Why Do We Need Reflex Saves?

[Vadskye]

This isn't a full proposal yet; I'm still working out the implications of this idea. Feedback is very much appreciated.

Reflex saves aren't like Fortitude or Will saves, which are about resisting hostile effects that are currently affecting you. Reflex saves represent two distinct things. First, they represent being able to move out of the way of a physical attack. That's the exact same as AC. Second, they represent being able to react quickly to things (like pit traps). That's the exact same as initiative. Why do we need Reflex saves?

For the purpose of this argument, I'm going to ignore purely mechanical considerations like "initiative checks don't scale, so they shouldn't be used to avoid pit traps". Maybe that just means that initiative checks should scale, but actually creating a functional system without Reflex saves is a secondary concern here; I'm just looking at this from a theoretical game design perspective.

One potential counterargument is that AC includes factors that shouldn't affect Reflex saves. Traditionally, armor has had zero effect on your ability to avoid a fireball or lightning bolt. However, let's take a closer look at this. Being hit by a fireball isn't like standing in an oven; we can safely ignore facts like "metal conducts heat" because that only takes place over a long period of time.

Suppose you have to grab a hot tray of cookies out of the oven. Would you rather do it bare-handed, wearing a leather glove, or wearing a full plate gauntlet? It seems clear to me that the more armor you have, the more protected you are from the immediate heat. There's no reason why armor shouldn't protect you from a fireball. It's easy to make the same argument about natural armor or a shield, too. In fact, I would say that all armor bonuses should apply against fireball. Insight and dodge bonuses help you avoid the flames in the same way that Dexterity and Reflex saves do, deflection bonuses deflect the fire away from you, and so on.

What about lightning bolt? Maybe metal armor shouldn't protect you much from that. Or maybe the current would be redirected to run through the armor instead of through your body, protecting you. This is a question for people who know their physics (electrical engineers?), but it doesn't mean we need a Reflex save. If armor shouldn't help, just make lightning bolt an attack against touch AC instead of regular AC! (Or if you decide that you shouldn't be able to dodge lightning, make it an attack against flat-footed AC. I really like that idea personally.) You still don't need a Reflex save.

I'm eager to see if there are spells or effects which simply don't make sense in a world without Reflex saves. But I'm pretty sure that you can represent everything just as well using a combination of AC and initiative checks.

[Ziegander]

You know what you're talking about here, Vadskye. Ultimately, Reflex saves resist damage. You know what else resists damage? AC. Armor Class or "defense" or whatever you want to call it should fully protect characters from damage. There does not need to be two separate mechanics for deciding how well a character can avoid damage.

[Just to Browse]

Because attacks/AC run off a terrible bonus system, and it's easier to tack on a level-appropriate defense mechanic than rewrite the hot mess that is armor class.

Touch AC doesn't scale. If you add a scaling factor to fix that, now you need to rewrite monster scaling, which means editing monster strength (which changes damage, you need to rewrite weapons now) or monster hit dice (which means you need to re-adjust HP, saves, skills).

Of course if you were rewriting a system from the top, you could also remove reflex. But then you could instead remove AC, and maybe make initiative and spells less terrible.

[NichG]

Lets go at this from the other direction. What is the mechanical role of abilities that currently target Reflex saves?

Structurally, they tend to be 'area damage effects'. Fort saves and Will saves tend to be more resistance against instakills and status effects. There's a 10% set of cases where reflex saves come into play to avoid area status effects instead (Web and the like), which is worth being aware of, but most of the time its area damage. So, if you change how Reflex saves work, you (in 90% of cases) are just changing the mechanical role of area damage effects.

So now the question is, what do you want the mechanical role of area damage effects to be? What do you want them to be good at doing or poor at doing? That will determine what kind of system you should attach to resisting them.

Currently, the existence of AC, Touch AC, and three Save types means that you have a spectrum of strengths and vulnerabilities. Reducing that breadth means that its easier to create a character that is resistant to all effects, and also that it is harder to specifically target a weakness. Replacing an element of that breadth with some other statistic lessens this effect or removes it entirely of course, but then it depends on how easy it is to control that statistic (e.g. optimizing initiative vs optimizing a saving throw)

The other consideration when merging statistics is that you create correlations. Someone who is concerned about area of effect damage will tend to always move early in the combat order (for Initiative), or will need to be heavily armored (for AC). This can influence what archetypes can be created within the system.

[erikun]

It's because they are two different sub-systems. While D&D3e did convert everything to a d20+bonus die roll, they never bothered to sync the different systems together in parallel. This is why, for example, climb checks are so much higher than AC, or why targetting a will save is so much easier than a touch attack. It's especially noticeable in D&D4e, where reflex is right next to AC but has slightly different modifiers. (Fun note: shields improve reflex in D&D4e, under the logic that a shield will block part of the Fireball blast.)

There is also the touch AC oddity, where a Lightning Bolt is resisted by armor but Shocking Grasp is not. And if you try removing touch AC, you need to give wizards a massive BAB boost to compensate... which just causes more problems with the D&D3e system.

[Grod_The_Giant]

I certainly see your point about AC and Initiative being pretty much the same as Reflex. However, I would rather fold them into Reflex than ditch Reflex in favor of them-- there's a much more appealing symmetry that way, and I think that's how Star War Saga Edition does it.

Armor is a bit of a weird spot, as you've identified. On the one hand, you don't have to dodge so much if you're wearing 50 pounds of metal. On the other, it's a lot harder to dodge while wearing 50 pounds of metal. That dichotomy is probably why Reflex and AC were broken apart, because it is difficult to deal with. If we're combining everything, I'd recommend ditching the idea of armor bonuses to Reflex altogether, and have it be purely a matter of DR. Maybe even go so far as to apply ACP to Reflex, to get a bit of a trade-off between damage avoidance and damage resistance.

[Plato Play-Doh]

Well...I've actually been fiddling with similar concepts. For more info, look in my signature. The defenses overhaul modifies AC and Reflex so they don't tread on each other's territory, among other things.

Seriously though, I think a better question is "why do we need AC?" not "why do we need reflex saves?"

[toapat]

all armor bonuses should apply against fireball

this is specifically dependant on how you define fireball as functioning.

If its an explosion, then it should bypass armor upto halfplate because it deals damage by crushing through pressure points on a foe, and only those armors contain ridged bascinets that would protect the primary pressure points which can kill you (ears, eyes)

However, a fuel air bomb would bypass armor and require an explicit dodge because it bakes the local area.

personally i believe its type 1 due to why the FAB was developed

[Jallorn]

I certainly see your point about AC and Initiative being pretty much the same as Reflex. However, I would rather fold them into Reflex than ditch Reflex in favor of them-- there's a much more appealing symmetry that way, and I think that's how Star War Saga Edition does it.

Actually, SWS has AC as Reflex, but Initiative as a skill. I rather like the SWS system for d20.


Armor is a bit of a weird spot, as you've identified. On the one hand, you don't have to dodge so much if you're wearing 50 pounds of metal. On the other, it's a lot harder to dodge while wearing 50 pounds of metal. That dichotomy is probably why Reflex and AC were broken apart, because it is difficult to deal with. If we're combining everything, I'd recommend ditching the idea of armor bonuses to Reflex altogether, and have it be purely a matter of DR. Maybe even go so far as to apply ACP to Reflex, to get a bit of a trade-off between damage avoidance and damage resistance.

Certainly not a bad idea. SWS is also good for a fun take on armor. It replaces level bonus to Reflex normally, which at first glance is bad past the first few levels, but there's a pair of really good talents that make armor really really good by allowing you to add half the armor bonus to your level bonus. Also it gives a small Fortitude bonus.

[Eldan]

I like the idea, but let's go a bit further, because AC is two things too. On the one hand it covers evasion. Dexterity, dodge bonuses, etc. On the other hand, it covers being difficult to hurt, almost like damage reduction.

So, how about we split it along different lines?

First:
A speed or reaction statistic, that rolls in touch AC, reflex saves, doge bonuses and initiative and allows you to avoid things that would hit you.

Second:
A damage resistence statistic, that involves armour and damage reduction. Perhaps use it as a "save against damage", even.

[FireJustice]

I got your idea. It makes sense.

But the entire d&d 3.5 math is really messed up.

It's not just reflex, so trying to fix just that is a enormous effort for very little result overall.

[OzymandiasX]

The function of AC currently is to model a combination of 'avoiding damage by not getting hit' and 'avoiding damage by absorbing/deflecting it.' For the vast majority of combat rolls, this simplification works well.

Reflex saves model getting the hell out of the way of something, be it a trap that opens under your feet or ducking to avoid the brunt of a fireball. Armor (in D&D) isn't considered here, as it sometimes would help (covering skin from the fireball) and would sometimes hinder (jumping off the ledge of the pit trap)... so the designers chose to just cancel it out.

You're right that there is definitely overlap between the two... but within the D&D system I think it is fairly well balanced and that combining these causes a lot of problems (as people have pointed out with how things scale with level, touch attacks, ways to increase AC that wouldn't apply to Ref scenarios, etc) Trying to 'tweak' it to combine AC and Ref causes more problems than it resolves.

If you're looking at building a new system from scratch, you'll need to work out the balance between these three aspects... and try to make it less complex. hehe Quite a challenge.

Avoidance - The ability to avoid a hit by not being there. This would be reduced by most armors. This would also be similar to Reflex-saves
Deflection - The ability of armor or a weapon to deflect damage away from vital areas or fleshy bits.
Absorption - The ability of armor to reduce (or negate) damage from an attack. This can vary by type of attack (bludgeoning, slashing, fire, etc) and by type of defense (leather, metal, cloth, etc).


Some systems I've played (HARN for example) detail all of these things by a series of item properties and charts that are quite realistic. Other systems (Savage Worlds) simplify into essentially "did it hurt or not?" I think D&D's system is a decent compromise between variety and over-simplification. Your mileage may vary, of course. :)

[Vadskye]

This is really nice feedback - thank you. I'll respond to specific points later, but I have another question. It seems like there's a general perception that AC is a bad system. Why is that? I've become rather fond of it. Fully separating "resistance" from "avoidance" could yield interesting results, but given that hit points are just an abstraction anyway, it has struck me as adding unnecessary complexity.

[OzymandiasX]

It seems like there's a general perception that AC is a bad system. Why is that?
AC and attacks that target AC don't scale at the same pace. By mid-levels, a fighter is almost guaranteed to hit with his first attack and is almost guaranteed to miss with his last. The scaling issues become even more wonky when comparing to attacks that target Touch AC.

[NichG]

AC and attacks that target AC don't scale at the same pace. By mid-levels, a fighter is almost guaranteed to hit with his first attack and is almost guaranteed to miss with his last. The scaling issues become even more wonky when comparing to attacks that target Touch AC.

Actually, this sounds ideal. One of the problems with D&D math is that the variance remains constant but the mean increases, so things become dominated by constant modifiers at high level. By adding attacks at increasing penalties, it broadens the range over which the details of attack bonus and AC matter compared to other things in the system.

For example, with a skill, I only have a range of 20 with respect to a fixed DC where my actual skill modifier matters. Beyond that I fail on a 20 or succeed on a 1. With an attack roll, I would have the same issue, but iterative attacks mean that by the time I'm Lv20 and have four attacks, my actual range of meaningful values is 35, not 20. Power attack expands this even more in principle.

So, I would not have actually called that out as the problem with AC. I think the main problem with AC is that its one of those things that you have to burn a LOT of character resources to make meaningful, but often its easier just to let it be really awful and go to secondary layers of protection instead. If I could have a choice between AC 30 or AC 10 and 50% miss chance, I'd generally take the latter.

[Vadskye]

Actually, this sounds ideal. One of the problems with D&D math is that the variance remains constant but the mean increases, so things become dominated by constant modifiers at high level. By adding attacks at increasing penalties, it broadens the range over which the details of attack bonus and AC matter compared to other things in the system.
Agreed. I often see this misunderstood, but that iterative scaling is very important.


So, I would not have actually called that out as the problem with AC. I think the main problem with AC is that its one of those things that you have to burn a LOT of character resources to make meaningful, but often its easier just to let it be really awful and go to secondary layers of protection instead. If I could have a choice between AC 30 or AC 10 and 50% miss chance, I'd generally take the latter.
So the issue is with the mechanics of AC, not with the concept of "a single number that represents how hard you are to hit"?

[OzymandiasX]

So the issue is with the mechanics of AC, not with the concept of "a single number that represents how hard you are to hit"?
Right. The 'single number' aspect is a simplification, but I think it works well. The mechanics work well at low levels, but by higher levels get silly, imo.