to add up their "Good Save" Class Levels before adding up their "Average Save" Class Levels.
(i.e.: "OK, Let's see, now I'm a Fighter 10/Ranger 10. Since they're both have Good Fort and Average Will, those stack, so I look up the 20th level Base Saves for those: +12 and +6. The Ranger has Good Reflex while the Fighter is Average Reflex, so those add up separately: +7 +3 = +10. Now let's kill some of them Fraction Monsters! [using 3.5 rules, the same character would have +14/+10/+6])
You might have missed a post I made realizing that with one small change, "Like Stacking" can 100% match the outcome of Fractional Saves:
If people are worried about those wierd corner cases (Rogue(3/4) 1/Wizard(1/2) 1 = BAB 0), there can even be a line saying that you can treat one level of a higher BAB Class as if it were of a lower BAB progression (if it would be advantageous).
In other words, the 'wasted' 3/4 is treated as a 1/2 so it stacks with another floating 1/2, adding up to 1: The same outcome as Fractional Saves (since adding fractions results in 1-1/4, rounding to 1)
...If this is still too complicated somehow, I think the simplest solution is separating out the "Good Save 1st Level Bonus" from the Save Progression, and making it function like Class Skills or Weapon Proficiencies: You don't gain any further benefit by taking ANOTHER Class with "Good Save X". Someone else mentioned this option also, though I don't think we really need to worry about the "Bonus Type": Class Skills don't seem to bother with that - It's just explicitly defined to not stack. The reason why I don't like this as much as "Like Stacking" is it only adjusts DOWN for multiple High Save Classes, not UP for multiple Poor Save Classes. That said, it's obviously the easiest solution to implement.

Ross Byers RPG Superstar 2008 Top 32 |

You might have missed a post I made realizing that with one small change, "Like Stacking" can 100% match the outcome of Fractional Saves:
Quandary wrote:If people are worried about those wierd corner cases (Rogue(3/4) 1/Wizard(1/2) 1 = BAB 0), there can even be a line saying that you can treat one level of a higher BAB Class as if it were of a lower BAB progression (if it would be advantageous).In other words, the 'wasted' 3/4 is treated as a 1/2 so it stacks with another floating 1/2, adding up to 1: The same outcome as Fractional Saves (since adding fractions results in 1-1/4, rounding to 1)
I didn't miss it, I just think that this is a case where the cure is worse than the disease: If we are worried about people not comprhending fractions, I think that your fix is excessively wordy and confusing. I'm willing to lose +1 once in a while, since with simple 'like stacking', it never gets worse than that.

Darwin |
...If this is still too complicated somehow, I think the simplest solution is separating out the "Good Save 1st Level Bonus" from the Save Progression, and making it function like Class Skills or Weapon Proficiencies: You don't gain any further benefit by taking ANOTHER Class with "Good Save X". Someone else mentioned this option also, though I don't think we really need to worry about the "Bonus Type": Class Skills don't seem to bother with that - It's just explicitly defined to not stack. The reason why I don't like this as much as "Like Stacking" is it only adjusts DOWN for multiple High Save Classes, not UP for multiple Poor Save Classes. That said, it's obviously the easiest solution to implement.
Simplest (for those who can't understand fractions) might be to just change the way saving throw totals are listed. instead of:
level fort reflex will
1 +2 0 +2
2 +3 0 +3
3 +3 +1 +3
4 +4 +1 +4
5 +4 +1 +4
6 +5 +2 +2
it could be listed like this:
Upon gaining a level in cleric, the saving throws increase by the below amount:
level fort reflex will
1 +2 0 +2
2 +1 0 +1
3 0 +1 0
4 +1 0 +1
5 0 0 0
6 +1 +1 +1

Mattastrophic |

To me, it looks like a partial solution to a problem that I'm not even convinced exists, at least not in terms of abuses (ie, save-dipping)...
Oh, it exists. I show why fractional save-and-BAB-stacking is a necessary change to the game right here.
-Matt

![]() |

Bagpuss wrote:To me, it looks like a partial solution to a problem that I'm not even convinced exists, at least not in terms of abuses (ie, save-dipping)...Oh, it exists. I show why fractional save-and-BAB-stacking is a necessary change to the game right here.
-Matt
My concern is about whether the problem actually happens in the wild and the extent to which it causes problems in the playing of the game.
I don't have a problem with fractional save or level stacking (although the latter looks a little more elegant in practice to me) and I'd include them as game options in the PFRPG rulesbook at the least. However, I think that what Jason's done with PrCs saves is dodgy because it's only for PrCs and breaks backwards compatibility whilst introducing new inhomogeneity. Also, I've never seen a save-dipped combo in the wild that was a problem, thus my comment about not being sure that it's actually a real problem that required changing PrC save progressions.

Epic Meepo RPG Superstar 2009 Top 16, 2012 Top 32 |
A fractional solution without fractions:
Don't list class BAB and saves at all. List ranks in Combat, Fort, Ref, and Will. For every 3 ranks you have in Combat, your BAB increases by +1. For every 3 ranks you have in a saving throw, your save bonus increases by +1. Determine class progressions for ranks in Combat and saves accordingly.

Ross Byers RPG Superstar 2008 Top 32 |

Simplest (for those who can't understand fractions) might be to just change the way saving throw totals are listed. instead of:
level fort reflex will
1 +2 0 +2
2 +3 0 +3
3 +3 +1 +3
4 +4 +1 +4
5 +4 +1 +4
6 +5 +2 +2it could be listed like this:
Upon gaining a level in cleric, the saving throws increase by the below amount:
level fort reflex will
1 +2 0 +2
2 +1 0 +1
3 0 +1 0
4 +1 0 +1
5 0 0 0
6 +1 +1 +1
This would make sense for leveling up a character from 1-20 organically. However, it is a really terrible idea for creating higher-level NPCs or replacement PCs, or even being able ot check the math on that organic character.