D&D General Tweak to Multiclass Spellcasting Rules: Do You Think it Would Break the Game?

[Silam]

This helps some casting multiclasses, basically where there's wanted overlap in spell lists, and hurts others. That's should not be read as "it's balanced", it should be read as "some combos are too powerful" plus the corrolary "this will discourage even more some multiclass combonations". That second part is even more important when looking at all of the half-casters out there.

Fair point. It’s true that this proposed change does not impact all combos equally, and it is worth going over each one to assess that impact. That the impact is not equal is not a problem in and of itself, but it could be if some already too strong combination became even stronger, or vice versa. But if two medium-powered permutations where one is slightly stronger than the other switch place in the power hierarchy, while still being in the middle of the pack otherwise, I think that’s just the cost of doing business: you tweak things, things will get tweaked, and it’s not fundamentally a problem.

Also, the entire "what is weaker" goes away when multiclassing between classes with the same spellcasting stat, making this strictly stronger in many cases.

For combos where all classes share the same casting stat, the upcasting mechanic would remain unchanged compared to RAW (not stronger). What would be stronger in that case would be the higher level prepared spells for spells that are in the intersection set of these various spell lists. The relevant question then becomes: how many such spells are there in the intersection set? I am not going to crunch these lists right now (though I would be curious to do the exercise, and I might later…) but at a glance, we can look at a simple bucketing of stat to classes:

• Int: Arcane Trickster, Artificer, Eldritch Knight, Wizard (everything but the Artificer has 100% overlap here).
• Wis: Cleric, Druid, Ranger
• Cha: Bard, Paladin, Sorcerer, Warlock

As noted above, Int is where there is the greatest overlap, since the AT and EK literally have the Wizard list itself as their list. For those cases where the list is actually the exact same, I think there could be room for an additional rule to go along the other two proposed originally… and it could go like this:

When multiple classes share exactly the same list (such as AT, EK and Wizard) then those levels (including fractional ones) get added up and treated as levels of the same class (e.g., Wizard levels) not only for the sake of slots and max prepared spell levels, but also for cantrips known and unique spells prepared.

In effect, this is a slight nerf in versatility. For example, an EK 6 / Wizard 6 would get the slots of a Wizard 8 (as normal in the RAW). But in the RAW, this build would get 2 + 4 = 6 cantrips and 4 + 10 = 14 spells prepared, whereas in this house rule, they would get 4 cantrips and 12 spells prepared. Not a big nerf, but at least it’s not 100% better with nothing in exchange. It’s also a bit simpler than RAW to reason about what you’re supposed to have (whereas other combos are admittedly more complicated, but hey, we’ll take it!).

How this interacts with half casters (and 1/3 casters) who have higher level spells on their spell list that they might not get to for a while, but suddenly match other spell lists, is also an abusable power boost.

I am not sure here. Can you name an example of a spell off the top of your head which seems abusable?

All in all, it leaves multiple avenues to be significantly stronger than the current caster multiclassing rules. I'd rather use the existing mechanisms of feats to represent things like Wizards with Metamagic or Sorcerers who have improved their casting under the narrative of study.

I totally agree with the desire to lean on feats… I’m a big fan of feats, and actually prefer that mechanic over multiclassing (in part because I find the 5e design of multiclassing to be clunky). My only concerns there are that 5e provides too few feats, and the feats which accomplish what I’m trying to get to in this thread (Metamagic Adept, Ritual Adept) are not that good. But, certainly, I see a lot of potential for new feats (or feat chains, a design I’m a big fan of).

Thanks for the feedback! Please let me know what you think, if you care to keep this discussion going!

 

[Blue]

Fair point. It’s true that this proposed change does not impact all combos equally, and it is worth going over each one to assess that impact. That the impact is not equal is not a problem in and of itself, but it could be if some already too strong combination became even stronger, or vice versa. But if two medium-powered permutations where one is slightly stronger than the other switch place in the power hierarchy, while still being in the middle of the pack otherwise, I think that’s just the cost of doing business: you tweak things, things will get tweaked, and it’s not fundamentally a problem.

I agree that it's not fundamentally a problem, but considering that casters are already generally at the higher end of the power scale, a small boost of power is a much larger issue than a small boost of power would be for a below average class like the Ranger.

So increase to top-tier choices can't be handwaved as part of the 'cost of doing business'. If something is already more powerful than average, we need to be very careful about increasing it more. That said, multiclassing pure caster generally isn't the highest power choice, which means we can accept it as you mention.

But we still need to look for those cases it is and make sure we don't inadvertently make them too much. Paladin/sorcerer may be an example of one already considered strong this would boost. Bladesinger/Fighter (EK) with full overlap in spell lists. But this also opens up combos previously not considered too powerful because the casting stats didn't line up. When you have cases such as a character picking buff/healing/support/utility spells where the casting stat doesn't matter as much you can get some interesting cases.

For combos where all classes share the same casting stat, the upcasting mechanic would remain unchanged compared to RAW (not stronger).

I don't believe this to be correct. Where they share a casting stat all of the weakness go away, but the strengths of overlapping spell lists remain. Unless there is a corner case where two classes share the casting stat and have zero spell list overlap, this must be stronger than RAW, since it has an advantage and no disadvantage.

When multiple classes share exactly the same list (such as AT, EK and Wizard) then those levels (including fractional ones) get added up and treated as levels of the same class (e.g., Wizard levels) not only for the sake of slots and max prepared spell levels, but also for cantrips known and unique spells prepared.

In effect, this is a slight nerf in versatility. For example, an EK 6 / Wizard 6 would get the slots of a Wizard 8 (as normal in the RAW). But in the RAW, this build would get 2 + 4 = 6 cantrips and 4 + 10 = 14 spells prepared, whereas in this house rule, they would get 4 cantrips and 12 spells prepared. Not a big nerf, but at least it’s not 100% better with nothing in exchange. It’s also a bit simpler than RAW to reason about what you’re supposed to have (whereas other combos are admittedly more complicated, but hey, we’ll take it!).

Versatility: Nerf in number of spells, large increase in being able to prepare higher level spells, large increase in which spells because it's going from small EK selection to your full spellbook. Rating: Boosted.
Power: able to prepare higher level spells than RAW. Rating: Boosted.
Final: Large boost.

I'm using the 2024 rules here, same basic idea for 2014. RAW, the EK6/Wiz6 would be able to prepare three 1st level spells from a very short list, and the Wizard would be able to prepare up to ten spells of level 3. Combined they can do 12 spells of up to level 4. So a decrease of two spells, but not only do we have access to level 4 spells for the wizard, we have access to levels 2 through 4 that the EK would not have had. And those spells can come from the spellbook (or the EK chosen spells), giving both more power and more versatility than the EK spells.

Increasing the level of the spells chosen for both classes and the selection of spells chosen by the EK is very hard to justify as a net reduction in versatility even with 12 instead of 14 spells.

 

[TigerDude]

I would say this adds complexity for something that no one will likely play anyway. The trouble with rules like this is that there will be something that someone wants to do and now they can't (or won't). Multiclass casters typically are weaker than single class anyway, multiclass tends to be either for a speciic concept or to make up for something (the warlock/sorc/bard/pally combos, which this wouldn't touch).

 

[TigerDude]

I would say this adds complexity for something that no one will likely play anyway. The trouble with rules like this is that there might be something that someone wants to do and now they can't (or won't). Multiclass casters typically are weaker than single class anyway. Multiclass tends to be either for a specific concept or to make up for something (the warlock/sorc/bard/pally combos, which this wouldn't touch).

 

[Silam]

I agree that it's not fundamentally a problem, but considering that casters are already generally at the higher end of the power scale, a small boost of power is a much larger issue than a small boost of power would be for a below average class like the Ranger.

Agreed, this needs to be handled carefully to avoid making the most powerful builds even more powerful.

But we still need to look for those cases it is and make sure we don't inadvertently make them too much. Paladin/sorcerer may be an example of one already considered strong this would boost.

So... I have not conducted the full analysis yet, but I have looked at this particular combo, and here are the numbers, according to 5eT while including WotC books only, and excluding 2014 content that's been reprinted in 2024:

• Paladin: 56 spells
• Sorcerer: 234 spells
• AND intersection of Paladin and Sorcerer lists: 5 spells, which are
  • Banishment (level 4)
  • Daylight (level 3)
  • Detect Magic (level 1)
  • Dispel Magic (level 3)
  • Magic Weapon (level 2)

So there is very little overlap, and Detect Magic would not be affected by the proposed change since it's 1st level and therefore available regardless. That leaves us with 4 spells in the overlapping set, which could be gotten at an earlier level with the proposed homebrew rule. Dispel Magic might be the most generally useful spell in that set, and it's certainly nice to access it sooner, but it does not feel like a dramatic power boost to me. So I don't think that particular combo is significantly stronger with the proposed change. In particular, the usual motivation for a Paladin/Sorc build is to upcast the various smite spells beyond what a straight-class Paladin ever could, and that particular mechanic remains unchanged, while also probably being significantly stronger than getting Dispel Magic a level or two sooner (though never sooner than a straight-class Sorcerer would).

Bladesinger/Fighter (EK) with full overlap in spell lists. [...] I don't believe this to be correct. Where they share a casting stat all of the weakness go away, but the strengths of overlapping spell lists remain. Unless there is a corner case where two classes share the casting stat and have zero spell list overlap, this must be stronger than RAW, since it has an advantage and no disadvantage.

What I meant when I said that it's no stronger than RAW was specifically about the upcasting. The upcasting is no stronger. Other aspects are stronger, indeed, and so you can make the valid case that overall it is stronger, and I agree.

Versatility: Nerf in number of spells, large increase in being able to prepare higher level spells, large increase in which spells because it's going from small EK selection to your full spellbook. Rating: Boosted.
Power: able to prepare higher level spells than RAW. Rating: Boosted.
Final: Large boost.

So if that is still too strong overall, we can look for further compensation mechanisms... leaning further into the versatility nerf, we might say that if a given spell level could not be prepared by any single class within the build, then all of the classes which contribute the necessary levels need to spend a preparation slot. In other words, if we come back to the example of the Wizard 6 / EK 6 who has an effective caster level of 8 and can therefore prepare 4th level spells, they would need to pay 2 prepared spells to prepare a 4th level spell (in addition to the other proposed change that they would get the number of prepared spells of a Wizard 8, which is 12, rather than the sum of those of a Wizard 6 and EK 6, which is 10 + 4 = 14). Unfortunately, this makes spell prep accounting more complicated, which is something to consider in terms of ergonomics... but would it be a big enough nerf in terms of balance?

I'm using the 2024 rules here, same basic idea for 2014. RAW, the EK6/Wiz6 would be able to prepare three 1st level spells from a very short list, and the Wizard would be able to prepare up to ten spells of level 3. Combined they can do 12 spells of up to level 4. So a decrease of two spells, but not only do we have access to level 4 spells for the wizard, we have access to levels 2 through 4 that the EK would not have had. And those spells can come from the spellbook (or the EK chosen spells), giving both more power and more versatility than the EK spells.

I see what you mean, but personally I don't think that this specific aspect is that big of a deal. Theoretically, 5e's neo-Vancian prep system allows a Wizard 6 to prepare 10 spells of level 3, and none of lower levels (assuming they had found this many spells from copying spellbooks and scrolls). By doing so, they would get a ton of versatility in their most powerful slot level (of which they possess 3, and can get 1 back with Arcane Recovery), but they would forfeit their 1st and 2nd level slots in exchange. No Wizard in their right mind would do that. It's a fine strategy to memorize a few more high level spells than you have slots for, meaning that you accept that you're guaranteed to be unable to cast all of your prepared spells of that level, but it would not be good to overdo this strategy. And since 5e places a big emphasis on upcasting, it is usually the reverse that we see: many low-level spells remain valuable to prep since they can be useful across a wide range of slot levels.

All that to say... using vanilla rules, a Wizard/EK multi-class build would be forced to use their EK prepped spell on lower level options, but they would probably want to do that regardless, and they may choose to prep fewer or even none at those levels via their Wizard prepped spells. If that build suddenly had a single pool of prepped spells drawing from a single list of known spells, rather than two distinct pools and lists, I think ultimately their low-level spell choices would still look quite similar. In other words, I don't think this requires burdening the proposed rules with any additional restrictions, since the extra flexibility of prepping fewer low-level spell is akin to "giving them enough rope to hang themselves with" rather than an actual benefit.

I would say this adds complexity for something that no one will likely play anyway.

I find this assertion confusing... but please do let me know if I'm missing something!

On one hand, if no one will play that anyway, then who cares about its complexity? It affects no one. Sure, it might not be worth the paper or digital bits it's printed on, but it has no practical downside if no one's choosing to be affected by it...

On the other hand, if no one wants to play a given multi-class combo, then I wonder if that is a symptom of the game design being unbalanced? Let's say Wizard/Sorcerer is one of those combos no one would likely play (not saying that's what you said, maybe you were thinking of some other combo, but it's just an example...) then why? Are they not playing it because it's not powerful enough (i.e., the game is unbalanced)? Or because thematically it is disagreeable (e.g., like a Barbarian/Wizard, where we could make the case that Rage being incompatible with spellcasting is a feature, not a bug, in order to retain the theme of these classes)?

The trouble with rules like this is that there will be something that someone wants to do and now they can't (or won't).

Are you saying this proposed homebrew rule makes some specific combos significantly weaker, such that they would be viable under vanilla rules but no longer viable in the homebrew? If so, do you have any specific examples where that seems to be a problem? Or if not, what did you mean there...?

Multiclass casters typically are weaker than single class anyway, multiclass tends to be either for a speciic concept or to make up for something (the warlock/sorc/bard/pally combos, which this wouldn't touch).

Right... multi-classing spellcasters is typically seen as a big no no, from an optimization perspective. Do you think there is value in that being the case? Would there be a problem in making multi-classed spellcasters come closer to the power of single-classed spellcasters?

---

Thanks for all this feedback, folks! Really appreciate it!

 

[Blue]

I see what you mean, but personally I don't think that this specific aspect is that big of a deal.

This here is where I think our big disagreement lies.

I see the multiclassed wizard getting higher level spells known so that they can use their highest level slots to their fullest as a big deal. I see converting EK spells from a short list of 1st level to 1-4th from the wizard's spellbook to be a significant boost for the EK casting as well. I'm stymied that this is "not a big deal" while your calibration calls going from 14 spells prepared to 12 spells prepared big enough to be considered a nerf.

Higher levels spells known are one of the chief balance points of multiclassing. Higher level spells are simply more powerful than a lower level spell upcast, if an upcast option is even available.

Theoretically, 5e's neo-Vancian prep system allows a Wizard 6 to prepare 10 spells of level 3, and none of lower levels (assuming they had found this many spells from copying spellbooks and scrolls). By doing so, they would get a ton of versatility in their most powerful slot level (of which they possess 3, and can get 1 back with Arcane Recovery), but they would forfeit their 1st and 2nd level slots in exchange. No Wizard in their right mind would do that. It's a fine strategy to memorize a few more high level spells than you have slots for, meaning that you accept that you're guaranteed to be unable to cast all of your prepared spells of that level, but it would not be good to overdo this strategy. And since 5e places a big emphasis on upcasting, it is usually the reverse that we see: many low-level spells remain valuable to prep since they can be useful across a wide range of slot levels.

Sure, but that's not the case under discussion. The case under discussion is "I have 1st through 4th slots, but can only use my 4th level slots for upcasting" vs. "I have a good selection of prepared spells for all of my slots".

Because the second of those is most assuredly a more powerful choice.

 

[Silam]

This here is where I think our big disagreement lies.

I see the multiclassed wizard getting higher level spells known so that they can use their highest level slots to their fullest as a big deal. I see converting EK spells from a short list of 1st level to 1-4th from the wizard's spellbook to be a significant boost for the EK casting as well. I'm stymied that this is "not a big deal" while your calibration calls going from 14 spells prepared to 12 spells prepared big enough to be considered a nerf.

I'm sorry, definitely did not want you to feel stymied by my comments. I think I may not have been sufficiently precise, and that this resulted in the same kind of misunderstanding as in my other post... I did not mean to say that the proposal overall is not a big deal. It certainly is a big deal to have access to a higher maximum level of prepared spells. My goal is for the overall tweaked system to still feel balanced, or as close to balanced as possible (it is very difficult, admittedly). But where I think we are having misunderstandings, here and before too, is that I'm trying to decompose the overall system into its various individual changes, and assess each one individually. When I said I thought it's not a big deal, I was talking very narrowly about this blurb:

I'm using the 2024 rules here, same basic idea for 2014. RAW, the EK6/Wiz6 would be able to prepare three 1st level spells from a very short list, and the Wizard would be able to prepare up to ten spells of level 3. Combined they can do 12 spells of up to level 4. So a decrease of two spells, but not only do we have access to level 4 spells for the wizard, we have access to levels 2 through 4 that the EK would not have had. And those spells can come from the spellbook (or the EK chosen spells), giving both more power and more versatility than the EK spells.

Please let me try to take another stab at it and be more precise in my response...

Right off the bat, I need to acknowledge that having access to level 4 prepared spells is stronger than not having access to them. Of course it is. I'm definitely not trying to shove that under the rug nor to sneak a fast one past the balancing gods

Let's look at the comparison concretely, for the EK6/Wiz6 build we've been discussing. We can explore the space endlessly but I'll try to narrow down the parameters for the sake of simplifying the discussion, and I'll assume this is a gish build since it contains EK, and therefore I'll further assume that the Wizard side is Bladesinger. We can redo the analysis for other combos, but this is just to help inform the hypothetical spell choices (e.g., if it were built as a Batman-style control-focused Wizard, it would certainly make different choices).

As we have mentioned previously, this build gets:

• With vanilla rules: 4 x 1st level spells from the EK side, and 10 prepared spells of up to 3rd level from the Wizard side.
• With the proposed homebrew rules: 12 spells of up to 4th level (same as a Wizard 8).

So, I think an important question is: are there fewer than 4 x 1st level spells that this build would like to prepare? Why is this relevant? Because if that build is going to want 4 x 1st level spells anyway, then the prep breakdown by level imposed by the vanilla rules is going to be organically carried over into the homebrew rules, and therefore the extra flexibility of the homebrew rules which would give the option of preparing fewer than 4 x 1st level spells won't actually be exercised (hence my comment that this specifically would not be a big deal)...

I would think this build definitely wants Shield and Mage Armor at a minimum (if the Wizard subclass was not Bladesinger, then MA could be ditched in favor of wearing an actual armor, but that is why I stated the assumptions upfront...), so that's 2 out of 4. If that's all they want from the 1st level, then yeah, the homebrew rules' flexibility are a benefit. But let's look at some other 1st level spells that might be relevant for a gish build:

• Absorb Element for another Reaction option besides Shield, very appropriate for a melee gish, and can be upcasted...
• Jump or Expeditious Retreat, as they might possibly want some mobility buff (probably not both, but maybe one of them... Jump does not require concentration and can be upcasted to affect allies too).
• False Life can be pretty useful for a gish, especially since it does not require concentration and can be upcasted.
• Feather Fall is a good utility spell for any Wizard, though its usefulness depends on whether the DM is devious in laying down traps...
• Sleep (or perhaps even Color Spray if you want to lean into the close-range aspect of the build) for some control option.
• Silvery Barbs is widely considered overpowered... might not be accepted at your table due to that, but if it is accepted and you have no qualms then you're probably taking it...
• Thunderwave is a decent "of cr*p" panic button spell if you've over-extended your stay on the frontlines.

You definitely don't want all of the above as that would take up most of your prepared spells, but I would think there are enough decent options there to want at least 2 of the above (even while constraining for a gish build, which means I've left out a bunch of ranged attacks and many other good spells). You could make do with just Shield and MA and give up Absorb Element, Feather Fall, etc... but would you really?

That was what I tried to convey when I said I didn't think it's a big deal. I was talking specifically about the low end, not the high end. Under vanilla rules, the character must have at least 4 of their 14 total prepared spells be the 1st level spells coming from the EK list, that is true and it is indeed a constraint. But under the proposed homebrew rules, even though that constraint is lifted, I think it's fairly unlikely that the character would prepare fewer than 4 x 1st level spells out of the 12 total they can prepare. And if they do choose to exercise that flexibility, they're leaving a lot of good stuff on the table, which is arguably not that great for them overall.

The EK6/Wiz6 possesses 2 x 4th level slots, and cannot regain any from Arcane Recovery (that is the same in vanilla rules and in the homebrew). In the proposed homebrew, they would have 12 total prepared spells of 1-4th level, so let's say they prep a single 4th level spells, and let's say that privilege costs them twice the amount of prepared spells (as in my latest proposed tweak), that would mean they have only 10 left for levels 1-3... If they wanted to prep 2 x 4th level spells, that would leave them only 8 prepared spells for levels 1-3. It's getting pretty squeezed... I, at least, feel like I would have a hard time choosing. If I had the choice between vanilla and these homebrew rules while using that particular build, I don't even know if I would actually want to pick the homebrew anymore ... basically we're looking at a choice between 1 x 4th level prepared spell + 10 others of levels 1-3, versus 14 prepared spells of levels 1-3... Considering how many 5e spells have good upcasting options (and that both builds have the same slots available regardless), I don't know if there is a single 4th level spell that would be worth my while here (CME comes to mind, but CME only outperforms the damage of Shadow Blade starting from 6th level upcast, so it's a wash before then; if you have a good magic weapon and rarely dwell in dim light dungeons, then CME could be better).

It might still be stronger overall though. I don't deny it, but I'm not sure it's that much stronger. We need to consider that we're talking about a hypothetical 12th level character with a fairly suboptimal build. A 12th level EK or a 12th level Wizard both probably outperform this 6/6 build with or without the homebrew rules. So maybe this is a bit of a strawman argument, in that sense, and we need to look for other multi-class builds that already come closer in power to single-class builds of the same level, and see how it compares over there...

I hope that makes sense...

What do you think?

 

[Fanaelialae]

I wouldn't implement it this way.

The penalty basically doesn't exist for multiclass combos that share the same stat (such as Sorlock, which is already a strong combo), while it makes it a semi-trap option for the combinations I assume it is intended to fix. For example, a sorcerer2/wizard3 probably shouldn't bother with Fireball, because the save DC will suck (which is why I refer to this as a trap). On the other hand, if they select a spell that doesn't depend on the casting stat (ie, Tongues) there's literally no cost at all. So IMO it's a straight buff with no penalty, assuming the player has sufficient wherewithall to avoid the trap options.

What I would do is make it a feat. Now there's always a cost, regardless of whether you're Sorlock or Sorcerer/Wizard. Keep the part about the spell list overlap but drop the part about using the weaker stat. Then playtest it. Maybe one feat allowing to use your combined spellcasting levels to learn spells works. Maybe it should require one feat per +1 spell level (meaning a sor10/wiz10 needs 4 of these feats to learn 9th level spells). I suspect the former should work, while the latter is overkill, but the only way to know for sure it to test it out.

 

[Silam]

I wouldn't implement it this way.

The penalty basically doesn't exist for multiclass combos that share the same stat (such as Sorlock, which is already a strong combo), while it makes it a semi-trap option for the combinations I assume it is intended to fix. For example, a sorcerer2/wizard3 probably shouldn't bother with Fireball, because the save DC will suck (which is why I refer to this as a trap). On the other hand, if they select a spell that doesn't depend on the casting stat (ie, Tongues) there's literally no cost at all. So IMO it's a straight buff with no penalty, assuming the player has sufficient wherewithall to avoid the trap options.

What I would do is make it a feat. Now there's always a cost, regardless of whether you're Sorlock or Sorcerer/Wizard. Keep the part about the spell list overlap but drop the part about using the weaker stat. Then playtest it. Maybe one feat allowing to use your combined spellcasting levels to learn spells works. Maybe it should require one feat per +1 spell level (meaning a sor10/wiz10 needs 4 of these feats to learn 9th level spells). I suspect the former should work, while the latter is overkill, but the only way to know for sure it to test it out.

That's a great idea! I hadn't thought of introducing the mechanic via a feat... it's an elegant way to add a cost while also removing the complexity of the weaker stat mechanic, which as you said is inevitably going to have an uneven effect.

If the feat gave just that feature, do you think it could be fine as a half-feat that also gives +1 to a mental stat? Or would that be too strong? I think I would prefer keeping it a single feat, otherwise the cost gets pretty high and I'm not sure anyone would be willing to pay that...

 

[Fanaelialae]

That's a great idea! I hadn't thought of introducing the mechanic via a feat... it's an elegant way to add a cost while also removing the complexity of the weaker stat mechanic, which as you said is inevitably going to have an uneven effect.

If the feat gave just that feature, do you think it could be fine as a half-feat that also gives +1 to a mental stat? Or would that be too strong? I think I would prefer keeping it a single feat, otherwise the cost gets pretty high and I'm not sure anyone would be willing to pay that...

It might be too strong for a +1. The way D&D stats work is weird (since 3e). If you have a 16 and add a +1, it doesn't do anything immediate. If you have a 17 and add +1, it has the same effect as if you had a 16 and added +2. If you have an odd score, you basically get the stat boost and the benefits. That's fine for weak feats like Actor.

Is this a weak feat? I'm inclined to say that in the situations where it is applicable, it's pretty strong, and therefore that it shouldn't have a +1. IMO, if I were playing something like a sor2/wiz3, I would pay a feat to have access to fireball. Sure, missing the opportunity to boost my primary casting stat would sting, but now I have a wizard subclass, meta magic, and fireball. All for delaying that +1 on my spell DCs. That's not a bad deal IMO.

If you think it needs a little something extra, I would add something other than a +1 stat.