average pc save bonuses by level.

[bjoern]

I'm wondering about what save bonuses are "par for the course" for each PC level?

For example , is a 5th level PC with an average save bonus of +4 below, at or above average for his level?

[AvatarVecna]

There's just too many variables to give you an accurate "average" estimate. Saves are based on base attributes and class levels, sure. But what if the stats rolled/purchased are particularly high/low? What if the character multiclasses using standard rules? What if they multiclass using fractional save bonus calculation? What if they're a gestalt character?

Ultimately, though, the biggest variable is optimization level; on the low end, we might be looking at a character who has a bad save and with their relevant stat sucking as well, for a total lvl 5 save bonus of -4; on the high end, Pun-Pun has infinite everything, and so has a save bonus of +infinite.

[Zaq]

It really varies. First, there's the obvious factor of different classes having different base save bonuses. Then you consider that key stats figure into it pretty heavily—a Cleric and a Sorcerer both have high base Will saves, for example, but the Cleric is likely to have a much higher one than the Sorcerer is, because the Cleric is almost guaranteed to have a much higher WIS than the Sorcerer is. Then there's the fact that multiclassing makes your saves better, so builds that dip around for a lot of frontloaded classes (which means mostly mundane/melee people) will have higher base saves than builds who single-class it. (PrCs are just as important as base classes in this regard, so a Wizard who dips into Master Specialist and Mindbender will have a higher Will save than an Artificer who doesn't PrC out, even if they have the same WIS.) Then you get all the little bonuses here and there—Cloaks of Resistance are universal enough to not make a huge difference from PC to PC (though classes/builds with less reliance on magic items can afford better ones faster), but there's racial bonuses (dwarves get a +2 against spells, for example, and halflings just get a +1 on everything), class-related bonuses (Divine Grace being the most obvious example), and so on.

With so many different factors in play, I feel like calculating an "average" is going to be a fool's errand, and it isn't likely to actually be that useful once you calculate it. You could start with four values per level (high save on stat you care about, high save on stat you don't care about, low save on stat you care about, low save on stat you don't care about), but that would ignore items, multiclassing, racial bonuses, and so on. By the time you factor everything in, you'd have such a jumble of factors that the numbers you get for your averages wouldn't really mean much.

[bjoern]

OK, let me ask the same question from the other direction.
What is the average save DC by CR ?


What I'm going for here is to figure out what save bonus is adequate for each level to give a good chance of making a saving throw vs a CR appropriate save DC. Probably making the save on 3 out of 4 throws or better I suppose.

[HolyCouncilMagi]

"Average" and "par for the course" are two quite different things.

I have no idea what the average is, but...

Generally, at level 1, it'd be good to have +2-3 to all Saves, if you can manage it. This, like all things, is easier for wizards/sorcerers; Dex 14 and Con 14 are pretty standard as far as stats go, and they get a good Will save. Sadly, this is annoying to pull off for characters whose stats don't line up so nicely. Melee classes (the ones with good Fort saves) also tend to want high Con anyway for the HP, and Wisdom and/or Dex end up in the dumping ground. Anything non-negative will do in a pinch.

Around level 5, you should be able to get some Resistance in item form to patch up your defenses. +5 or more in non-favored Saves will serve you well at this point.

Getting closer to level 10, items and other resources should be keeping you up to par if your build isn't doing that by itself (for example, with Paladin levels). Anybody who's trying to keep themselves in one piece during level-appropriate boss fights should be sporting a +15 or so across the board.

15, you're hitting the pinnacle of things you can add to your Saves, so if you're well-constructed, a little under 25 is a nice benchmark.

DCs don't go up much between CR 16+ enemies and there aren't many new options to explode your Saves, so you can mostly let your classes run their course.

This is assuming a low-op game with reasonable tactics (though not necessarily reasonable builds) among your enemies. The above will generally get you through the game in terms of Saves if the DCs are similar to what appropriately-CRed monsters' abilities have. Obviously, it's not quite so nice if the optimization levels increase in regards to generating enemy Save DCs.

[AvatarVecna]

OK, let me ask the same question from the other direction.
What is the average save DC by CR ?


What I'm going for here is to figure out what save bonus is adequate for each level to give a good chance of making a saving throw vs a CR appropriate save DC. Probably making the save on 3 out of 4 throws or better I suppose.

This will run into a similar problem. All character builds have a CR attached to them; since such builds can be optimized for save DC, it can still vary rather wildly, going right back up to DC Infinite with Pun-Pun's spells. Even if you don't do that, and you're only using monsters that already exist, you still run into the problems of how many monsters that is; combined with the fact that lots of monsters have spell-like abilities and/or actual spellcasting that they could optimize, there's just too much material to sift through. And all of that is before monster advancement and template-stacking are even considered.

Hell, I recently started trying to figure out how important each save is in 5e based on how often a spell/monster calls for such a save and how bad failing the save would be. I start 4 days ago, have been working off and on, and I'm still nowhere close to finishing the basic documentation (not to mention other things people have asked me to take into account). And 5e's still on Core only; 3.5 have dozens of supplements on its own, not to mention all the 3.0 stuff that didn't get replaced by a 3.5 version.

EDIT: Sigh...

If I were DMing, I'd probably say that a character's overall average save bonus should be roughly equal to their level in a low-op game, and higher in higher-op games (depending on exactly how "high-op" those games are); for reasonably high-op games, I'd probably expect a character's overall average save bonus to be somewhere between "lvl*1.5" and "lvl*2.5".

[Malcador]

Using the approximately 4500 monsters in ezkajii's recently-released Monster Compendium (http://www.giantitp.com/forums/showthread.php?402179-The-Monster-Compendium), I calculated 10 + 1/2 HD + highest ability modifier for each monster (because that's the formula used for most monster ability DCs, and it also provides a good upper bound for spell save DCs), then averaged for each CR. For CRs between 1 and 20, the results look fairly linear; the average is approximately 12 + (6/5)*CR. Note that this estimate does not account for any of the following:

You will encounter some monsters more frequently than others.
Not every monster determines save DCs using its highest ability modifier.
Not all monsters have abilities with save DCs.


If your save bonus is X, then you can calculate P(1d20+X >= 12+(6/5)*CR) = (8 - (6/5)*CR + X)/20. So if you want a 75% success chance against saves with DC 12+(6/5)*CR, then your save bonus needs to be at least 7+(6/5)*CR. Note the constant term; nobody has +8 to a save at level 1, but as you level up this will become more attainable.

If you follow HolyCouncilMagi's suggestions, these data suggest that you will start with about a 50% chance to succeed, increasing to 75% by level 15. AvatarVecna's suggestion (between 1.5*level and 2.5*level) does about the same thing.

(Somebody please check my math; I'm a logician with a spreadsheet, not a trained statistician.)

[Seerow]

If I were DMing, I'd probably say that a character's overall average save bonus should be roughly equal to their level in a low-op game, and higher in higher-op games (depending on exactly how "high-op" those games are); for reasonably high-op games, I'd probably expect a character's overall average save bonus to be somewhere between "lvl*1.5" and "lvl*2.5".

That seems to be on the high end to me. Even lvl*1.5 means by level 20 you are expected to have a +30 saving throw to every save. Even a character with a good save and a +5 cloak is pulling a +17 at that level, with somewhere between a +2 and a +12 stat bonus. So the +30 simply isn't happening, and +19 is possible for a character with a good save.



Of course with significant multiclassing you can rack up quite good saves. But even then I would consider level * 1.5 to be a high save bonus, and a low save bonus can easily be .5*level.


Edit: Malcador: I think the biggest issue with your math is that most monster save DCs aren't based on their highest stat. DCs tend to be based on con, or a mental stat. Think how many brute monsters are out there with 25-30 strength and then a poison or some other effect keying off of a 20 con, or an SLA keying off a 12 cha, or something like that. For your analysis you count the ridiculously high strength score most monsters have as their ability mod for save DCs, which results in dramatically inflated numbers.

[AvatarVecna]

That seems to be on the high end to me. Even lvl*1.5 means by level 20 you are expected to have a +30 saving throw to every save. Even a character with a good save and a +5 cloak is pulling a +17 at that level, with somewhere between a +2 and a +12 stat bonus. So the +30 simply isn't happening, and +19 is possible for a character with a good save.

Of course with significant multiclassing you can rack up quite good saves. But even then I would consider level * 1.5 to be a high save bonus, and a low save bonus can easily be .5*level.


My assumption for low-op is "Average Save=ECL"; my assumption for high-op is "Average Save>(ECL*1.5)". Your example (a build with a +5 cloak and 20 levels in a class with all good saves) is not highly optimized, unless a monk has become a lot more powerful than this forum would suggest. Beyond that, good saves are assumed to be higher that bad saves; this is just what the average should be. The only other thing I'll mention is that this was my quick-and-dirty general estimate. It will, of course, be wrong for any number of reasons.

I'll be back later with a well-optimized character that hits a rather high average.

[Seerow]

My assumption for low-op is "Average Save=ECL"; my assumption for high-op is "Average Save>(ECL*1.5)". Your example (a build with a +5 cloak and 20 levels in a class with all good saves) is not highly optimized, unless a monk has become a lot more powerful than this forum would suggest. Beyond that, good saves are assumed to be higher that bad saves; this is just what the average should be. The only other thing I'll mention is that this was my quick-and-dirty general estimate. It will, of course, be wrong for any number of reasons.

I'll be back later with a well-optimized character that hits a rather high average.

I wasn't proposing a character with a good save and a cloak was high op. However the question at hand is what is the average saving throw of a player character? You cannot assume a character taking a half dozen classes providing good saving throws as your baseline assumption. To get a good baseline, you look at the simplest scenario. And that is a single class character with standard gear. Especially since you have to consider that is almost certainly what the game is actually balanced around. Because let's face it, as much as we all love our crazy character builds, nobody expects that WotC playtested a character with more than 3 classes, and the vast majority of characters in their testing were almost certainly single classed player characters. (Incidentally this is a large part of why the game falls apart at high levels, but what can you do?)

And in that scenario, a character with a good save is having trouble cracking +20, much less +30. A character with a bad save is actually potentially in the abysmal +11-12 range. A character will on average have two good saves and one bad save, so I'd actually expect the overall average for a 20th level character to be somewhere between a +17 and +20, depending on how the good saves line up with their prioritized stats. Fighters are going to average lower. Monks are going to average higher. I find it really hard to imagine a table where every character in the party has an average for their saving throws of +30.


Just to take a sampling of my last few characters I've used in actual games, I've got:
-Straight Mystic Ranger. At level 9 he has +10/+9/+7 (average: +8.6). By level 20 assuming no free wishes he'd have +21/+20/+15 (average: 18.6)

-Cleric4/Ordained Champion3/Prestige Paladin 2/Bone Knight1. At level 10 he has +18/+9/+18 (average: +15). By level 20 he'd finish off Ordained Champion and have 9 levels of Bone Knight, giving him at 20 +36/+20/+36 (average: 30.66).

-Sorcerer4/Cleric1/Runesmith2/Swiftblade5. At level 12 this character has +14/+10/+16 (average: 13.33). By level 20 he will have finished up Swiftblade and taken 4 levels of Abjurant Champion giving him +21/+17/+22 (average:20)


All three of these are real characters I've played in campaigns fairly recently. The Ranger ends up a bit better all around because at higher level he has reason to want to invest into a +6 item for Dex, Con, and Wisdom, even though they aren't primary. Even with that edge, he's averaging slightly below +1 per level, and is well below the benchmark of 1.5x level that is supposed to be the average.

The Cleric/Paladin has great all around saves, with multiple classes granting good fort and will, as well as using Serenity to get Wisdom based Divine Grace and used as a casting stat, making Wis his favorite stat, and boosting his saves very high. Yet despite what he has going in his favor here, he barely scratches the surface of what you are calling acceptable (1.5x level), and is nowhere near qualifying for the 2.5x level required to be at the high end of average.

The Sorcerer gish has 5 different classes with varying good saves, providing a nice even spread of saving throws. He has no reason to boost Dex or Wis, so doesn't waste the money on those items, and in the end despite the advantages gained from multiclassing averages roughly +1 per level.




But who knows? Maybe I'm just playing in a way low op environment, and am awful at building characters. What characters are you people playing that end with +50 to all saves as a side effect to whatever you were actually trying to achieve? I am seriously curious here, because IME saves at the level the Paladin I showed has are way above average. +1/level seems to fit the mold pretty well for most of my characters, but looking through dropbox characters that other players in my group play seem to average closer to .75/level.

[HolyCouncilMagi]

Well, I don't know about other posters, Seerow, but my own response was just based on my experience with level-appropriate encounters in games I've been a part of. I don't really play the way of minimization, and I don't think most people actually play that way either, though obviously if you're trying to talk about high-op it's the goal. (For the uninitiated, minimization play refers to things like trying to keep your AC at least 19+highest attack bonus of enemies and having unbreachable saves, in order to minimize risk to the character; in other words, defensive optimization, with the "minimization" referring to minimal risk.) I have a lot more fun in my games if my characters have a reasonable but not severe chance of failing their Saves against big guys every so often, but a lot of that has to do that my D&D games are very... Um... The way WotC would have envisioned them.

As such, when I give baselines for where I like to be, it's based on giving myself a major but not insurmountable chance to make Saves against the higher ends of correctly-CRed enemy DCs. As an example, I come up with the "around 25" range for bonuses to Saves by level 15 because at that point, the things you need to be "up to par" against are, for example, the Slow effect of a Greater Stone Golem. With a DC of 31, +25 gives you success a little under 75% of the time, which is about where I enjoy it and about the practical threshold of where you should bother aiming with most DMs (because, y'know, if the party is impervious to the monsters as listed, the natural response on the DM's part will usually just be to crank up the DCs more).

[AvatarVecna]

Even with that edge, he's averaging slightly below +1 per level, and is well below the benchmark of 1.5x level that is supposed to be the average.

I think I've figured out what's causing this disagreement. To clarify, "Average Save=1.5*ECL" is not what I consider average for most characters, merely what I consider the lower-end of high-op play. The high end of high-op play is more like "AS=ECL*2+". "AS=ECL" is pretty standard; having an average save lower than your level isn't uncommon, but it's hardly high-op. What's more, even low-op characters will usually have an average close to their level, even if it tends to be on the low side; if their average is significantly lower (say, "AS=ECL/2"), it's usually because they've over-optimized for something other than saving throws (for instance, most ubercharger builds consist of 2-4 low-level dips into classes with only a good Fort save; their average save sucks).

The higher numbers are things that can be reached, just not without straying into high-op play. Of course, that's not to say that all high-op characters have saves that high, or that a character optimized for high saves and nothing else is high-op, just that characters with saves that high aren't generally going to be low-op characters.

I'm still working on that build, BTW. I've even avoided certain char-op tricks because they didn't fit the idea. And it looks like it's still gonna get some powerful spellcasting on top of it's ridiculous saves, so there's that.

[Seerow]

So are you going to show some sample characters from your high level games? Because I still want to see how you are getting saving throws that high on a typical character. I can understand if you specifically like having high defenses and build every character with multiple stacking save bonuses to hit that threshold, but when talking about an average, that sort of thing is useless.


Also the Greater Stone Golem has a pretty high DC for its level. A CR16 Blue Dragon has a DC of 27 for its breath weapon. Other dragons seem to be around the same range. CR17 Marilith tops out at a DC 25. CR16 horned devil tops out at DC27. Every one of these monsters in addition to their big high DC ability also has a bunch of other abilities with DCs significantly lower.

This is just from a quick sweep of dragons/demons/devils in the SRD since they're big easily searchable entries. Point is, DC31 is not the norm at CR16. It's actually a very high DC for that level. Someone with a +16 save bonus (1/level) will have a roughly 50% success rate on most of these abilities. Meanwhile somebody with the supposed target of +25 is going to be passing the hardest DCs from all of these encounters on a 2. That is not normal. That is a character with exceptional defenses, not the baseline expected average.

Edit:

The higher numbers are things that can be reached, just not without straying into high-op play. Of course, that's not to say that all high-op characters have saves that high, or that a character optimized for high saves and nothing else is high-op, just that characters with saves that high aren't generally going to be low-op characters.


Nobody cares about what numbers "can be reached". Yes I can build a character who hits +50 in all saves by focusing on saves. That character might even be capable of doing something else. The point is that the question being asked was what are the target numbers for a PC to have for saves, and your answer provides numbers that go beyond ludicrous. If your point all along was about what was possible, not what was reasonable or what is expected, then you've done nothing but add confusion to the OP who is seeking something completely different.


I'm still working on that build, BTW. I've even avoided certain char-op tricks because they didn't fit the idea. And it looks like it's still gonna get some powerful spellcasting on top of it's ridiculous saves, so there's that.


Why do you need to work on a build?

Go grab a character sheet from a game you have played in recently, and post the stats. If you are building something specifically to showcase the level you are looking for, rather than having an example already close at hand, you've already pretty much lost this argument.


edit2: as a reminder:

If I were DMing, I'd probably say that a character's overall average save bonus should be roughly equal to their level in a low-op game, and higher in higher-op games (depending on exactly how "high-op" those games are); for reasonably high-op games, I'd probably expect a character's overall average save bonus to be somewhere between "lvl*1.5" and "lvl*2.5".


So you're expecting saves = level as a low op. saves = level * 1.5 as the minimum in high op. This was clearly not a statement in any way of what is possible, but rather very clearly a statement of what is expected. All I want to know is what is it that every high op character in the game is doing to get saves literally double what I demonstrated with my characters? I don't want to see a build tailored specifically to saving throws unless you want to claim whatever it is you are using for that build is bog standard for every high op character out there.

[AvatarVecna]

I spent 10 seconds coming up with a vague estimate to at least try to answer the OP's question, and I get wall-of-text after wall-of-text saying that the average character, build with next to no optimization, could accomplish such a thing, and therefore my 10-second estimate is wrong and I am stupid. Never mind that some people play taking more than one class. Or. Or play with fractional save progression. Or gestalt/tristalt. Nope! Stupid. Stupid stupid AvatarVecna.

Think what you will, Seerow. I think, and have always thought, that "AS=ECL" is average and acceptable on a character. I also acknowledge the existence of characters that get higher than that. I'm done here. Go find someone else to pick on.

[Seerow]

I spent 10 seconds coming up with a vague estimate to at least try to answer the OP's question, and I get wall-of-text after wall-of-text saying that the average character, build with next to no optimization, could accomplish such a thing, and therefore my 10-second estimate is wrong and I am stupid. Never mind that some people play taking more than one class. Or. Or play with fractional save progression. Or gestalt/tristalt.

2 out of 3 of the samples I posted use 4+ classes. And gestalt/tristalt is certainly not what anyone would use to set the standard for expected numbers.


Look, I wouldn't be holding you to the 10 second guestimate if after having it pointed out that the guestimate was way off, you let it go.



Nope! Stupid. Stupid stupid AvatarVecna.

Think what you will, Seerow. I think, and have always thought, that "AS=ECL" is average and acceptable on a character. I also acknowledge the existence of characters that get higher than that. I'm done here. Go find someone else to pick on.

I am not calling you stupid, I am pointing out that the estimate is wrong. Which is important in the context of this being a thread where somebody actually wanted a usable answer, and what was provided to them by other posters here had a very high probability of creating unrealistic expectations. And I still remain open to the possibility that there are high op tricks I simply am not aware of that provide monstrous save bonuses at very little cost, but I haven't seen them mentioned yet, except for a bit about multiclassing (already accounted for in my estimates) and tristalt.

For what it's worth I am sorry, and did not intend anything said as a personal attack.

[Vrock_Summoner]

Pfft. Greater Stone Golems are sort of borked in general. I mean, DC 31, on a Slow effect (hurts melees unbelievably), plus Magic Immunity (huge pain in the arse for lower-op casters) and DR 10/Adamantine to give the rangers a hard time, just to top it all off. And that Slow is a Free action. The only thing salvaging them in a lot of low-op games is their nonexistent grey matter meaning that if the DM is playing them right they shouldn't be too hard to outwit. That and they lack flight, so subverting them entirely rather than fighting is usually an option.

I do agree that Saves tend to increase faster as you level, since more options open up for increasing them. But not at that excessive a rate.

For me, it's usually something like +1 in weak saves at level 1 (unless you have something RP-related going on; weak-willed characters are oddly popular, or maybe Wis is just the universal dump stat), +3-4 at level 5, 8-9 at 10, 16-17 at 15, and maybe a little over 20 at 20.

In Core games, this applies less, as there really isn't a whole lot to hope for outside of attribute increases and Resistance bonuses. Well, I mean, there's Paladin or Blackguard, but those come with a lot of RP baggage that doesn't work for all characters.

[Arbane]

These creature design guidelines (http://www.d20pfsrd.com/bestiary/rules-for-monsters/monster-creation) are for Pathfinder, but they might still be useful.

Looks like the high DCs are usually around 11+ (CR * 4/5), and the low DCs are supposed to be 8 + (CR / 2).

The 'optimal' amount for a save is "lol don't need to I'm immune", but as a general rule I'd think succeeding 75% of the time is pretty good, so a save of +(level) is good if you can manage it.

[Jowgen]

In terms of hard numbers and calculations, I can't contribute much to this discussion (which has gone in a way that makes me kinda sad :smallfrown: ), but I do have some anecdotal data that might be insightful.

At the current point, I'm playing a very multiclassed character (Ranger 2, Paladin 2, Fighter2, Stalker of Kharash 2, Shadow Scout 2, Sword of Righteousness 1, if anyone cares). Being a Magic-blooded Feytouched who transition-classed to get Half-fey, I have high cha that adds onto my multi-classing boosted saves. My lowest save (Will) is at about 3/4 of my ECL, while my other two saves average at just about x2 my ECL (AS=1.6*ECL). I can provide two obervations based on this.

1st, being able to either mostly or entirely ignore one save (e.g. Undead for Fort), significantly skews any estimate regarding the importance of saves, as being able to focus on less than all 3 frees up considerable resources to boost the other two. Coversely, this means that the DM may shift-things around to focus on the non-immune saves, increasing the importance of having them high, so that may lead some counter-scaling depending on the game.

2nd, x2 ECL in saves seems like too high an estimate to me. Adjusting to the power-level of the group I'm part of, the DM has consistently been using creatures/encounters around 2-3 CRs higher than the average party level, and challanging ones at that (e.g. 2 Living Holocausts in the Elemental Plane of fire today). I have terrible luck with saving throw rolls, but my character is yet to have failed on anything but a natural 1. The party also has 2 arcane casters that are reasonably well optimized, and the saves for the spells that they have been using aren't much harder that those we've been facing (maybe 3 to 4 higher, on average I'd guess).

This is obviously all very anecdotal and should be taken with a spoon of salt. Going beyond this and drawing on the trends I myself have noticed, I'd probably estimate as follows (my opinion, YMMV):

Save = 1 x ECL is the unoptimized goal to play "fair", or the minimum for op-levels ranging from decent to the lower-end of high.

Save = 1.5 ECL is the awesome for unoptimized, fair-to-good at decent-level op, and just about fair for lower-end op

Save = 2 x ECL is outlandishly high for unoptimized, great for decent and lower-end-of-high op, and still respectable good for real OP.

Anything above seem like it would go into Saveing-throw-monster overdrive, which is the point where racking up immunities is likely far more practical/acchieveable.

So yeah, I'd go for ECL x 1.5 as the golden respectable middle-ground for most common playing groups. You won't be unbeatable in the presnece of anything but the softest DM, but you should be able to handle most things most of the time with anything but the harshest of DMs.

This is my 2 coppers, and again, I am really no expert. Hope it helps in some way :smallsmile: