[Fixing 5E] Saving Throws - Fortitude Reflex Will

[Jester David]

Sorry - you and I have very different expectations on the system in that case.

This is specifically and explicitly about how the system breaks down for what you call those "two, maybe three" times. That the campaign reaches above level 15 is a given.

So please stop dismissing the issue just because you don't care high level play is broken. Thank you.

It's about fixing the right problem, not making changes to all levels of a campaign to fix a problem that only occurs at ultra-high level shouldn't be done lightly. Not working to fix a symptom of a larger problem rather than fixing the underlying cause,

First, how many times has this actually happened in your campaign? Is this an actual problem at your table or a theoretical problem based on reading the rules?
Second, what about monsters facing the PCs? Many have minuses to stats, so that DC 19 at level 17+ becomes as impossible as the DC 21.

Again, if the problem is just that monsters who have saves of DC 21+ are un-fun for the player, then cap saves at DC 20. Just have a hard cap. Or 19. That fixes the problem without requiring changes how the other 98.6% of saves rolled over the campaign work. Without increasing success across the board.

Or, as I pointed out but you ignored, you could address the root problem: saves become harder and harder for most PCs, with PCs proficient in saves simply retaining their odds of success. However, the solution to this problem is to reduce PC spellcaster accuracy - which is problematic - or calculate monster saving throws differently than player saving throws.

 

[dave2008]

I tried that system and it's horribly unbalanced. Fort spells will never succeed, reflex spells will succeed nearly twice as much - especially as levels go up and will is in the middle.
You'd have to rebalance the entire MM which I also considered for different combinations, but definitely not worth it.

EDIT: Ran the numbers again. Across all levels likelihood that an enemy would save: 52% fort, 39% reflex, 44% Will. At 17-20 we're looking at 66% fort, 45% reflex, 56% will. Unacceptable imo.

My results: 47,46,46 overall. 59,54,58 at 17-20.



Initially I was opposed to the idea of Str + Dex = Reflex, but the two are used for quite a few things in D&D so it makes sense - escape grapple, break out of restrain, etc (I had a list at one point). I tried many different combinations and in the end this one was the best mathematically and pretty darn good thematically (arguable based on opinions).


I heavily considered that system, and I really liked that armor became DR, but we decided against it. It could simplify some things.



Anything combined with Con will be far more likely to succeed than the rest because it's already quite high on most monsters.


I do exactly that - I have a list for them in the homebrew. It's basically the same except for Bard who could potentially choose Reflex or Will.


Definitely. Current Intelligence saves are definitely Will saves no matter how you combine the ability scores.


I considered this option - it was my original fix, but it doesn't solve the issues at all. As the game advances in levels Constitution save spells are still significantly less likely to succeed and Intelligence spells succeed 70% of the time on average. It's rather ridiculously unbalanced. Giving half prof to all doesn't solve this unbalance.

I think you have sold me! Thank you!

 

[CapnZapp]

It's about fixing the right problem, not making changes to all levels of a campaign to fix a problem that only occurs at ultra-high level shouldn't be done lightly. Not working to fix a symptom of a larger problem rather than fixing the underlying cause,

First, how many times has this actually happened in your campaign? Is this an actual problem at your table or a theoretical problem based on reading the rules?
Second, what about monsters facing the PCs? Many have minuses to stats, so that DC 19 at level 17+ becomes as impossible as the DC 21.

Again, if the problem is just that monsters who have saves of DC 21+ are un-fun for the player, then cap saves at DC 20. Just have a hard cap. Or 19. That fixes the problem without requiring changes how the other 98.6% of saves rolled over the campaign work. Without increasing success across the board.

Or, as I pointed out but you ignored, you could address the root problem: saves become harder and harder for most PCs, with PCs proficient in saves simply retaining their odds of success. However, the solution to this problem is to reduce PC spellcaster accuracy - which is problematic - or calculate monster saving throws differently than player saving throws.

Thank you for at least/at last moving past the point where we have to discuss this is a problem. Your suggestion, capping saves at DC 20, is indeed a simple one.

On ther other hand, I would be grateful if you stop questioning the premise for the discussion. Don't call it "ultra-high level". It's dismissive. If not outright false. You know I'm talking about essentially every double-digit level. At least one third of the game. Then: asking me how many times this has actually happened is first of all irrelevant. Secondly, it is insulting, since it insinuates that unless I meet some arbitrary quota of yours, the problem isn't large enough or even real. So, I'm asking you as respectfully as I can: stay the bleep out of the thread if all you're here for is to question its premise, m'kay?

.
.
.

Okay. That's better. Now then. To your questions.

Q. Is this an actual problem at your table?
A. Yes

Q. Second, what about monsters facing the PCs?
A. What about them? I'm not talking about that.

Q. if the problem is just that monsters who have saves of DC 21+ are un-fun for the player, then cap saves at DC 20
A. Well, that's a simplistic description. The real problem is that DCs can and will be roughly twenty points higher than save bonuses. The real problem is that good game design would have made sure the target number you need to roll generally does not go outside something like the 4-16 range. It's strange to be asking "was bounded accuracy completely and utterly forgotten by the saving throw designer!?"

Q. the solution to this problem is to reduce PC spellcaster accuracy - which is problematic - or calculate monster saving throws differently than player saving throws
A. Now you have lost me. What problem? I'm not talking about PC spellcasters. I'm not concerned about monster saves. Again, this is not about monsters facing the PCs.

Q. the root problem: saves become harder and harder for most PCs
A. Well, I would more call this a symptom of the root problem, rather than the root problem in itself. But yeah, we seem to be in agreement here.

What I would call the root problem is how the designers did not stop saves from going from "very difficult" til "outright impossible". The root problem is how you are given six saves, but not any reasonable path to maintaining them. I'm quote okay with heroes having weak spots and achilles heels.

But the save subsystem of 5E is more like achilles legs, or an achilles torso.

---

It is quite unacceptable to ask a player to either resign herself to having three out of six defenses where you basically face a zero percent success rate. Or, to completely divert any attempt at being actually best at anything just to patch those holes... and still have holes gaping wide.

A save system where it's essentially up to pure chance if the monster decides to attack you where you can't defend yourself at all is a broken system. It's like you roll the dice, roll tails you keel over and die with nothing you can do about it, roll heads, you get your shot but still face a sizeable risk of failing.

There simply is nothing heroic with such pathetic defenses. It suits D&D high level play extremely badly.

---

Finally, what do you think about the F/R/W suggestion?

 

[CapnZapp]

I spent a lot of time on this area and have a good understanding of all the variables involved so let me know if you have any questions. Also feel free to PM me if you'd rather discuss over chat.

I prefer open discussion if it's all the same to you.

Also, let me summarize your proposal for clarity brevity and for those who dislike links to outside sites.

Saving throws condensed to Fortitude, Reflex, and Will.

Fortitude uses your Constitution modifier (and replace Con saves)

Reflex uses the average of your Strength and Dexterity modifiers (and replace Str and Dex saves)

Will uses the average of the highest two modifiers from your Intelligence, Wisdom, and Charisma modifiers (and replace Int, Wis and Cha saves).

Each class only gain one save proficiency:

Barbarian: Fortitude
Bard: Reflex or Will
Cleric: Will
Druid: Will
Fighter: Fortitude
Monk: Reflex
Paladin: Will
Ranger: Reflex
Rogue: Reflex
Sorcerer: Fortitude
Warlock: Will
Wizard: Will

First thing on the agenda: is this a correct representation?

 

[CapnZapp]

Assuming it is, a few thoughts:

First I consider the way Fort saves aren't based on averages a potential weakness, but when I realized you are probably counting on the fact that Con is the least likely ability to be "average", I see that might not be an issue after all. Nice.

Then, a question. Am I right in guessing you're not actually planning to change how a feat such as Resilient works? That is, you still need to pick an actual ability for the feat, rather than directly picking something like Reflex or Will?

One concern however, and this is more like a review of your document than anything else. You present a lot of numbers. But I'm afraid that assumes an equal distribution of monsters, which is highly unlikely in practice. The overall idea is still sound, but to present numbers with your degree of accuracy isn't really appropriate here.

Intelligence saves might account for 2% of the saves in the Monster Manual, but after meeting three NPC wizards all casting Feeblemind, that number loses all relevance.

I'm not accusing you of doing anything wrong. I'm just asking you to consider that equality on paper might not be such a valuable measure to strive for. And perhaps to replace percentage numbers (that suggest accuracy) with less defined terms.

For example:
Strength saves: Common overall, common damage, common hard control, and pervasive light control.
Intelligence saves: Very rare overall, pervasive damage, pervasive hard control, common light control.

By avoiding putting numbers on it, you avoid implicitly suggesting the numbers mean something, that they're scientifically calculated. When in reality they might mean nothing at all, and will probably vary wildly from one campaign to the next. After all, you are only describing trends, and vague likelyhoods.

 

[Kryx]

Also, let me summarize your proposal for clarity brevity and for those who dislike links to outside sites.

I purposefully have not done so because the proposal is based on math and that math is contained in that document or in spreadsheets.

Saving throws condensed to Fortitude, Reflex, and Will.

Fortitude uses your Constitution modifier (and replace Con saves)

Reflex uses the average of your Strength and Dexterity modifiers (and replace Str and Dex saves)

Will uses the average of the highest two modifiers from your Intelligence, Wisdom, and Charisma modifiers (and replace Int, Wis and Cha saves).

Small correction: Take the average score of str/dex and then get the modifier. So 17 + 11 would be 28 - 20 = 8 / 2 / 2 = +2. Whereas 3+0/2 = +1.

First thing on the agenda: is this a correct representation?

Other than the average part, yes. I have the ability score part in a note on the page.

First I consider the way Fort saves aren't based on averages a potential weakness, but when I realized you are probably counting on the fact that Con is the least likely ability to be "average", I see that might not be an issue after all. Nice.

I'm not sure I understand your meaning. Con saves in the normal system (and therefore in this system) are the most consistent among monsters. Monsters have more and more con as they level, therefore more and more con save modifier or fort modifier.

Then, a question. Am I right in guessing you're not actually planning to change how a feat such as Resilient works? That is, you still need to pick an actual ability for the feat, rather than directly picking something like Reflex or Will?

Strength Saving throws do not exist in my game. Reflex saving throws do. Resilient's power is largely unchanged. Previously the best choice for casters was Con - that value is unchanged. Some people may have taken Wisdom in the normal rules, but that is rather rate. Even if they do take Resilient (Will) the power has barely nudged. They now have proficiency in 25% of all saves instead of 19%. Still less than 35% of all saves on Fort. Reflex is definitely more powerful than before, but is inline with Fortitude and Will. Besides Resilient would take 1/5 of the player's ASIs. Assuming 2 ASIs are on ability score improvements that's 1/3 of the player's free choice on a defensive option. It's a good option (especially for casters), but isn't too powerful imo.

One concern however, and this is more like a review of your document than anything else. You present a lot of numbers. But I'm afraid that assumes an equal distribution of monsters, which is highly unlikely in practice. The overall idea is still sound, but to present numbers with your degree of accuracy isn't really appropriate here.

My DPR spreadsheet and this spreadsheet are attempts to understand the balance of the game as it stands. To do that I must use a distribution of monsters from the game. The game, not your game or my game. If you instead wanted to see how much CR 6 creatures would save I could do that, but that doesn't give us the high level view required to understand the math behind the game's system.

Intelligence saves might account for 2% of the saves in the Monster Manual, but after meeting three NPC wizards all casting Feeblemind, that number loses all relevance.

Exactly a problem I'm attempting to fix. Niche saves that are incredibly likely to succeed and incredibly deadly. I'm not sure what your point is here.

I'm just asking you to consider that equality on paper might not be such a valuable measure to strive for.

I think equality on paper is exactly what I'm striving for. Equality on paper attempts to transition to equality in game. Notice, however, that Will is still much less common than Reflex. That is because the results of a Will save are often much more bothersome than the typical damage from Reflex. Equality doesn't mean exact numbers - it means numbers that have equal value. In my opinion the current numbers have the most equal value of any of the combinations I tried and far more equal value than the RAW rules.
I attempted to isolate the severity of the saving throw via my Damage, Hard CC, Light CC system. If you think that can be improved to model the game's system better please do make suggestions.

And perhaps to replace percentage numbers (that suggest accuracy) with less defined terms.
By avoiding putting numbers on it, you avoid implicitly suggesting the numbers mean something

Numbers are what matters. Words mean different things to different people. Numbers are true values unmanipulated by my words which inherently have bias not only in the choice of words, but also in the breakpoints I choose to set.
I want to deal in numbers.

that they're scientifically calculated. When in reality they might mean nothing at all, and will probably vary wildly from one campaign to the next. After all, you are only describing trends, and vague likelyhoods.

They are calculated from the whole of the monster manual and definitely do mean something. They tell us how the core game is balanced. Once again I'm not balancing for your game or my game. Those games will obviously only use a subset of the overall distribution and that's fine. But assuming the GM doesn't purposefully select for certain aspects (higher AC, higher Str, etc) then an average assortment should reflect this model pretty accurately.

 

[CapnZapp]

They are calculated from the whole of the monster manual and definitely do mean something. They tell us how the core game is balanced. Once again I'm not balancing for your game or my game. Those games will obviously only use a subset of the overall distribution and that's fine. But assuming the GM doesn't purposefully select for certain aspects (higher AC, higher Str, etc) then an average assortment should reflect this model pretty accurately.

Well... if hypothetically the only monster in the entire MM that used attacks requiring Int saves was Orcs, your number would be "less than 1%" but in reality, probably closer to 75%. That is, if we can assume three out of four campaigns feature Orcs at one time or another. Do you see my point?

Besides, I don't think equality among saves is necessarily a priority or that it automatically leads to a better game. Again hypothetically, if a common save was deadlier than an uncommon save, it wouldn't necessarily be a problem.

Of course, I agree when you consider Int saves to be a problem, since they are very rare yet so very dangerous. It sets players up for a most unfair choice - they can't be expected to focus on such a rare save; yet the consequences of not doing so can be devastating.

Anyway. Thanks for your ideas.

I will consider your coupling of the six saves (that is, Con becomes Fort all by itself, Strength and Dexterity becomes Reflex, and everything else becomes Will).

I will probably stick to my simple "pick your highest score" rather than average. It's only fair to let you ask yhy, when I heard you say you want dump stats to retain a meaning?

Well, because for Fortitude, it doesn't matter.

And for Reflex, well, almost no character picks both Strength and Dexterity. You go either Strength and heavy armor, or you go Dexterity and light armor. So it seems overly harsh to ask the martial characters to focus on both, just for the sake of getting their one save up.

A bit of the same with Will. Few characters focus on more than the one mental stat they use for their spellcasting.

 

[CapnZapp]

(cont'd)

As for Resilient, my question was that if you take the average of, say, Wisdom and Charisma to compute a character's Will save, it becomes important to know if the feat applies to abilities still, or to your new save categories.

That is, do you pick Resilient (Wisdom) or do you pick Resilient (Will)?

In the first case, the actual Will save is only impacted half as much, since the final Will score does not only depend on Wisdom, but another value too.

 

[Kryx]

Well... if hypothetically the only monster in the entire MM that used attacks requiring Int saves was Orcs, your number would be "less than 1%" but in reality, probably closer to 75%. That is, if we can assume three out of four campaigns feature Orcs at one time or another. Do you see my point?

I'm interested in the balance of the game, not the balance of any specific campaign that uses the game. As I said above I expect a campaign that pulls from the system without bias will generally reflect the balance of the system.

I don't think equality among saves is necessarily a priority or that it automatically leads to a better game.

I do. I had quite negative experiences throwing mind flayers at my players who had all justifiably dumped their Intelligence score as it has very little value in 5e. I had negative experiences with players prioritizing int, cha, or dex save spells over any of the others as those saves were almost always the weakest on monsters.

I am a firm believer that better balance leads to a better game. This balance leads to more options for players which is one of my main goals.

Of course, I agree when you consider Int saves to be a problem, since they are very rare yet so very dangerous. It sets players up for a most unfair choice - they can't be expected to focus on such a rare save; yet the consequences of not doing so can be devastating.

Exactly.

I will probably stick to my simple "pick your highest score" rather than average. It's only fair to let you ask yhy, when I heard you say you want dump stats to retain a meaning?

That will result in monsters who are more likely to succeed than RAW success levels and dump stats for players, but that's your choice.



I'm glad my ideas and work on this area could be an interesting thing for you to read and consider. Happy rolling.

 

[Jester David]

On ther other hand, I would be grateful if you stop questioning the premise for the discussion. Don't call it "ultra-high level". It's dismissive. If not outright false. You know I'm talking about essentially every double-digit level. At least one third of the game.

The focus on level is very relevant form a balance & design perspective. If the problem affects 1/3rd of the game (I would argue it's probably closer 1/4th or 1/5th, but whatever), then the solution should be focused on those levels, and need not focus as much on low levels. If 5% of saving throws are an issue, the solution shouldn't equally or noticeably impact the other 95% of saving throws.

Finally, what do you think about the F/R/W suggestion?

I think it's flawed. And I don't think it fixes the problem.

With two proficient saves, a character is proficient in 2/6ths of saving throws. A character proficient in one of three saves has similar odds of success on any given saving throw. Reducing the number of saves doesn't change that ratio. Yes, it means you have one save you're really good at, one you're so-so, and one you're poor with (rather than one good, two to three so-so, and likely a couple poor). But you still have the saving throw where you have a low bonus.
Similarly, increasing the odds of a character having a good ability score to add to their save doesn't significantly change the chances of succeeding with a non-proficient saving throw, especially at high levels. At high levels you might still have a high disparity between DCs and save bonuses; the change might only be a +1 or +2 bonus at most. So instead of facing the DC 21 saving throw with a +1 they're facing it with a +3. Which isn't all that much better. So it doesn't entirely resolve your problem.


However, at low levels - when things were arguably balanced - you still have that +1 bonus to your worst save, and your middle saves are much higher. So it makes saves easier across the board. Even using your "1/3rd of the game figure" it has an unwanted impact on 2/3rds of the game.
So just in a design vacuum it seems like poor design.

However, games don't work in a vacuum and interact with people. This might leads to the problem that occurred in 4e where characters have three dump stats with minimal game impact. This promotes and rewards min-maxing: you're encouraged to have Con & Dex since those are the stronger of the two paired stats. And once you have one character with a high Cha, the rest of the party is better off with Wis.
This is unless your class hits off a weaker stat. But those classes become slightly inoptimal and more MAD. Classes like the Strength based fighter or the wizard become subtly weaker, since they have a stat overlap.

And it also makes many monsters tougher. For example, dragons. You used to target a dragon's Dexterity saves because they were the weakest. Now, since it's Reflex (and Intelligence) the dragon becomes dodgy. They're the most egregious, but any monster that has one or two dump stats becomes slightly more likely to save. And since - unlike PCs - monsters aren't build via a point buy, they'e more likely to benefit.

Q. Second, what about monsters facing the PCs?
A. What about them? I'm not talking about that.

It's no more fun for the DM to face an impossible saving throw. To have their cool monster killed with no chance to succeed.
And it's less fun for everyone else at the table who might not get a chance to act because their initiative was lower than the spellcaster's. Yeah, most level 15 PCs are just going to fail against the CR 17 goristro's DC 21 Strength saving throw (the lowest CR creature I could find with a DC above 20). But with a -2 it's never going to make a level 17 PC's DC 19 Intelligence saving throw. Ever. And with a +0 Dex, +1 Wis, and +2 Cha it's not going to make many other saves either, even with it's Magic Resistance. Feeblemind is an auto-win.

Q. the solution to this problem is to reduce PC spellcaster accuracy - which is problematic - or calculate monster saving throws differently than player saving throws
A. Now you have lost me. What problem? I'm not talking about PC spellcasters. I'm not concerned about monster saves. Again, this is not about monsters facing the PCs.

It's the same problem. Both sides face an identical issue: the increasing disparity between save DCs and save bonuses. Because saving throws don't following the standard rules of DCs (aka bounced accuracy).
Which is tied to spellcasters having DCs that increase, impacting their "accuracy" in combat.
That's the root cause of the problem. If you want to fix the problem go after the actual cause. Perhaps change how saves are calculated, possibly just for monsters, or possibly for all characters.

Then: asking me how many times this has actually happened is first of all irrelevant. Secondly, it is insulting, since it insinuates that unless I meet some arbitrary quota of yours, the problem isn't large enough or even real.

There are a lot of "theoretical" problems posted on these boards. Things that are an issue on paper, upon reading the PHB, but may not a serious problem at the table.
I don't think it's unfair to ask if something is a problem that has never occurred but *might*, a problem that has occurred once, or a problem that has occurred repeatedly.