Why is targeting Reflex good at high levels and Fort at low levels?

[jmax]

I've seen lots of people say that Reflex is the save you want to target at high levels and many people say that targeting Fort is good at low levels. I repeated the advice in my new spellcasting handbook, but when asked where the assertion originally came from, I realized I didn't know. I think I originally got it from Treantmonklvl20's wizard guide, but I'm not certain.

Does anyone know where that comes from? Or are we talking "Moon is goddess; it is known" here?

(Credit to RexDart for making me question my knowledge.)

[tyckspoon]

It depends a lot on what your DM likes to use in combats, but basically there is a trend for low level enemies to be smaller, weaker, and often basic humanoids, monstrous humanoids, and low-tier animals or other smaller and more fragile opponents (ie, lower HD, not much in the way of immunities, and depending on specific type may have a poor base Fortitude save, and if they have notable status bonuses they're more likely to be in Dex.) Then as you increase CR stuff starts getting inflated in size and HD, which comes with big Constitution bonuses and base save totals, and the things that -aren't- scaled that way start becoming more and more likely to just be immune to most kinds of Fortitude saves ... but being ridiculously enormous usually comes with a poor Dex, so even if something has good base Reflex progression by type it will likely have Reflex as its weakest save due to that being the lowest stat.

I'd probably lean toward Will as the most effective low-level save type, tho - it has the most restrictions, with things like being Language-Dependent or Mind Affecting, but in low levels you are usually the most likely to be dealing with things where these don't matter, and a successful Will-targeting spell will often take one more enemies out of a fight entirely.

[Biffoniacus_Furiou]

Opponents' saves are based on two things, ability scores and maybe HD depending on creature type. In the lower levels, ability scores are likely to contribute more to this. In the higher levels, good saves from HD are going to be extremely high. Most monster types (based on encounter numbers) have good Fort saves but poor Reflex saves.

Even if it doesn't have a good base save from HD, monsters' Con scores tend to go up as their CR goes up, but not their Dex scores unless it's a Dex-focused opponent.

Also keep in mind that the higher level you get, the higher the chances that a given opponent will be outright immune to whatever Will save effect you're throwing at them. Thus their Will save typically won't even matter, unless you're using something nothing is immune to, like Time Hop.

So most higher level opponents won't have Reflex saves as high as their other saves. Whether from not having it as a good save from HD, or having a lower Dex than Con, it's generally reliable that most enemies' reflex saves are lower than their other saves in the higher levels.


With all that in mind, from the mid to high levels any given spellcaster should have at least one rank in every knowledge skill that identifies a creature by type (Arcana, Dungeoneering, Nature, Religion, the Planes, maybe Local). This automatically entitles you to making a check to identify a given creature upon first seeing/detecting it with no action required on your character's part. So get a notecard and on it list each creature type, which knowledge skill identifies it, which good/poor saves it has for its HD, and any flat immunities from type it gets. These things would be common knowledge for anyone who has the appropriate knowledge skill, so even if you can't remember them, your character sure could. You can somewhat metagame these checks based on which knowledge skill you use and what the DC is since it's based on HD, if you're told what the DC is. Play long enough and it's generally easy to figure out what a given creature's type would be based on its appearance anyway. Once you know a creature's type, you'll generally have a good idea of what its poor base saves are and you can pick what spells you use based on that.

[jmax]

It depends a lot on what your DM likes to use in combats, but basically there is a trend for low level enemies to be smaller, weaker, and often basic humanoids, monstrous humanoids, and low-tier animals or other smaller and more fragile opponents (ie, lower HD, not much in the way of immunities, and depending on specific type may have a poor base Fortitude save, and if they have notable status bonuses they're more likely to be in Dex.) Then as you increase CR stuff starts getting inflated in size and HD, which comes with big Constitution bonuses and base save totals, and the things that -aren't- scaled that way start becoming more and more likely to just be immune to most kinds of Fortitude saves ... but being ridiculously enormous usually comes with a poor Dex, so even if something has good base Reflex progression by type it will likely have Reflex as its weakest save due to that being the lowest stat.

Yeah, that's essentially my intuition. I just didn't have any evidence to back it up. Saintheart has since provided that evidence in the thread where this first came up.

[AvatarVecna]

Before I start seriously researching, I'll just state my expectations for mage-building:

Ideally, you are a wizard who knows exactly what you're going to be fighting. And thus, you can prepare the perfect spell for that particular enemy. Maybe the spell in question is useless against 99% of monsters, but it's devastating for this one. But if you don't know what you're going to be fighting, or if you have to pick a spell that will stay useful for the rest of your career (the way sorcerers do), you've gotta pick general all-purpose spells that maximize their overall usefulness across all targets, because you have no idea what all monsters are going to be involved in the campaign. Incidentally, this is part of why spellcasting guides tend to suggest you make a buff build. This is partly because it lets you be overpowered in a way that doesn't overshadow the party, but it's also because you don't necessarily know what debuffs are best for the enemy, but you basically always know what buffs are really good for your party.

From that perspective, if you want to be a blaster/debuffer, there are three saves to target. Will saves, you have to be careful with: Aberrations, Dire Animals, Dragons, Fey, Monstrous Humanoids, Outsiders, and Undead all have good Will saves. Additionally, Humanoids will tend towards class levels; the most dangerous of those will be casters, who will also have good Will saves. Finally, most Constructs, Oozes, Plants, Undead, and Vermin are mindless, giving them straight-up immunity to vast swathes of Will save effects. And that's one of the harder immunities to work around. This leaves your target types for generic Will save spells as Elementals, Giants, and Magical Beasts (and non-Dire Animals). That's...not a long list.

Now, instead of going type by type for Fort and Ref, I'm gonna focus on two particular creature types, because they exist across a wide range of CRs: Animal, and True Dragons. Animals have good Fort/Ref save from HD, so it's a matter of Dex vs Con, really. Small animals will tend to have higher Dex than Con...and also have low HD, low NA, and thus low CR. Big animals will tend to have lower Dex and higher Con...and also have high HD, high NA, and thus high CR. Thus, we can conclude that lower CR animals will tend to have better Ref than Fort saves, and higher CR animals will tend to have better Fort than Ref saves. Now, let's talk about dragons: dragons have good saves across the board, and tend to start with okay-ish Con, but Dex 10. Later, their Con will increase, but their Dex will remain 10. So for low CR dragons, Ref starts off a little bit behind Fort, but as CR grows, so too does Fort's lead over Ref. That's just the kind of thing you might know off-hand if you've spent enough time absorbing statblocks.

So, serious research. What I'm gonna do is go through the monster manual one statblock at a time, recording each monster's saves and CR, and then I'll sort them and see how the data shakes out by CR.

Spoiler: Raw data

Show

Fort	Ref	Will	CR
7	3	11	7
15	10	15	17
7	6	7	5
1	4	4	3
14	12	12	14
14	13	15	16
18	18	20	23
0	2	-5	0.5
0	1	-5	1
0	0	-5	2
1	1	-4	3
2	1	-3	5
5	3	0	7
10	7	5	10
6	3	2	3
5	5	4	4
2	2	2	2
6	5	6	4
18	11	13	16
14	11	11	14
4	8	4	3
6	10	6	5
12	14	10	8
7	1	2	3
9	5	10	8
4	4	4	2
6	7	7	4
9	8	10	5
9	4	3	5
18	12	8	12
11	7	5	8
5	4	9	6
9	5	11	13
3	10	2	6
4	7	4	2
3	5	7	8
2	4	1	2
11	8	6	7
3	3	5	4
6	7	2	4
3	6	5	3
7	7	6	7
9	7	6	7
2	5	4	2
7	6	9	7
5	5	7	5
4	7	2	3
8	9	10	10
3	2	0	1
12	6	11	9
10	6	6	6
22	19	19	20
16	9	9	10
5	3	3	2
18	8	11	13
16	7	9	11
19	14	14	17
17	10	15	14
3	6	4	2
3	6	3	11
6	6	7	7
14	9	10	9
2	5	6	3
5	5	10	8
14	14	12	11
8	7	5	5
12	12	11	9
8	8	6	6
11	11	10	8
9	11	8	7
16	16	15	16
15	14	15	13
3	6	4	2
3	3	3	1
19	19	21	20
4	4	11	11
9	10	3	6
8	6	2	3
15	9	4	7
10	8	4	6
19	9	6	9
16	12	8	8
6	6	5	3
7	6	4	2
7	10	6	2
12	9	9	7
8	5	8	4
9	8	7	5
3	5	3	0.33
14	13	12	9
13	12	11	8
3	7	4	2
8	7	6	3
8	7	5	4
8	7	3	4
16	14	9	12
4	5	6	3
5	4	4	3
6	5	5	4
9	7	7	5
10	8	8	7
13	10	11	9
15	11	12	11
18	13	15	14
19	14	16	16
22	16	19	18
23	17	20	19
26	19	23	20
28	20	25	22
6	5	5	3
8	6	6	4
10	8	9	6
12	9	11	8
15	11	13	11
17	12	15	14
19	14	17	16
21	15	19	18
23	17	21	19
25	18	23	21
28	20	25	23
29	21	27	25
5	4	4	3
8	6	6	4
9	7	8	5
12	9	11	8
14	10	12	11
17	12	15	13
18	13	16	16
21	15	19	18
22	16	20	19
25	18	23	21
27	19	24	22
29	21	27	24
7	5	5	4
10	7	8	5
11	8	9	7
14	10	12	10
16	11	13	13
18	13	17	15
20	14	18	18
23	16	21	20
25	17	23	21
28	19	26	23
30	20	27	24
32	22	30	26
4	3	3	2
6	5	5	3
8	6	6	4
10	8	8	6
12	9	9	8
15	11	11	10
17	12	13	12
19	14	15	15
21	15	17	17
23	17	19	18
25	18	20	19
28	20	24	21
5	4	4	3
6	5	5	4
9	7	8	6
10	8	9	8
13	10	12	10
15	11	13	12
18	13	16	15
19	14	17	17
22	16	20	19
23	17	21	20
26	19	24	21
28	20	25	23
6	5	7	3
8	6	8	5
10	8	11	7
12	9	13	9
15	11	15	12
17	12	17	15
19	14	19	17
21	15	21	19
23	17	23	20
25	18	25	22
28	20	28	23
29	21	29	25
5	4	5	3
7	6	7	5
9	7	9	7
11	9	11	9
13	10	13	11
16	12	15	14
18	13	17	16
20	15	19	19
22	16	21	20
24	18	23	22
26	19	25	23
29	21	27	25
8	6	8	5
10	7	10	7
12	9	12	9
14	10	14	11
17	12	16	14
18	13	18	16
21	15	20	19
23	16	23	21
26	18	26	22
28	19	28	24
31	21	31	25
33	22	33	27
6	5	7	4
9	7	9	5
10	8	11	7
13	10	14	10
15	11	15	13
18	13	18	15
19	14	19	18
22	16	22	20
23	17	24	21
26	19	27	23
29	20	29	24
32	22	32	26
13	8	9	9
9	8	4	7
5	4	8	7
3	8	6	3
4	0	-1	0.5
4	0	-1	1
5	7	3	3
8	9	7	6
9	8	10	13
0	6	0	1
3	9	1	3
5	13	2	5
9	19	5	7
11	22	9	9
12	25	10	11
4	-1	0	1
7	0	1	3
10	1	2	5
15	4	7	7
17	6	9	9
19	7	10	11
0	4	0	1
3	7	1	3
5	11	2	5
9	17	7	7
11	20	9	9
14	23	10	11
4	0	0	1
7	2	1	3
10	4	2	5
15	9	7	7
17	14	9	9
19	16	10	11
2	1	-1	0.5
2	1	-1	1
1	5	5	3
3	4	1	3
4	4	6	3
9	2	5	6
3	5	3	0.5
3	4	2	0.5
6	7	5	3
7	8	8	7
12	12	11	10
19	-	19	17
14	9	6	12
4	-	-4	1
6	-1	0	3
5	6	4	4
7	9	7	5
7	9	7	8
9	10	9	8
6	7	7	4
4	2	2	7
0	2	5	1
1	4	6	3
16	6	10	11
14	4	9	10
14	3	6	9
25	13	13	17
12	3	4	7
13	6	7	8
13	6	7	9
17	8	13	13
7	4	5	5
7	8	5	6
4	1	-2	1
3	3	0	1
4	0	0	1
4	0	-1	0.5
5	3	2	1
3	1	-1	0.33
3	2	3	10
3	2	3	7
6	5	6	13
4	3	4	11
14	12	14	16
11	6	5	8
14	7	4	8
0	2	5	3
8	7	5	4
1	4	2	1
10	11	8	9
13	11	10	12
6	6	6	6
6	7	7	5
2	4	4	4
14	8	10	11
8	2	4	6
7	5	9	9
4	2	0	0.5
2	7	6	4
11	14	11	11
5	5	4	3
13	10	9	9
6	5	2	2
4	1	-1	0.5
0	4	1	1
7	8	7	3
9	5	3	4
10	6	4	5
10	6	4	6
11	7	4	7
11	7	5	8
12	8	5	9
12	8	5	10
13	9	6	11
9	5	3	6
10	6	4	7
10	6	4	8
11	7	4	9
11	7	5	10
12	8	5	11
12	8	5	12
13	9	6	13
9	5	3	6
10	6	4	7
10	6	4	8
11	7	4	9
11	7	5	10
12	8	5	11
12	8	5	12
13	9	6	13
6	7	7	12
7	6	8	15
4	2	5	9
4	10	4	7
2	1	-1	0.25
21	12	13	12
3	5	1	1
3	3	5	2
7	8	7	6
8	8	7	8
13	12	10	13
4	7	10	13
7	10	8	7
1	3	0	1
3	1	1	0.5
8	5	4	5
12	6	4	5
12	6	4	5
6	3	3	4
9	3	3	4
9	3	3	4
5	2	4	2
6	5	4	2
6	5	4	2
8	5	4	5
10	7	4	5
10	7	4	5
6	3	2	3
8	5	2	3
8	5	2	3
16	12	13	14
17	12	11	14
17	12	11	14
17	8	12	11
20	8	12	11
20	8	12	11
3	3	0	3
9	7	3	5
3	7	6	7
3	6	3	3
3	6	3	3
4	2	3	3
3	4	3	3
3	6	3	3
3	4	3	3
4	3	3	3
4	2	3	3
3	4	3	3
4	3	3	3
4	1	-1	0.5
5	5	6	4
3	4	9	8
8	9	16	17
6	5	5	4
4	10	9	8
4	2	8	5
13	8	20	15
5	7	8	8
7	7	11	10
7	6	9	9
6	5	8	7
12	9	10	9
8	7	6	5
16	11	10	11
12	10	23	18
11	11	19	16
9	11	17	14
7	12	12	7
6	0	1	3
12	2	2	7
7	1	3	8
9	-2	-2	7
15	1	1	12
9	-4	-4	3
6	-4	-4	4
8	-3	-3	5
3	0	-2	0.5
3	2	6	4
5	7	3	3
9	5	2	4
7	6	4	3
6	0	0	3
7	7	2	5
11	11	11	7
3	0	0	0.5
3	1	-1	0.5
4	5	4	1
17	8	4	12
8	7	6	10
5	5	5	5
4	3	4	5
10	6	3	7
18	13	9	9
10	8	8	12
2	4	5	3
3	4	4	2
6	5	6	3
8	7	8	6
12	10	11	10
2	5	5	2
2	5	5	4
8	6	5	4
1	3	4	3
3	5	7	8
7	5	5	5
9	2	4	6
5	5	5	8
4	5	1	2
0	1	2	0.33
0	3	3	1
1	3	4	2
2	5	5	3
3	5	6	4
3	3	7	5
4	7	8	6
5	7	10	7
6	7	11	8
1	1	3	2
8	7	3	7
10	8	4	8
10	7	6	9
12	10	9	10
14	15	13	13
2	5	7	7
12	8	7	9
10	7	3	7
7	7	8	8
8	8	5	5
9	5	2	5
1	6	3	1
0	5	3	1
0	6	4	4
0	6	4	5
2	6	1	0.5
3	7	3	2
5	7	3	4
10	14	7	8
4	6	2	3
4	6	2	2
3	3	0	1
38	29	20	20
12	2	4	6
4	2	2	2
26	13	21	21
5	4	4	3
7	6	6	5
13	10	10	9
10	1	7	8
4	3	4	2
5	-1	0	1
11	4	3	5
16	8	8	11
8	3	6	7
17	6	15	14
9	7	6	3
16	12	15	13
4	6	4	7
7	16	13	15
1	5	5	4
3	3	3	2
1	2	5	3
3	12	9	6
8	6	3	5
6	6	3	2
1	4	6	5
5	14	14	11
7	6	6	6
6	7	5	6
5	5	5	3
5	3	3	3
7	5	5	6
13	9	9	8
11	10	5	9
1	5	4	3
3	6	9	5
6	7	11	7
0	0	3	0.5
0	-1	3	0.5
1	-1	4	1
2	2	5	2
2	0	6	3
4	3	8	4
4	4	9	4
5	5	10	5
6	5	12	6
6	6	2	2
3	4	1	0.5
4	5	1	0.5
2	4	2	0.1
5	4	2	2
9	6	3	4
10	7	3	4
7	4	1	2
6	3	2	2
5	6	1	1
2	4	1	0.25
5	7	2	2
6	4	2	2
9	6	3	4
4	5	1	0.33
5	5	1	1
4	4	0	0.16
3	4	2	0.5
12	7	6	7
2	5	2	0.33
5	4	2	1
5	4	2	1
7	5	2	2
6	4	2	1
5	5	1	1
5	7	2	2
6	7	2	3
2	4	1	0.16
8	5	2	2
4	4	2	1
2	4	1	0.16
6	4	1	1
3	6	1	1
7	8	3	8
2	5	2	0.25
4	4	0	0.25
5	4	0	0.5
4	6	1	0.5
2	4	1	0.12
2	4	2	0.16
11	6	3	4
4	5	2	1
8	7	3	2
11	9	4	4
4	6	2	2
2	5	1	0.33
2	5	1	0.5
3	6	1	1
3	6	2	2
6	7	3	3
8	10	4	5
3	6	2	1
9	11	5	9
8	7	3	4
2	3	2	0.1
2	4	1	0.25
14	9	5	6
15	9	6	7
11	8	5	5
5	5	1	1
7	5	2	2
3	3	2	1
3	0	0	1
4	0	1	2
5	0	2	2
5	0	0	2
2	0	0	0.33
8	2	2	4
6	0	3	3
6	2	2	3
2	2	0	0.12
2	2	0	0.25
2	2	0	0.5
3	3	1	1
6	4	2	2
9	6	4	6
15	9	8	9
4	0	0	0.25
4	0	0	0.5
5	0	0	1
6	1	1	3
10	3	3	7
15	6	6	10
25	12	13	12
2	3	0	0.25
2	3	0	0.5
4	3	0	1
5	4	1	2
8	5	2	5
12	8	5	8
20	12	10	11

Spoiler: Average Saves By CR

Show

CR	Fort	Ref	Will
<1	+2.604	+2.708	+0.542
1	+3.091	+3.116	+0.909
2	+4.396	+4.313	+2.666
3	+4.543	+4.171	+3.314
4	+6.295	+5.136	+4.522
5	+7.380	+5.900	+4.960
6	+8.000	+6.483	+5.828
7	+8.666	+6.917	+6.292
8	+9.171	+7.571	+7.371
9	+12.188	+9.313	+8.375
10	+11.444	+8.111	+8.555
11	+13.630	+10.333	+9.926
12	+15.066	+9.600	+8.666
13	+13.000	+9.733	+11.266
14	+15.750	+11.250	+13.833
15	+14.625	+11.875	+15.500
16	+16.500	+13.000	+16.000
17	+18.125	+12.714	+16.250
18	+19.714	+14.429	+19.429
19	+22.125	+16.125	+20.125
20	+24.222	+18.666	+21.222
20<	+27.226	+19.226	+25.935

In retrospect, this isn't too surprising. There's monsters with good Reflexes in the higher CRs, but they're very much not the default. Fortitude overtakes Reflex quickly, and it doesn't take too long for Will saves to do so as well, even taking the weak-willed monsters into account. Early on, when really weak enemies are serious threats to you, targeting Fortitude is frequently a good idea - but only a hair better of an idea than targeting Ref. Past CR 2, averages suggest Reflex is the weaker save worth targeting.

What all of this post doesn't really take into account is the kind of spells that force Fort and Ref saves. Namely, Ref save spells tend to be direct damage, while Fort is usually conditions, and maybe damage. Rule of thumb: the most useful condition to inflict on an enemy is "dead".

EDIT: ...and unsurprisingly, it turns out that what I was doing has been done before, and better, multiple times. I should probably google before I start putting in tons of work lol.

[Troacctid]

What I really wish we had was a survey of monster stats over the course of a given adventure path. After all, in actual gameplay, you aren't going to be fighting every monster in the Monster Manual, in order, from lowest CR to highest. Some enemy types are going to be more common than others. For example, humanoid is one of the least common creature types in the Monster Manual, but it's the most common creature type in most adventures.

[pabelfly]

Before I start seriously researching, I'll just state my expectations for mage-building:

Ideally, you are a wizard who knows exactly what you're going to be fighting. And thus, you can prepare the perfect spell for that particular enemy. Maybe the spell in question is useless against 99% of monsters, but it's devastating for this one. But if you don't know what you're going to be fighting, or if you have to pick a spell that will stay useful for the rest of your career (the way sorcerers do), you've gotta pick general all-purpose spells that maximize their overall usefulness across all targets, because you have no idea what all monsters are going to be involved in the campaign. Incidentally, this is part of why spellcasting guides tend to suggest you make a buff build. This is partly because it lets you be overpowered in a way that doesn't overshadow the party, but it's also because you don't necessarily know what debuffs are best for the enemy, but you basically always know what buffs are really good for your party.

From that perspective, if you want to be a blaster/debuffer, there are three saves to target. Will saves, you have to be careful with: Aberrations, Dire Animals, Dragons, Fey, Monstrous Humanoids, Outsiders, and Undead all have good Will saves. Additionally, Humanoids will tend towards class levels; the most dangerous of those will be casters, who will also have good Will saves. Finally, most Constructs, Oozes, Plants, Undead, and Vermin are mindless, giving them straight-up immunity to vast swathes of Will save effects. And that's one of the harder immunities to work around. This leaves your target types for generic Will save spells as Elementals, Giants, and Magical Beasts (and non-Dire Animals). That's...not a long list.

Now, instead of going type by type for Fort and Ref, I'm gonna focus on two particular creature types, because they exist across a wide range of CRs: Animal, and True Dragons. Animals have good Fort/Ref save from HD, so it's a matter of Dex vs Con, really. Small animals will tend to have higher Dex than Con...and also have low HD, low NA, and thus low CR. Big animals will tend to have lower Dex and higher Con...and also have high HD, high NA, and thus high CR. Thus, we can conclude that lower CR animals will tend to have better Ref than Fort saves, and higher CR animals will tend to have better Fort than Ref saves. Now, let's talk about dragons: dragons have good saves across the board, and tend to start with okay-ish Con, but Dex 10. Later, their Con will increase, but their Dex will remain 10. So for low CR dragons, Ref starts off a little bit behind Fort, but as CR grows, so too does Fort's lead over Ref. That's just the kind of thing you might know off-hand if you've spent enough time absorbing statblocks.

So, serious research. What I'm gonna do is go through the monster manual one statblock at a time, recording each monster's saves and CR, and then I'll sort them and see how the data shakes out by CR.

Spoiler: Raw data

Show

Fort	Ref	Will	CR
7	3	11	7
15	10	15	17
7	6	7	5
1	4	4	3
14	12	12	14
14	13	15	16
18	18	20	23
0	2	-5	0.5
0	1	-5	1
0	0	-5	2
1	1	-4	3
2	1	-3	5
5	3	0	7
10	7	5	10
6	3	2	3
5	5	4	4
2	2	2	2
6	5	6	4
18	11	13	16
14	11	11	14
4	8	4	3
6	10	6	5
12	14	10	8
7	1	2	3
9	5	10	8
4	4	4	2
6	7	7	4
9	8	10	5
9	4	3	5
18	12	8	12
11	7	5	8
5	4	9	6
9	5	11	13
3	10	2	6
4	7	4	2
3	5	7	8
2	4	1	2
11	8	6	7
3	3	5	4
6	7	2	4
3	6	5	3
7	7	6	7
9	7	6	7
2	5	4	2
7	6	9	7
5	5	7	5
4	7	2	3
8	9	10	10
3	2	0	1
12	6	11	9
10	6	6	6
22	19	19	20
16	9	9	10
5	3	3	2
18	8	11	13
16	7	9	11
19	14	14	17
17	10	15	14
3	6	4	2
3	6	3	11
6	6	7	7
14	9	10	9
2	5	6	3
5	5	10	8
14	14	12	11
8	7	5	5
12	12	11	9
8	8	6	6
11	11	10	8
9	11	8	7
16	16	15	16
15	14	15	13
3	6	4	2
3	3	3	1
19	19	21	20
4	4	11	11
9	10	3	6
8	6	2	3
15	9	4	7
10	8	4	6
19	9	6	9
16	12	8	8
6	6	5	3
7	6	4	2
7	10	6	2
12	9	9	7
8	5	8	4
9	8	7	5
3	5	3	0.33
14	13	12	9
13	12	11	8
3	7	4	2
8	7	6	3
8	7	5	4
8	7	3	4
16	14	9	12
4	5	6	3
5	4	4	3
6	5	5	4
9	7	7	5
10	8	8	7
13	10	11	9
15	11	12	11
18	13	15	14
19	14	16	16
22	16	19	18
23	17	20	19
26	19	23	20
28	20	25	22
6	5	5	3
8	6	6	4
10	8	9	6
12	9	11	8
15	11	13	11
17	12	15	14
19	14	17	16
21	15	19	18
23	17	21	19
25	18	23	21
28	20	25	23
29	21	27	25
5	4	4	3
8	6	6	4
9	7	8	5
12	9	11	8
14	10	12	11
17	12	15	13
18	13	16	16
21	15	19	18
22	16	20	19
25	18	23	21
27	19	24	22
29	21	27	24
7	5	5	4
10	7	8	5
11	8	9	7
14	10	12	10
16	11	13	13
18	13	17	15
20	14	18	18
23	16	21	20
25	17	23	21
28	19	26	23
30	20	27	24
32	22	30	26
4	3	3	2
6	5	5	3
8	6	6	4
10	8	8	6
12	9	9	8
15	11	11	10
17	12	13	12
19	14	15	15
21	15	17	17
23	17	19	18
25	18	20	19
28	20	24	21
5	4	4	3
6	5	5	4
9	7	8	6
10	8	9	8
13	10	12	10
15	11	13	12
18	13	16	15
19	14	17	17
22	16	20	19
23	17	21	20
26	19	24	21
28	20	25	23
6	5	7	3
8	6	8	5
10	8	11	7
12	9	13	9
15	11	15	12
17	12	17	15
19	14	19	17
21	15	21	19
23	17	23	20
25	18	25	22
28	20	28	23
29	21	29	25
5	4	5	3
7	6	7	5
9	7	9	7
11	9	11	9
13	10	13	11
16	12	15	14
18	13	17	16
20	15	19	19
22	16	21	20
24	18	23	22
26	19	25	23
29	21	27	25
8	6	8	5
10	7	10	7
12	9	12	9
14	10	14	11
17	12	16	14
18	13	18	16
21	15	20	19
23	16	23	21
26	18	26	22
28	19	28	24
31	21	31	25
33	22	33	27
6	5	7	4
9	7	9	5
10	8	11	7
13	10	14	10
15	11	15	13
18	13	18	15
19	14	19	18
22	16	22	20
23	17	24	21
26	19	27	23
29	20	29	24
32	22	32	26
13	8	9	9
9	8	4	7
5	4	8	7
3	8	6	3
4	0	-1	0.5
4	0	-1	1
5	7	3	3
8	9	7	6
9	8	10	13
0	6	0	1
3	9	1	3
5	13	2	5
9	19	5	7
11	22	9	9
12	25	10	11
4	-1	0	1
7	0	1	3
10	1	2	5
15	4	7	7
17	6	9	9
19	7	10	11
0	4	0	1
3	7	1	3
5	11	2	5
9	17	7	7
11	20	9	9
14	23	10	11
4	0	0	1
7	2	1	3
10	4	2	5
15	9	7	7
17	14	9	9
19	16	10	11
2	1	-1	0.5
2	1	-1	1
1	5	5	3
3	4	1	3
4	4	6	3
9	2	5	6
3	5	3	0.5
3	4	2	0.5
6	7	5	3
7	8	8	7
12	12	11	10
19	-	19	17
14	9	6	12
4	-	-4	1
6	-1	0	3
5	6	4	4
7	9	7	5
7	9	7	8
9	10	9	8
6	7	7	4
4	2	2	7
0	2	5	1
1	4	6	3
16	6	10	11
14	4	9	10
14	3	6	9
25	13	13	17
12	3	4	7
13	6	7	8
13	6	7	9
17	8	13	13
7	4	5	5
7	8	5	6
4	1	-2	1
3	3	0	1
4	0	0	1
4	0	-1	0.5
5	3	2	1
3	1	-1	0.33
3	2	3	10
3	2	3	7
6	5	6	13
4	3	4	11
14	12	14	16
11	6	5	8
14	7	4	8
0	2	5	3
8	7	5	4
1	4	2	1
10	11	8	9
13	11	10	12
6	6	6	6
6	7	7	5
2	4	4	4
14	8	10	11
8	2	4	6
7	5	9	9
4	2	0	0.5
2	7	6	4
11	14	11	11
5	5	4	3
13	10	9	9
6	5	2	2
4	1	-1	0.5
0	4	1	1
7	8	7	3
9	5	3	4
10	6	4	5
10	6	4	6
11	7	4	7
11	7	5	8
12	8	5	9
12	8	5	10
13	9	6	11
9	5	3	6
10	6	4	7
10	6	4	8
11	7	4	9
11	7	5	10
12	8	5	11
12	8	5	12
13	9	6	13
9	5	3	6
10	6	4	7
10	6	4	8
11	7	4	9
11	7	5	10
12	8	5	11
12	8	5	12
13	9	6	13
6	7	7	12
7	6	8	15
4	2	5	9
4	10	4	7
2	1	-1	0.25
21	12	13	12
3	5	1	1
3	3	5	2
7	8	7	6
8	8	7	8
13	12	10	13
4	7	10	13
7	10	8	7
1	3	0	1
3	1	1	0.5
8	5	4	5
12	6	4	5
12	6	4	5
6	3	3	4
9	3	3	4
9	3	3	4
5	2	4	2
6	5	4	2
6	5	4	2
8	5	4	5
10	7	4	5
10	7	4	5
6	3	2	3
8	5	2	3
8	5	2	3
16	12	13	14
17	12	11	14
17	12	11	14
17	8	12	11
20	8	12	11
20	8	12	11
3	3	0	3
9	7	3	5
3	7	6	7
3	6	3	3
3	6	3	3
4	2	3	3
3	4	3	3
3	6	3	3
3	4	3	3
4	3	3	3
4	2	3	3
3	4	3	3
4	3	3	3
4	1	-1	0.5
5	5	6	4
3	4	9	8
8	9	16	17
6	5	5	4
4	10	9	8
4	2	8	5
13	8	20	15
5	7	8	8
7	7	11	10
7	6	9	9
6	5	8	7
12	9	10	9
8	7	6	5
16	11	10	11
12	10	23	18
11	11	19	16
9	11	17	14
7	12	12	7
6	0	1	3
12	2	2	7
7	1	3	8
9	-2	-2	7
15	1	1	12
9	-4	-4	3
6	-4	-4	4
8	-3	-3	5
3	0	-2	0.5
3	2	6	4
5	7	3	3
9	5	2	4
7	6	4	3
6	0	0	3
7	7	2	5
11	11	11	7
3	0	0	0.5
3	1	-1	0.5
4	5	4	1
17	8	4	12
8	7	6	10
5	5	5	5
4	3	4	5
10	6	3	7
18	13	9	9
10	8	8	12
2	4	5	3
3	4	4	2
6	5	6	3
8	7	8	6
12	10	11	10
2	5	5	2
2	5	5	4
8	6	5	4
1	3	4	3
3	5	7	8
7	5	5	5
9	2	4	6
5	5	5	8
4	5	1	2
0	1	2	0.33
0	3	3	1
1	3	4	2
2	5	5	3
3	5	6	4
3	3	7	5
4	7	8	6
5	7	10	7
6	7	11	8
1	1	3	2
8	7	3	7
10	8	4	8
10	7	6	9
12	10	9	10
14	15	13	13
2	5	7	7
12	8	7	9
10	7	3	7
7	7	8	8
8	8	5	5
9	5	2	5
1	6	3	1
0	5	3	1
0	6	4	4
0	6	4	5
2	6	1	0.5
3	7	3	2
5	7	3	4
10	14	7	8
4	6	2	3
4	6	2	2
3	3	0	1
38	29	20	20
12	2	4	6
4	2	2	2
26	13	21	21
5	4	4	3
7	6	6	5
13	10	10	9
10	1	7	8
4	3	4	2
5	-1	0	1
11	4	3	5
16	8	8	11
8	3	6	7
17	6	15	14
9	7	6	3
16	12	15	13
4	6	4	7
7	16	13	15
1	5	5	4
3	3	3	2
1	2	5	3
3	12	9	6
8	6	3	5
6	6	3	2
1	4	6	5
5	14	14	11
7	6	6	6
6	7	5	6
5	5	5	3
5	3	3	3
7	5	5	6
13	9	9	8
11	10	5	9
1	5	4	3
3	6	9	5
6	7	11	7
0	0	3	0.5
0	-1	3	0.5
1	-1	4	1
2	2	5	2
2	0	6	3
4	3	8	4
4	4	9	4
5	5	10	5
6	5	12	6
6	6	2	2
3	4	1	0.5
4	5	1	0.5
2	4	2	0.1
5	4	2	2
9	6	3	4
10	7	3	4
7	4	1	2
6	3	2	2
5	6	1	1
2	4	1	0.25
5	7	2	2
6	4	2	2
9	6	3	4
4	5	1	0.33
5	5	1	1
4	4	0	0.16
3	4	2	0.5
12	7	6	7
2	5	2	0.33
5	4	2	1
5	4	2	1
7	5	2	2
6	4	2	1
5	5	1	1
5	7	2	2
6	7	2	3
2	4	1	0.16
8	5	2	2
4	4	2	1
2	4	1	0.16
6	4	1	1
3	6	1	1
7	8	3	8
2	5	2	0.25
4	4	0	0.25
5	4	0	0.5
4	6	1	0.5
2	4	1	0.12
2	4	2	0.16
11	6	3	4
4	5	2	1
8	7	3	2
11	9	4	4
4	6	2	2
2	5	1	0.33
2	5	1	0.5
3	6	1	1
3	6	2	2
6	7	3	3
8	10	4	5
3	6	2	1
9	11	5	9
8	7	3	4
2	3	2	0.1
2	4	1	0.25
14	9	5	6
15	9	6	7
11	8	5	5
5	5	1	1
7	5	2	2
3	3	2	1
3	0	0	1
4	0	1	2
5	0	2	2
5	0	0	2
2	0	0	0.33
8	2	2	4
6	0	3	3
6	2	2	3
2	2	0	0.12
2	2	0	0.25
2	2	0	0.5
3	3	1	1
6	4	2	2
9	6	4	6
15	9	8	9
4	0	0	0.25
4	0	0	0.5
5	0	0	1
6	1	1	3
10	3	3	7
15	6	6	10
25	12	13	12
2	3	0	0.25
2	3	0	0.5
4	3	0	1
5	4	1	2
8	5	2	5
12	8	5	8
20	12	10	11

Spoiler: Average Saves By CR

Show

CR	Fort	Ref	Will
<1	+2.604	+2.708	+0.542
1	+3.091	+3.116	+0.909
2	+4.396	+4.313	+2.666
3	+4.543	+4.171	+3.314
4	+6.295	+5.136	+4.522
5	+7.380	+5.900	+4.960
6	+8.000	+6.483	+5.828
7	+8.666	+6.917	+6.292
8	+9.171	+7.571	+7.371
9	+12.188	+9.313	+8.375
10	+11.444	+8.111	+8.555
11	+13.630	+10.333	+9.926
12	+15.066	+9.600	+8.666
13	+13.000	+9.733	+11.266
14	+15.750	+11.250	+13.833
15	+14.625	+11.875	+15.500
16	+16.500	+13.000	+16.000
17	+18.125	+12.714	+16.250
18	+19.714	+14.429	+19.429
19	+22.125	+16.125	+20.125
20	+24.222	+18.666	+21.222
20<	+27.226	+19.226	+25.935

In retrospect, this isn't too surprising. There's monsters with good Reflexes in the higher CRs, but they're very much not the default. Fortitude overtakes Reflex quickly, and it doesn't take too long for Will saves to do so as well, even taking the weak-willed monsters into account. Early on, when really weak enemies are serious threats to you, targeting Fortitude is frequently a good idea - but only a hair better of an idea than targeting Ref. Past CR 2, averages suggest Reflex is the weaker save worth targeting.

What all of this post doesn't really take into account is the kind of spells that force Fort and Ref saves. Namely, Ref save spells tend to be direct damage, while Fort is usually conditions, and maybe damage. Rule of thumb: the most useful condition to inflict on an enemy is "dead".

EDIT: ...and unsurprisingly, it turns out that what I was doing has been done before, and better, multiple times. I should probably google before I start putting in tons of work lol.

Here's a question for you, if you're playing with the raw data. What are the median averages for the saves, and do the differ significantly from mean average? I'm wondering if the mean values of saves are weighed towards statistical outliers at all.

If you'd go to the trouble, I'd suggest checking out, say, level 5, 10, 15 and 20 just to see if it's worth doing for the other levels.

[jmax]

What I really wish we had was a survey of monster stats over the course of a given adventure path. After all, in actual gameplay, you aren't going to be fighting every monster in the Monster Manual, in order, from lowest CR to highest. Some enemy types are going to be more common than others. For example, humanoid is one of the least common creature types in the Monster Manual, but it's the most common creature type in most adventures.

For adventure paths that have searchable PDFs, one might be able to write a script to extract and parse the stat blocks. I imagine you'd have to tune it a little for each specific module - but maybe not.

Hmm, here's an idea. Some of the VTTs have digital module files that they use to generate all the adventure material within the environment. Those would have to be in machine-friendly format already - if you can download them, and if they aren't in for stupid proprietary format, parsing that out should be fairly easy. You'd just have to be careful to capture the number of each enemy type.

[AvatarVecna]

I'm not a statistician so idk what a "median average" is, and Google isn't helping, but I can at least rearrange things once O get home.

[Ramza00]

I realized I didn't know. I think I originally got it from Treantmonklvl20's wizard guide, but I'm not certain.

Does anyone know where that comes from? Or are we talking "Moon is goddess; it is known" here?

(Credit to RexDart for making me question my knowledge.)

Carniovre Psion Guide used to have images that explain this. You can still see the text for people copy pasted it on other websites, but the 3.5 wizards dot com boards are gone, so unless you have a web archive link you will not see this simple excel graph that showed average cr and usually fort, reflex, will where they are for the mean monster.

I think Dictum Mortuum also had this in one of his handbooks such as his wizard's handbook, but do not quote me on this.

Pretty much

Big Monster, go for reflex, often has good fort due to HD, Size Increases, and higher Con.
Fort and Will can be done with most things regardless of body shape. That said Humanoids, Outsiders, etc often have better Will than non Humanoids, non Outsiders.

If Humanoid shape and playing Pathfinder CMB / CMD it makes sense to use spells or attacks with CMB, but pretty much everything else there is a 2/1 ratio where the monster gains 2 defense for 1 offense such as Caster Level increase, BAB increase, etc. Thus CMB / CMD is often not optimal if the monster is not humanoid.

I'm not a statistician so idk what a "median average" is, and Google isn't helping, but I can at least rearrange things once O get home.

Mean means you take all the numbers add them together and divide by the total amount of numbers, like 1+2+3+4+5+6 divided by 6 equals 3.5

Median is the middle arrangement in a group. Useful if you have have a large inequality in the data set, such as real world wealth where 5 people may make 30k, and then a rich person but not really rich person makes 180k and thus the Mean for those 6 people in a single room is 55k per person (5*30+180) / 6, but the Median is 30k for you put those 30k people on the number line and you look for person 3 and person 4 in that sequential data set.

[pabelfly]

I'm not a statistician so idk what a "median average" is, and Google isn't helping, but I can at least rearrange things once O get home.

There are three types of averages.

Mean average is the sum of a set of numbers, divided by how many are in that set of numbers. This is the average you worked out.
Median average is if you list the numbers in a group in numerical order and find the middle number
Mode average is what number is most common in a set of numbers.

The reason I think a median average would be interesting is because it ignores outliers in numbers that might otherwise skew the data. For an extreme example, if you had test scores of 10, 10, 15, 15, 20 and 100, the median average would be 15, because that's the middle number, while a mean average would be 28.3, the latter of which isn't a very useful number.

Hope this makes sense.

[AvatarVecna]

There are three types of averages.

Mean average is the sum of a set of numbers, divided by how many are in that set of numbers. This is the average you worked out.
Median average is if you list the numbers in a group in numerical order and find the middle number
Mode average is what number is most common in a set of numbers.

The reason I think a median average would be interesting is because it ignores outliers in numbers that might otherwise skew the data. For an extreme example, if you had test scores of 10, 10, 15, 15, 20 and 100, the median average would be 15, because that's the middle number, while a mean average would be 28.3, the latter of which isn't a very useful number.

Hope this makes sense.

Ohhhh I see the disconnect. Those other things aren't averages. Average and Mean are the same thing. Mode and median are not averages, but I know what those are. Honestly all three plus the range would be useful info.

[AvatarVecna]

Here's a question for you, if you're playing with the raw data. What are the median averages for the saves, and do the differ significantly from mean average? I'm wondering if the mean values of saves are weighed towards statistical outliers at all.

If you'd go to the trouble, I'd suggest checking out, say, level 5, 10, 15 and 20 just to see if it's worth doing for the other levels.

Spoiler: MM Median Saves

Show

CR	Fort	Ref	Will
<1	+2	+3	+1
1	+3	+3.5	+1
2	+4	+4.5	+2.5
3	+5	+4	+3
4	+6	+5	+5
5	+8	+6	+5
6	+8	+7	+5
7	+9	+7	+6
8	+10	+7	+7
9	+12	+9	+9
10	+12	+8.5	+9
11	+14	+9	+10
12	+15	+11	+8
13	+14	+9	+11
14	+17	+12	+15
15	+17.5	+13	+16.5
16	+18	+13	+16
17	+19	+14	+16.5
18	+21	+15	+19
19	+22	+16	+20
20	+23	+17	+21
20<	+28	+19	+25

I think this would be more accurate with more books, and I agree this kind of analysis would be more helpful if we were looking at an adventure module or even a full campaign.

[Anthrowhale]

I was curious about the median difference between the highest and lowest save, so I used AvatarVecna's raw data to get:

Spoiler: median max minus min

Show

CR	Median Max Save Minus Min Save-
1/3 or less	1.5
1/2	4
1	4
2	3.5
3	3
4	3
5	4.5
6	5
7	5
8	5
9	5
10	4
11	7
12	7
13	5
14	5
15	5
16	5
17	5
18	6
19	6
20	6
21 or more	8

So, the typical difference is about 5 and the difference is pretty flat across levels. Thus, targeting the weak save is typically worth ~25% or +10 to the casting stat.

[Saintheart]

I think this would be more accurate with more books, and I agree this kind of analysis would be more helpful if we were looking at an adventure module or even a full campaign.

Would you get a lot of difference out of default prewritten WOTC modules? I had the impression creature selections in modules tended to stick to the MM1 anyway, for simplicity and so the adventures had the widest reach to potential customers.

Admittedly I'm biased since most of my time is spent rereading Red Hand of Doom. That one is 90% MM1: hobgoblins, goblins, ogres, some Spawn of Tiamat and some devils right towards the end. There's four young dragons in it, but these are default MM1 as well. The differences most crop up in the few monsters that have class and prestige class levels, but I guess I could go through and pull out the saves and CRs from it if people are interested.

[AvatarVecna]

Would you get a lot of difference out of default prewritten WOTC modules? I had the impression creature selections in modules tended to stick to the MM1 anyway, for simplicity and so the adventures had the widest reach to potential customers.

Admittedly I'm biased since most of my time is spent rereading Red Hand of Doom. That one is 90% MM1: hobgoblins, goblins, ogres, some Spawn of Tiamat and some devils right towards the end. There's four young dragons in it, but these are default MM1 as well. The differences most crop up in the few monsters that have class and prestige class levels, but I guess I could go through and pull out the saves and CRs from it if people are interested.

It's not necessarily about different content being used, but different percentages. As Troacctid said, it's not like the party is going to face every single enemy in the monster manual exactly one time and then never again - and that's what my averages represent. There are 50 CR 5 monsters in the monster manual, but odds are you're gonna face half of those at most on your path to lvl 20, and that's assuming your DM makes an extremely diverse campaign. Knowing how WotC designs adventures, what kind of monsters they choose and how they're spread out, how many you face and when, that's useful information. If WotC has a tendency to spam low-level humanoids who tend to just be brute warriors (melee or ranged, barely matters), then what spells work best against them is gonna be very different from if the adventure was necromancers and undead. The first game will see you almost universally wanting to attack their will save, while the latter game, trying to get them to fail a will save is a waste of time and spell slots.

[Troacctid]

It's not necessarily about different content being used, but different percentages. As Troacctid said, it's not like the party is going to face every single enemy in the monster manual exactly one time and then never again - and that's what my averages represent. There are 50 CR 5 monsters in the monster manual, but odds are you're gonna face half of those at most on your path to lvl 20, and that's assuming your DM makes an extremely diverse campaign. Knowing how WotC designs adventures, what kind of monsters they choose and how they're spread out, how many you face and when, that's useful information. If WotC has a tendency to spam low-level humanoids who tend to just be brute warriors (melee or ranged, barely matters), then what spells work best against them is gonna be very different from if the adventure was necromancers and undead. The first game will see you almost universally wanting to attack their will save, while the latter game, trying to get them to fail a will save is a waste of time and spell slots.

Actually, in low-level premade adventures, the most prominent types are humanoid, followed by undead! So you really should be prepared for both.

[ericgrau]

The number of targets is an important factor too. There was a Red Hand of Doom campaign log tally where reflex was the save caused the most party trouble by far, because they tend to hit the most targets and parties have multiple members. Though at high level you can get some good multi-target will and fort save spells that hit a large area. And since it's multi-target, the foes are less likely to be huge. At lower level you have slow for example, but it's a bit deceptive in how many it can hit. It targets a tight area which limits it outdoors, and you need to be able to see all the targets which limits it indoors. How many times have you taken for granted the presence of a foe that only an ally can see? Unlike large AOE spells where you only need to be able to see the center.

The nice thing is you can get both single and multi-target spells and cast them as the situation requires. Because unlike PCs sometimes there is 1 and sometimes there are multiple.

[jmax]

The number of targets is an important factor too. There was a Red Hand of Doom campaign log tally where reflex was the save caused the most party trouble by far, because they tend to hit the most targets and parties have multiple members.

That works both ways, though. Call Avalanche is lovely for smashing nice, neat squares of hobgoblin soldiers marching in formation

[Seward]

The brief answer is the type of enemies and how they scale.

In low levels most enemies are vulnerable to fort save effects (stunning, nausea, necromancy, etc). They are also mostly a single hit die with constitution no higher than 13. So compared to what DC's a specialized spellcaster can get at level 1, it's effective. Meanwhile reflex save spells are only doing 1d4-1d6 damage, not enough to kill anything even if you fail a save, or do effects like trip or entangle, rather than depriving the enemy of meaningful actions. Many enemies are also high dex size small creatures, more likely to save or at least as likely to save (eg, a goblin warrior has a similar reflex and fort save)

Monsters, especially brute monsters, though, scale at 1CR per 3-4 hit dice. Further they scale by growing to larger size, which adds a massive constitution bonus for each size category higher than medium. They also all have good fort saves, except Aberrations, but even they get big enough bonuses from hit die scaling+size to have decent fort saves. However size growth lowers dexterity and even if size doesn't increase, dex doesn't tend to grow larger on anything but classed roguelike characters. So the reflex save grows slower than the fort save vs what a PC can do, and the spell effects for failing a reflex save are now doing meaningful damage, or completely incapacitating the enemy. Also highly advanced creatures may make pretty much any save on a "2" (dire animals, for example, have all good saves) but many reflex save spells still do something even if you save. Finally a lot of more common higher level enemies become flatly immune to many fort save effects, where they might only be resistant to things like energy damage or have holes in their protections (eg, lightning vs devils).

It isn't a hard and fast rule. Honestly I target will save a lot more at low levels and expect it to work (color spray, daze, sleep, glitterdust, calm emotions etc) and never expect fort save spells to be reliable, except rare ones that target undead (which lack constitution scores and have weak fort saves) that the undead aren't simply immune to. But I am skeptical of relying on save or XXX in any event, it's a very swingy thing vs one target and the level scaling usually means in higher levels it's likely to fail. I prefer if using such spells in combat to choose multi-target spells, with the assumption that at least SOMEBODY will fail, so the action isn't wasted, or spells where it still does something useful even if save is made (as with glitterdust revealing invisibility, or stinking cloud still being a fog cloud that blocks vision etc).

[jmax]

Further they scale by growing to larger size, which adds a massive constitution bonus for each size category higher than medium.

However size growth lowers dexterity

That's not strictly true, actually. Advanced monsters get bonuses to Con (and Strength) based on size increases, but for primary listings, it's a correlation rather than a guarantee. It is, however, a very strong correlation.

Likewise for Dex in terms of advancement, but extra-big printed monsters don't usually have Dex decreases - nor do extra-small have substantial increases.

Only the actual Size modifiers (AC, attack rolls, and combat maneuver stuff) are guaranteed based on size.

even if size doesn't increase, dex doesn't tend to grow larger on anything but classed roguelike characters. So the reflex save grows slower than the fort save vs what a PC can do

That one I think is a very strong correlation, and it's likely responsible for a large part of why Reflex doesn't scale as well on monsters.

In low levels most enemies are vulnerable to fort save effects (stunning, nausea, necromancy, etc). They are also mostly a single hit die with constitution no higher than 13. So compared to what DC's a specialized spellcaster can get at level 1, it's effective. Meanwhile reflex save spells are only doing 1d4-1d6 damage, not enough to kill anything even if you fail a save, or do effects like trip or entangle, rather than depriving the enemy of meaningful actions.

This is good insight. I hadn't been thinking about how little damage the direct-damage spells deal at 1st level.

Finally a lot of more common higher level enemies become flatly immune to many fort save effects, where they might only be resistant to things like energy damage or have holes in their protections (eg, lightning vs devils).

Yeah, you definitely have to watch for the immunities. Freedom of Movement will ruin your day for a lot of the Reflex-targeting battlefield control, but it's not common as an inherent trait.

It isn't a hard and fast rule. Honestly I target will save a lot more at low levels and expect it to work (color spray, daze, sleep, glitterdust, calm emotions etc) and never expect fort save spells to be reliable, except rare ones that target undead (which lack constitution scores and have weak fort saves) that the undead aren't simply immune to.

Now that I'm thinking about it, I think most of the straight-up save-or-lose spells at very early levels do target Will rather than Fort.

But I am skeptical of relying on save or XXX in any event, it's a very swingy thing vs one target and the level scaling usually means in higher levels it's likely to fail. I prefer if using such spells in combat to choose multi-target spells, with the assumption that at least SOMEBODY will fail, so the action isn't wasted, or spells where it still does something useful even if save is made (as with glitterdust revealing invisibility, or stinking cloud still being a fog cloud that blocks vision etc).

Yeah, I talk about that in my guide. Though if you spam save-or-lose enough, eventually the target will get unlucky. Then you get to find out if they have good safety nets :-P

[Anthrowhale]

This is good insight. I hadn't been thinking about how little damage the direct-damage spells deal at 1st level.

I spent some time looking into this awhile ago. CR 1 monsters typically have ~12 hp. Using averages, you can think of the 1st level benchmark damage spells as:
Sunstroke (Sandstorm): 1/2 nonlethal + Fort-or-fatigue, close range
Light of Lunia (BoED): 1/2 or 1 undead/evil outsiders, Ranged touch, 30' range, 1 round activation but precastable.
Lesser Orb of <X>: 1/3 <X>, SR:No, Ranged Touch, Close Range
Burning Hands: 1/5 Fire, Refl/2, 15' cone.
Power Word Pain: 1/4 / round
Magic Missile: 1/4, Medium Range
Hail of Stone: 1/5, SR:No, 5' radius, 1 round activation

If you are willing to spend a feat, Bloodline of Fire provides:
Lesser Orb of Fire: 3/4 Fire, SR:No, Ranged Touch, Close Range
Burning Hands: 1/2 Fire, Refl/2, 15' cone.

A feat + the sorcerer fire sphere or fire domain wizard gives:
Burning Hands: 4/5 Fire, Refl/2, 15' cone.

With 2 feats, Iron Will + Reserves of Strength additionally gives:
Magic Missile: 1/2, Medium Range
Hail of Stone: 4/5, SR:No, 5' radius, 1 round activation

Overall, nothing is an instakill. Sunstroke and Light of Lunia do significant (1/2 kill) damage. Spending a feat on boosting caster level has a similar effect for the fire spells which scale better at later levels. I discount the Reserves of Strength approach, because you really need 3 feats (Iron Will+RoS+Quick Recovery) to deploy it effectively.

[jmax]

I spent some time looking into this awhile ago. CR 1 monsters typically have ~12 hp. Using averages, you can think of the 1st level benchmark damage spells as:
Sunstroke (Sandstorm): 1/2 nonlethal + Fort-or-fatigue, close range
Light of Lunia (BoED): 1/2 or 1 undead/evil outsiders, Ranged touch, 30' range, 1 round activation but precastable.
Lesser Orb of <X>: 1/3 <X>, SR:No, Ranged Touch, Close Range
Burning Hands: 1/5 Fire, Refl/2, 15' cone.
Power Word Pain: 1/4 / round
Magic Missile: 1/4, Medium Range
Hail of Stone: 1/5, SR:No, 5' radius, 1 round activation

If you are willing to spend a feat, Bloodline of Fire provides:
Lesser Orb of Fire: 3/4 Fire, SR:No, Ranged Touch, Close Range
Burning Hands: 1/2 Fire, Refl/2, 15' cone.

A feat + the sorcerer fire sphere or fire domain wizard gives:
Burning Hands: 4/5 Fire, Refl/2, 15' cone.

With 2 feats, Iron Will + Reserves of Strength additionally gives:
Magic Missile: 1/2, Medium Range
Hail of Stone: 4/5, SR:No, 5' radius, 1 round activation

Overall, nothing is an instakill. Sunstroke and Light of Lunia do significant (1/2 kill) damage. Spending a feat on boosting caster level has a similar effect for the fire spells which scale better at later levels. I discount the Reserves of Strength approach, because you really need 3 feats (Iron Will+RoS+Quick Recovery) to deploy it effectively.

Yeah, that definitely puts the low-level damage spells in perspective. I definitely prefer Entangle and then plinking from afar until they're all dead.

Though it's worth noting that, at 12 hp, the Fighter probably isn't one-shotting level-appropriate enemies at 1st level either.

That said, an average level-appropriate encounter will have only one CR1 enemy for four 1st-level PCs. A team-on-team combat will have fractional-CR enemies, which may fall to a single hit from either a 1st-level spell (even at CL 1) or the Fighter's pointy object of choice.

[pabelfly]

Though it's worth noting that, at 12 hp, the Fighter probably isn't one-shotting level-appropriate enemies at 1st level either.

A Fighter with just 18 STR gets +6 damage with a two-handed weapon. If you're wielding a 1d12 or 2d6 weapon, that's a better than 50% chance of a one-shot kill. And I haven't even picked a single feat or race to manage that.

[jmax]

A Fighter with just 18 STR gets +6 damage with a two-handed weapon. If you're wielding a 1d12 or 2d6 weapon, that's a better than 50% chance of a one-shot kill. And I haven't even picked a single feat or race to manage that.

Fair point. Certainly better odds than blasting at 1st level.

[Anthrowhale]

Using Eldariel's averages with an 18 stat level 1 PC vs. a CR 1 enemy:

1. A will save-or-lose works 65% of the time.
2. A fort save-or-lose works 55% of the time.
3. Two-handed sword with BAB+1 works 50% of the time.
4. Sunstroke does ~60% damage.

It's actually remarkably similar.

A mass save-or-lose definitely does stand out though, since the effect of multiple targets is more substantial. Color Spray (Will), Entangle (Refl), Impeding Stones (Refl), and Sticky Floor (Refl) look interesting.

[jmax]

Using Eldariel's averages with an 18 stat level 1 PC vs. a CR 1 enemy:

1. A will save-or-lose works 65% of the time.
2. A fort save-or-lose works 55% of the time.
3. Two-handed sword with BAB+1 works 50% of the time.
4. Sunstroke does ~60% damage.

It's actually remarkably similar.

A mass save-or-lose definitely does stand out though, since the effect of multiple targets is more substantial. Color Spray (Will), Entangle (Refl), Impeding Stones (Refl), and Sticky Floor (Refl) look interesting.

*yoink* Bookmarked the Eldariel link. Thanks!

I never really paid much attention to Sunstroke because it scales so poorly, but you're right - at 1st level, that's not bad. However, the Fighter with the greatsword can try as many times as she wants, whereas the caster only has a few shots.

[ericgrau]

In low levels most enemies are vulnerable to fort save effects (stunning, nausea, necromancy, etc). They are also mostly a single hit die with constitution no higher than 13. So compared to what DC's a specialized spellcaster can get at level 1, it's effective. Meanwhile reflex save spells are only doing 1d4-1d6 damage, not enough to kill anything even if you fail a save, or do effects like trip or entangle, rather than depriving the enemy of meaningful actions. Many enemies are also high dex size small creatures, more likely to save or at least as likely to save (eg, a goblin warrior has a similar reflex and fort save)

At level 1 it's 1d4-1d6 but at levels 2 and 3 you double and triple that. The average difficult foe at level 1 is CR 1/2 (4 = EL 3, which is a "difficult" fight). That's around 5 HP. At CR 1 you might get into the low teens. I don't usually get any damage spells at level 1 but at levels 2-3 it's really good. Especially when it's save half and you can hit 2 at times so allies can finish off the enem(y/ies). Rather than risking all or nothing on a level when you really want to avoid risks. If you get lucky and wipe the floor with one fight but unlucky and someone dies the next one, that's still a dead PC. Better to struggle but still win when the dice are bad. Routine fights have less HP than that and it's quite the opposite of not killing on the failed save: Often they die even on a passed save. Especially if there are multiple fights per day as usually is the case at low level. The CR will be lower in expectation of spent PCs. Another trick after you get the slots, level 3-4 or so, is to charge up your familiar with a shocking grasp. Have him hit first, then cast a spell to do double duty in round 1. In general low enemy HP makes damage work out well at low levels. Dead is as good or better than disabled.

I also like other save partial or no save spells but that's getting a little into a tangent. Web and ray of enfeeblement are very nice at low levels for example.

[jmax]

At level 1 it's 1d4-1d6 but at levels 2 and 3 you double and triple that. The average difficult foe at level 1 is CR 1/2 (4 = EL 3, which is a "difficult" fight). That's around 5 HP. At CR 1 you might get into the low teens. I don't usually get any damage spells at level 1 but at levels 2-3 it's really good. Especially when it's save half and you can hit 2 at times so allies can finish off the enem(y/ies). Rather than risking all or nothing on a level when you really want to avoid risks. If you get lucky and wipe the floor with one fight but unlucky and someone dies the next one, that's still a dead PC. Better to struggle but still win when the dice are bad. Routine fights have less HP than that and it's quite the opposite of not killing on the failed save: Often they die even on a passed save. Especially if there are multiple fights per day as usually is the case at low level. The CR will be lower in expectation of spent PCs. Another trick after you get the slots, level 3-4 or so, is to charge up your familiar with a shocking grasp. Have him hit first, then cast a spell to do double duty in round 1. In general low enemy HP makes damage work out well at low levels. Dead is as good or better than disabled.

I also like other save partial or no save spells but that's getting a little into a tangent. Web and ray of enfeeblement are very nice at low levels for example.

A lot does come down a lot to how tough the encounter is and whether it's a bunch of weak enemies or a single strong one. There's a big difference between 4 1st-level PCs fighting a single bugbear and the same PCs fighting 6 goblins even though those are both ECL 2 ("Very Difficult") encounters.

The point about 1st-level play being extra deadly on both sides due to high randomness and low hit points is a good one.

I cover save-and-you-suck-anyway and no-save-just-suck in my (still a working draft) spellcasting handbook.

With regards to the familiar and Shocking Grasp, etc. - just remember that dead familiars are expensive. They're (always) squishier than you are, and tiny or smaller ones provoke attacks of opportunity when entering an enemy square to deliver touch spells. I like Deliver Touch Spells better for buffing allies since the risk is much smaller.

[ericgrau]

A lot does come down a lot to how tough the encounter is and whether it's a bunch of weak enemies or a single strong one. There's a big difference between 4 1st-level PCs fighting a single bugbear and the same PCs fighting 6 goblins even though those are both ECL 2 ("Very Difficult") encounters.

The point about 1st-level play being extra deadly on both sides due to high randomness and low hit points is a good one.

I cover save-and-you-suck-anyway and no-save-just-suck in my (still a working draft) spellcasting handbook.

With regards to the familiar and Shocking Grasp, etc. - just remember that dead familiars are expensive. They're (always) squishier than you are, and tiny or smaller ones provoke attacks of opportunity when entering an enemy square to deliver touch spells. I like Deliver Touch Spells better for buffing allies since the risk is much smaller.

I should add that I like to use false life and mage armor on my familiar. And even without false life the familiar gets in his one hit and bolts his next turn for the rest of the fight. I actually tried this and the familiar was good at soaking an attack of opportunity (especially against foes with reach) and taking a hit as well instead of the party. If the monster even bothered with him since he was usually better off hitting a PC. Pretty soon the DM wised up and wouldn't even use an attack of opportunity with reach foes nor others half the time, saving it for a PC instead who could actually do something the rest of the fight. The familiar never died nor came close the whole campaign. I'm not sure the false life was really necessary. But even if he did die xp is a river. When you are behind you get extra xp and gain it back quickly. An extra action in the first round of the fight and soaking a hit or 2 ideally (a pro not a con) makes the trade-off an easy choice. Plus if a familiar can't survive 1 round, then a PC can't survive 2 or 3 rounds, and a dead PC is way more expensive. If the familiar dies instead of a PC or even only increases the PC's odds a bit, that's a big win.