An alternative spell duration mechanic (saving throws vs defence)

[Zustiur]

I have seen complaints on these forums about both 3e's saves and 4e's defences. I think both camps have some good points in this regard. The only other solution I've seen proposed was 'players roll everything', which I personally am not a fan of. While pondering these three mechanics and the 'flatter math' goal of 5E, I had an idea. Before I elaborate on my idea, here are some of the complaints I've seen previously. These can be used as a reference point:

• 3E saving throws mean that the attacker doesn't play an active part in determining the success of the effect.
• 3E also couples this with long durations where you know from the start how long the duration will be because it is determined when the spell is cast (and this must be tracked)
• 3E's 'save or suck' spells can result in you having literally nothing to do on your turn.
• Defences mean that the defender (particularly referring to PCs) doesn't play an active part in determining the effect.
• 4E's saving throws (1-9 fail, 10-20 succeed) often result in 1 round duration, or less, so why not just make all spells 'until the end of your next turn'?

I believe my mechanic addresses each of these concerns without being as one-sided as 'players roll everything'.

The Mechanic
Attack vs Defence to set the initial DC
Subsequent saves are measured against that DC
A natural 20 always succeeds in both of these cases.
Where
Attack = d20 + stat modifier + spell level
Defence = 10 + stat modifier + optional modifiers such as feats, race, items etc (to be determined later)
Save = d20 + stat modifier + optional modifiers such as feats, race, items etc (to be determined later)
Individual spells may have other conditional rules. You might for example, have a +2 to the attack roll for a certain spell which has a minor impact, thus giving it a greater likely duration.

An example of how this mechanic might play out:
A cleric casts hold person on an orc. He rolls the attack roll and gets a natural 13 on his dice. He adds his Wisdom bonus (4) and the spell level (2) for a total of 19. This beats the orc's [relevant-stat] defence of 13, so the orc is held. At the end of orc's turn, it rolls a saving throw and gets a natural 14 on the dice. The DM adds this to the orc's [relevant-stat] modifier and gets a total of 17. This isn't enough to break free. On the subsequent round the DM rolls a total of 8 and still hasn't broken free. On a 3rd round the DM rolls a total of 20 and the orc is no longer held.
In a subsequent battle, the cleric uses the same spell on another orc. This time the attack roll total is only 13. The orc rolls a natural 10 on it's first saving throw attempt and breaks free after only one round.

The higher the initial attack roll, the longer an effect is likely to last, yet the effect can still be ended within 1 turn if the target gets lucky on its saving throws.

Additional Clauses
Sustain. I feel that particularly powerful spell effects should require sustaining. For example, Hold Person might be 'save ends' and 'sustain minor', with either condition getting you out. i.e. save ends, or until the spell is not sustained, whichever happens first. This factor should help to alleviate 'save or suck' by tying up the caster as well as the target. Note that some spells should take more than a minor action to sustain, particularly those which deal ongoing damage, or which knock one or more targets unconscious. e.g. sleep.
Buffs and debuffs. Assuming the 'advantage/disadvantage' mechanic survives, I see no reason not to have that as the effect from buffs and debuffs. Allies might provide advantage on saving throws that the recipient makes. Enemies might in turn provide disadvantage on the target's saving throws.
Modifiers (Feats/Items/etc). I'd want to keep these bonuses to a minimum so that the attack rolls don't become impossible. One idea is to have these only affect saving throws, without affecting defences. Thus making durations shorter without making spells useless.

Pros and Cons
Pros:

• This provides what I feel is a more satisfying 'save ends' mechanic because, 'save ends' will frequently last longer than one round
• Getting a really good attack roll to set the condition plays a part in determining the duration
• It prevents/avoids the need to track a number of rounds
• It means that neither side can be really sure when the effect will end
• It highlights the difference in strength between the attacker and the defender.
• If the 'sustain' part is included, then 'save or suck' is partially balanced by reducing the caster's capabilities (and incidentally this makes it harder to 'go nova')
• It means that having a poor defence stat doesn't just mean you get affected more often, it means those effects are more powerful too. Likewise, having a good defence means both that you'll be affected less frequently, and that you'll break out of the effect sooner.

Cons

• It is more complex to teach/learn
• You must write down the DC when the effect starts
• It requires more dice rolling
• If combined with the 'sustain' part, it requires tracking on both sides (the caster and the target) because one must remember to sustain, and the other must remember the save DC.

If you spot other pros or cons (or indeed complaints about existing systems), please list them for me and I'll try to keep this post up to date.


I believe my mechanic addresses the existing concerns, and has pros which outweigh the cons. It is neither one edition's system nor the other, instead being a new entity taking inspiration from both prior mechanics. It is also somewhat more dynamic than either of the previous systems, which will hopefully lead to more interesting play experiences. With all that said; what your opinions on the mechanic?

 

[Mustrum_Ridcully]

The biggest drawback - sometimes the attack roll will be so high that the defender cannot save at all. I think having a saving throw to end the spell is fine, but I would simply use a static DC that is likely to be beatable.

I am also not convinced that the spell level should adjust the difficulty. Not in D&D Next, with bounded accuracy. Rolling high already gives you the benefit that you succeed.

What I might see as interesting is the idea that a spell doesn't expend if you cast it unsuccesfully at a target. Basically, it just teaches you some time to break through the enemies defense, but you will break it eventually. Maybe that violates the Vancian Magic aesthetics for some, but I believe it's a good way to reduce swinginess, and it feels thematically "right" to me - spells become a feared ability, you're never really save from them. But, at the same time, if your will is strong enough, you can break throug them.

 

[delericho]

I suspect what you have here is a mechanic that reads a whole lot better than it plays. It does indeed fix the 'problem'... but unfortunately I suspect the need to write down the saving throw of the effect when it first applies will prove a great deal of hassle for not enough gain.

A compromise I might consider would be to have the caster give a fixed saving throw DC for all his spells - sort of like 3e does, but without the spell level adding to the total. That way, you only need to worry about different numbers if multiple casters hit the target with spells.

But even so, that may be more trouble than it's worth.

 

[Zustiur]

The biggest drawback - sometimes the attack roll will be so high that the defender cannot save at all. I think having a saving throw to end the spell is fine, but I would simply use a static DC that is likely to be beatable.

I am also not convinced that the spell level should adjust the difficulty. Not in D&D Next, with bounded accuracy. Rolling high already gives you the benefit that you succeed.

On the first point, I have to admit that I'm comparing more with 3E than 4E. I mean that in 3E, once the spell has affected you, you're stuck until the duration runs out. This mechanic replaces the 'xdY turns' spell duration whilst also looking a bit like the 'save ends' mechanic of 4E. In practice may have more in common with 3E's spell duration than 4E's 'save ends'.

I'm also uncertain about the spell level being part of the attack roll/DC. I listed that more as a starting point. I fully expect numbers and exact details of the mechanic will need alteration. I didn't want to clutter up the first post with too much detail - rather get the idea out in the open first and see how people react. I agree with you; StatMod + Spell Level looks too powerful given what we know so far of 5E's saves. Perhaps it should simply be d20 + StatMod. Or perhaps individual spells should have different modifiers, where the greater the effect, the harder it is to 'hit'.

The key components are really: Attack vs Defence, followed by duration determined by saving throw. That is 4E's method, but my mechanic skews the total spell duration look more familiar to players of 3E.

 

[Zustiur]

I suspect what you have here is a mechanic that reads a whole lot better than it plays. It does indeed fix the 'problem'... but unfortunately I suspect the need to write down the saving throw of the effect when it first applies will prove a great deal of hassle for not enough gain.

Quite possible, though I hope it's good in play too. However, each mechanic needs some tracking.
AD&D and 3E both required you to track a number of rounds or turns. This in itself can be quite a headache.
4E required you to roll a dice instead of tracking rounds. Instead we end up tracking many variations on spell duration. Is it until the end of the caster's next turn? The beginning of the caster's next turn? Sustain? Save ends? Until the end of the target's next turn? Until some other weird condition is met, such as 'when I move' or 'if I attack the wrong guy'?
I personally found this form of tracking to be more irritating than tracking rounds. I also find the 'save ends' mechanic to be a bit 'flat' in play. Firstly because it never really varied from needing a 10, and secondly because all too often the effect only lasted one round.
Conversely, my mechanic only requires you to write down a single number (the DC). This number doesn't change from round to round. Nor does it necessarily have to live in a game with many other potential 'spell ends' conditions.

A simpler alternative might be to have:
Spell duration = Attack Total - Defence Total.
If you 'hit' you automatically know how many turns the spell will last. If you beat the target's defence by 5, it will last 5 rounds. No rolling for duration and no rolling for subsequent saves. However, I still prefer the mechanic I outlined above for the single reason that it means you won't know exactly how long the spell will last.

 

[Mustrum_Ridcully]

The key components are really: Attack vs Defence, followed by duration determined by saving throw. That is 4E's method, but my mechanic skews the total spell duration look more familiar to players of 3E.

If your focus of the rule was that each round the target makes an ability based saving throw - against a DC set by the caster's abilities (rather than D&D 4's static 10), then yes, I think it could work. I believe some spells in 3.5 did already use such a mechanic (Hold Monster/Person, I believe).

I think D&D 4 wanted to get a way a bit from a kind of "double-reward" system that this mechanic tends to bring. If you have a good defense against the spell, it's not just likely that it will fail in the first place, it's likely it won't last long. And if the enemy is good with his attack, it's not just likely the ability succeeds in the first place, it's likely that the target won't be able to end the effect anytime soon.

So, attack vs defense was based on non-static values, but the save was moved to basically a static 50 % (or 55 %) success chance, so the double reward would no lo longer apply. It is less swingy this way, but if you like some more swinginess and/or want the double reward, your system should work. And I'd definitely prefer it about something like 1 round / level type of durations. (And it could also be applied for much longer-lasting spells. For example, a charm spell may allow not one save per round, but once each day, but certain actions could also trigger a save. Say, forcing someone to do something against his obvious best interest or nature, or attacking him, or something like that)

What it definitely fits is my idea to get rid of spell durations expressed in time units. I find this makes magic too... "mundane" to some extent, if a spell always lasts exactly 5 minutes or some such duration, and it also creates some tracking of time to a depth that you normally wouldn't need to do.

 

[delericho]

4E required you to roll a dice instead of tracking rounds. Instead we end up tracking many variations on spell duration. Is it until the end of the caster's next turn? The beginning of the caster's next turn? Sustain? Save ends? Until the end of the target's next turn? Until some other weird condition is met, such as 'when I move' or 'if I attack the wrong guy'?

I personally found this form of tracking to be more irritating than tracking rounds. I also find the 'save ends' mechanic to be a bit 'flat' in play. Firstly because it never really varied from needing a 10, and secondly because all too often the effect only lasted one round.

I certainly found all the tracking of things to be a pain! Every round it seemed that some new spell was taking effect, or ending, or there was a condition to apply or remove. It seemed that for everything they simplified, they rewarded themselves by adding a new layer of complexity in its place.

Conversely, my mechanic only requires you to write down a single number (the DC). This number doesn't change from round to round. Nor does it necessarily have to live in a game with many other potential 'spell ends' conditions.

You only have to track a single number if there's only one spell (possible spell-caster). Multiples mean multiple numbers to track.

Additionally, you only skip the bulk of 4e tracking if you also get rid of all the variations on spell duration... which is fine, but which works just as well with 4e if you switch everything to "save ends".

If the single-round durations are too short, a simple solution would be to increase the threshold on saves - '15' or '18' would seem good values, depending on how forgiving you want the system to be.

A simpler alternative might be to have:
Spell duration = Attack Total - Defence Total.

Avoid this like the plague - that subtraction is a killer. I'm often amazed at just how poor so many people are even at simple arithmetic, and as soon as the numbers go above 10, subtraction really seems to suffer.

 

[Minigiant]

Cannot we just do the damage route.

The Attacker rolls the damage and duration
The Target rolls the saving throws.

The cleric casts hold person and roll for the duration of 1d6+Wis rounds.
The orc rolls the Charisma saving throw to not be held and paralyzed.

 

[Mustrum_Ridcully]

Avoid this like the plague - that subtraction is a killer. I'm often amazed at just how poor so many people are even at simple arithmetic, and as soon as the numbers go above 10, subtraction really seems to suffer.

So true. I don't know why, but differences are really hard on me as well.

IIRC, computer algorithm are also the least precise when it comes to substractions.

Cannot we just do the damage route.

The Attacker rolls the damage and duration
The Target rolls the saving throws.

The cleric casts hold person and roll for the duration of 1d6+Wis rounds.
The orc rolls the Charisma saving throw to not be held and paralyzed.

Variable duration or fixed duration makes little difference from the handling at the table. But a save each round makes a difference - on the one hand, a bit more work-load, on the other hand, you don't need to track how many rounds you were going, and on the gripping hand, it makes the (player of the) target of the spell feeling more active as he gets to have something to do even if a spell is otherwise debilitating.

 

[Zustiur]

If your focus of the rule was that each round the target makes an ability based saving throw - against a DC set by the caster's abilities (rather than D&D 4's static 10), then yes, I think it could work. I believe some spells in 3.5 did already use such a mechanic (Hold Monster/Person, I believe).

I think D&D 4 wanted to get a way a bit from a kind of "double-reward" system that this mechanic tends to bring.

Indeed, that is part of the intended design. I like the idea that a character with good will defence will have less trouble breaking out of will-based spell effects. Perhaps another alternative is to take all 4 components, instead of trying to make do with three.
1) Defence Value
2) Attack Roll
3) Saving throws
4) Caster based DC
Where Caster based DC is a static number as per 3E. Basically 10+ability modifier+spell level (assuming that fits with 5E maths). This is more forgiving for two reasons - firstly it means you can find out the DC later if you don't note it down immediately, secondly, you won't ever have a DC based on a natural 20 to contend with.

Another piece of the puzzle that I left out of the original post is my assumption that 'spell caster disruption' be possible. Again, I didn't want to include this lest the post become too long and complicated, but my intention is that there be a way of preventing 'sustain'. I haven't worked out the nuts and bolts of it, but I am thinking it would play roughly like this:

NPC Cleric cast's hold person on the party fighter and sets the DC at 19. The fighter fails to save against the effect on his turn. Seeing this, the party wizard hits the cleric with Melf's Acid Arrow, causing X damage. The cleric must then take a concentration check of some sort in order to sustain the spell next round.

Exactly what rules are required to match my imagination of the scenario... I'm not certain yet.