View source for Talk:Mana-Based Spellcasting (3.5e Variant Rule)

== Ratings ==
{{Rating |rater=Enigma
|rating=favor
|reason=Would love to try this, its well thought out and more flavorful than spell slots.
}}
{{Rating |rater=Spanambula
|rating=like
|reason=Basically what others have said. Still needs some official adapting for 1/2 casters, but this just feels so much more fun than pure Vancian casting.
}}
{{Rating |rater=Elohim
|rating=favor
|reason=I find the basic idea to be very adaptable to most classes, and many creation systems. i think a unified base for the varied classes is good for ease of understanding, as well as modification. I also see the variant as making other mechanics modification more streamlined and easy to implement.
}}
{{Rating |rater=Wildmage
|rating=favor
|reason=love it, was a little unsure about the power lvl but after having a player using it for a couple of months i a campaign im in love
}}
{{Rating |rater=DanielDraco
|rating=like
|reason=I've played with this rule before, and I have to say I do like it quite a bit more than Vancian casting. However, it is inescapable that spellcasters are, on the whole, already overpowered such that other classes are strictly inferior, and this only adds to their capabilities. As such it really cannot be considered balanced unless it is combined with rules that bring casters in line with other classes. On its own, for that reason, I would see it as being incomplete. It's fun, and workable, and that earns it a 3/4, but I won't give it a 4/4 because it widens the gap.
}}

{{Rating |block=RC 
|rater=Rithaniel
|rating=favor
|reason=This variant rule does so much correct that has been failed at in the past, that it makes my head spin. Right off the bat, let me simply state that the idea behind this variant rule is extremely clever, and should not be forgotten. Having said that, this system works solves a blaring problem that I have always had with point based systems: you have thousands of spells per day at higher levels. With this system, instead of gaining a bigger pool, you have a relatively set pool, but simply take less out of it at a time. This, I like in it's ingenuity, and must simply tip my hat to it. In addition to this, the triple recovery mechanic makes for a very interesting time getting your spells back. All in all this variant rule earns a big thumbs up from me. Favored.}}

{{Rating
|rater=Ganteka Future
|rating=Like
|reason=Hooray for variant rules that people might actually use. It isn't campaign specific, or play style specific or... well, it's general enough to be used by anyone who dislikes the core spellcasting style and would prefer to try something else. Easy to understand. Easy to implement. Heck, you could implement it mid-campaign if you really wanted without messing with the story of the game at all. That's cool. 
}}

{{Rating
|rater=Jota
|rating=like
|reason=Although there is the issue of wizards becoming more akin to sorcerers in this mana-based system (something I can hardly consider a good thing, with wizards being what they are) I doubt such a change truly makes a big difference given that a wizard's power lies in his versatility of spell selection rather than in his volume of usable spell levels per rest. With that said, this system is very user friendly and seems to scale well at face value. The fatigue system is a nice reflection of what magic is often seen as in less crunch-based descriptors (read: books) and allows the character to push himself, something which isn't provided for under the current system. The flexibility of the system is a great plus as well, although I am not sure spell strength scales in proportion with mana costs. That, however, is more a problem with spells themselves than with the system. Furthermore, this system actually caps the spellcaster's strengths to some extent in that the number of high level spells that can be cast is more limited than as standard. It also does screw with some feats a little bit, but none that are integral to the game or in ways that cannot be accommodated for with a little ingenuity. Between the added nuances of the system, the expanded versatility available from lower level spell slots, and the restrictions on higher level spells, this may actually be an improvement on the traditional spellcasting system, as opposed to just a variant. }}

{{Rating |block=RC
|rater=TK-Squared
|rating=favor
|reason=Something something dark side... I also like it.
}}

{{Rating |block=RC |rater=Sam Kay |rating=favor |reason=What more is there to say? I've never really liked spells per day...}}

{{Rating|rater=Leziad|rating=favor|reason=Way better than either the current spell-per-day mechanic or he UA spell point system.}}

{{Rating
|rater=Ghostwheel
|rating=oppose
|reason=This variant has a few problems that make it less usable in my mind; first, without changes it shares a problem with the Vancian magic system in the 15-minute workday, where the mage (or sorcerer) might use up their highest level spells and then force the party to call for a rest. An attempt is made to address this problem by suggesting a change to recharge times, but if you have a character recharge in a much shorter space of time (to account for possible encounters spaced shortly one after another) you can fall prey to allowing the mage to cast spells that are problematic virtually at will. To contrast, the [[Publication:Unearthed Arcana/Recharge_Magic|Recharge Magic]] variant has a built-in protection against specifically problematic spells that the DM still wants to keep in the game, but wishes to limit--a specific recharge time. The DM is free to change the recharge time of specific spells, a mechanic already inherent in the system, allowing them some control of spells that become problematic when used often in a session. This variant has nothing against it, and while I usually enjoy point-based systems, the downsides of this one by far outweigh any upsides it might have in my mind.}}

{{Rating|block=RC |rater=Sulacu |rating=like |reason=This variant is simple, usable, and does away with most of the tedium of playing a wizard (keeping track of spell slots), which is enough to warrant a like, and make me want to try it for myself. However, it is also rather incomplete, making mention only of wizards and sorcerers. I think that, for this article to be sufficiently complete, it should incorporate tables for every core spellcasting class; that is, wizard, sorcerer, cleric, druid, bard, ranger and paladin. Provided the full divine spellcasting classes could make use of the mage table, this would mean adding tables for partial spellcasting (0-6th and 1st-4th respectively). That way, it will be usable in pretty much any game with fairly little work on the DM’s part. A mention of prestige classes with independent spellcasting, or multiclassing into two or more spellcasting classes, could also help the completion of the article. Still, this is overall a great article.}}

{{Rating 
|rater=Foxwarrior
|rating=favor
|reason=Not only is it less accounting than usual for a mana-based system, it also fits most written and visual fantasy magic systems better than Vancian or other mana systems do.
}}

{{Rating 
|rater=Franken Kesey
|rating=favor
|reason=One of the things which made spell-casters suck was being forced to only have a certain number of spells of each level. But if they had no option to have mostly 3rd level spells, instead of wasting space with used low level spells. This flaw in the core system has turned many to preferring psionics over spellcasting. However, spell-casting has so much more versatility than psionics. Thus this variant, keeps the versatility, while having a balanced way to fix the rules. Furthermore This variant deviates from adding costs based on level -- which is excellent.
}}

== Design Question ==

Before I rate this Surgo, I'd like to know if you think spell power follows the same growth pattern as character level. This is a how things actually are question, since you built this for current spells, not a how should they be question. Specifically: do you think that 1 spell of level X+1 = 2 spells of level X? If not, is 1 spell of level X+1 better than or worse than 2 spells of level X in general? - [[User:Tarkisflux|TarkisFlux]] 22:27, December 16, 2009 (UTC)

:I just noticed this. Unless you're circumventing the round down rule, a level 1 mana caster only recovers strain from sleeping since half their level rounds to 0. Is this intentional? Similarly, the half level rule means that at level 8 onward you only need 7 hours of rest to fully recover and decreases consistently to only 4 hours at level 20  (assuming 20 or less in the casting stat at start, level bonuses to casting stat, and BoG item increases). Is that recovery period decrease intentional as well?

::It's not intentional, but it might be a feature -- I'll have to think about that one. As for the "how things are" question -- neither. Since this is a "how things are", I don't think that question has an answer -- you're asking me to quantify the difference of a level X and X+1 spell, and there are so many spells and so many variables that that just isn't possible and I think it would be a mistake to attempt to do that and then build a system based off of that for balance. I also don't think it's important to the balance of the overall system, which is similar to (but not quite) the spells-per-day numbers of a spellcaster's higher levels of spells. [[User:Surgo|Surgo]] 00:03, December 17, 2009 (UTC)

:::That last bit actually answers my question in a bizarre way. <s>I'll hold off on favor until you decide if that's a now-intentional feature or if you're going to change the recovery mechanic somehow.</s> - [[User:Tarkisflux|TarkisFlux]] 01:19, December 17, 2009 (UTC)

::::Not a feature, I'm editing it to make level 1 a special case. [[User:Surgo|Surgo]] 01:24, December 17, 2009 (UTC)

:::::Blah, edit conflicted. I realized it wouldn't change my thoughts anyway. - [[User:Tarkisflux|TarkisFlux]] 02:06, December 17, 2009 (UTC)

== Recovery Mechanic ==

Something I forgot to include in this page, which would probably change Tarkis's favor, is that the recovery mechanics are designed to be pluggable. You have three of them here -- fatigue-based casting; recharge magic; and the sleep-eight-hours for those dinosaurs who still believe in the 4-encounter workday. They can, and sometimes are, plugged in and out depending on what you're looking for. I just need to find a way to fit this text in to the overall narrative. [[User:Surgo|Surgo]] 03:06, December 17, 2009 (UTC)

:Yeah, that probably would. Tweaking that resolves all of the spell points issues I think you even can resolve without making serious changes to spells. If you're considering tweaking these anyway, I should point out that (barring spell buffs) your half level or cast stat mod, whichever is less, only matters at 20 given the same stat adjusts I used above. Quarter strain or cast stat mod, whichever is less, is a little more consistant but might fail some of the other adjusts you have in mind. - [[User:Tarkisflux|TarkisFlux]] 03:34, December 17, 2009 (UTC)

::I'm having a really tough time working the extra words about the pluggable recovery mechanic into the page's narrative. Hrm... [[User:Surgo|Surgo]] 03:43, December 17, 2009 (UTC)

:::There, a footnote should do okay -- no added length. [[User:Surgo|Surgo]] 04:00, December 17, 2009 (UTC)

== Tarkisflux's Criticism ==

Reading Tarkisflux's favor, I thought overnight about the criticism of being able to cast more as you go up in level. I personally believe that is a feature (and why it's designed that way), but I can understand why some people wouldn't like it. I'm thinking of adding in a second footnote, or part of the first, an example recovery mechanic that replaces the normal recharge with a recharge that's static based on caster level. What does everyone else think? [[User:Surgo|Surgo]] 14:30, December 17, 2009 (UTC)

:I think that since you intend people to adjust to taste you might be well served by staying away from specifics. A short discussion on what adjusting each variable does to a caster might get you there more effectively. For example, it might be nice to spell out that increasing or decreasing the strain pool doesn't have a strong impact on how many spells you can cast in a day if the recharge mechanic is turned up to high, but it does impact how many spells they can cast in a series of encounters before they need to recharge. And that the recharge amount limits total day castings more strongly than strain limit unless it's fairly low. Or that the refresh at night thing being eliminated only impacts the game at low levels or if recharge is turned down. So I guess some higher level interaction talk to better equip people who don't think about this stuff all of the time to make the adjustments they want. - [[User:Tarkisflux|TarkisFlux]] 23:16, December 17, 2009 (UTC)

== So basically.. ==

It's just like Ledgendary Quest, with some minor modifications. It was a good idea 7yrs ago and it's still a good idea. I think I like LQ's way of doing it though, no free spells for one thing. =P http://legendaryquest.com/ If you haven't heard of it, although given the similarities.. [[Special:Contributions/115.130.7.253|115.130.7.253]] 23:34, January 16, 2010 (UTC)

:Yes, I have never heard of it. I play very, very few systems. [[User:Surgo|Surgo]] 23:38, January 16, 2010 (UTC)

== Epic ==

How do you handle epic casting? Just keep going the same way? So 4th lvl spells are free at 23rd lvl, etc.
And how are epic spells cast? If at all. [[Special:Contributions/115.130.7.253|115.130.7.253]] 23:38, January 16, 2010 (UTC)

:You don't gain any new spell slots in epic anyway, so I don't see a convincing reason to extend the table to get your spells even cheaper (they're already technically getting cheaper because your mana keeps going up). 10th level slots, 11th level slots, etc., should cost maybe 12 mana. [[User:Surgo|Surgo]] 23:40, January 16, 2010 (UTC)

::If you're taking suggestions, I'd recommend what the OP did and just keep going up the table -- ''if'' you actually ''have'' those higher-level spell slots, from taking the [[SRD:Improved Spell Capacity|Improved Spell Capacity]] feat or suchlike. So if you took 5 levels of Improved Spell Capacity, for example, then your highest-level spell slot would be 14th and so you'd be throwing 9th-level spells for 0 Strain. But if you haven't taken the feat and don't have those higher-level spell slots, then you don't get the cost break. --[[User:Chuckg|Chuckg]] ([[User talk:Chuckg|talk]]) 16:27, 9 August 2016 (UTC)

== What about Bards? ==

That's my question, what about the other classes? like bard, duskblade, hexblade, artificer, etc.? Is there some sort of special equation used to figure out what their strain/mana cost would be? I'm wondering b/c I think this is the <u><b>best</b></u> spell casting alternative, and I would like to apply it to some other classes to see how they look.
[[Special:Contributions/98.28.174.92|98.28.174.92]] 13:36, April 23, 2010 (UTC)

:Every full caster should have the mana cost given in the table. There needs to be one made for non-full casters, like Bard -- I'll get on that. As far as spell-like abilities go, I have not attempted to fit them to the system. [[User:Surgo|Surgo]] 18:42, April 23, 2010 (UTC)

::Hey:  Could you make the partial spellcasters soon?  That would be really cool.  --[[User:Parakee|Parakee]]<sup>[[User Talk:Parakee|Talk]]</sup> 13:21, 29 May 2011 (UTC)

== What about Multiclass Spellcasters? ==

How does this variant deal with characters that have more than one spellcasting class? Say you're a wizard/sorcerer, you have multiple strain tolerances and accumulate different types of strain, how do they intermix? 

Will strain from all your spellcasting classes count against your highest strain tolerance? If this is the case, that means that this variant may cause trouble with multiclass characters, especially if prestige classes like the mystic theurge are involved. A dual-spellcaster build will inevitably have lower base casting stats, therefore lower strain resistance than a regular wizard or cleric. As such, if both its modes of spellcasting count towards the same strain tolerance, a build like this wouldn't be able to make the most of its versatile spellcasting using this variant rule, as it would quickly overtax itself, especially at low-to-mid levels.

However, giving a character multiple independent strain cost tables and strain tolerances would cause trouble with this fairly simple rule system that assumes a single source of strain. 

For the sake of completion, I think this article should mention the possibility (or impossibility) of dual or multi-spellcasters, and possibly include rules to make them work (or exclude them).  --[[User:Sulacu|Sulacu]] 21:28, 21 September 2011 (UTC)

:Ouch, now this is a toughie. Requires thought. [[User:Surgo|Surgo]] 02:29, 22 September 2011 (UTC)

:: I suggest: Determine Strain Tolerance, Strain Cost, and Strain Recovery separately for each class. Only cast as one class in any given round. Fatigue and Exhaustion apply to all classes regardless of how they occurred.--[[User:Ideasmith|Ideasmith]] 02:43, 22 September 2011 (UTC)

::: Additional word on this would be nice; as it is, you can dip a level into any other class to gain an easy +14 or so to your strain tolerance level. --[[User:Ghostwheel|Ghostwheel]] ([[User talk:Ghostwheel|talk]]) 18:38, 8 October 2013 (UTC)

== Bonus/Domain Spells? ==

Great system!  Is there any way to handle bonus spells from wizard specialization or certain feats?  What about domain spells?

:I would suggest reducing the strain/mana costs of all specialist school spells by 1. Maybe prepaired domain spells could have mana/strain costs halved?

== Pushing Yourself Even Further ==

It occures to me that the variant of the UA spell point system that utilized Fatigue and Exhaustion conditions also included an option that allowed you to push yourself even harder to continue casting spells after your spell points have been consumed, at the cost of taking lethal and non-lethal damage equal to the spell point cost.

You're Mana/Strain system utilizes the exhaustion and strain mechanics, but without making you less powerful first (getting fatigued once half your spell points were gone/dropping to half spell points when you become fatigued). It has been pointed out that the spellcasters under this system could fire the same number of high level spells as a standard spellcaster in a single encounter, but not much more than this. I also note that you have added to the effects of exhaustion induced by this system that it prevents further spellcasting until the condition is removed. Well what if it's not ''neccissary'', just advisable to stop once you become exhausted?

Here's what I'm talking about: You're Strain Tolerance is reached, you've failed your save and become exhausted - now you have a choice. As you feel the exhaustion set in you can stop spellcasting and let the spell fizzle away into nothing, or you can continue anyways, pushing yourself to the limits and beyond causing yourself to become Sickened immediately after completing the spell. If you stop casting the spell that exhausted you, but cast another spell while exhausted you also become sickened. This sickened condition can be removed only by recovering all of your mana/strain, thus it can hasten you trip back to becoming exhausted due to the penalty it imposes on saves. Furthermore, if you cast spells while both sickened and exhausted, you must make a save just as against exhaustion (except that your strain is even higher above its tolerance and you suffer the penalty from the sickened condition) or become Nauseated, which ''will'' put a stop your spellcasting. Pushing yourself to this extreme is even harder to recover from and takes one week of complete rest to remove the nauseated condition. Attempting to cast spells while nauseated in this manner causes you to fall unconcious until such time as the DM deems appropriate (though likely either until you recover from the nauseated condition or would recover from unconciousness caused by subdual damage normally).

So, in summary, what I am proposing will have the net effects of: Allows spell casters to really push it in one single encounter, taking a great risk that just might save the party while decreasing their effectiveness in different but compairable ways so as to leave the decision over whether to proceed dependant on the situation; Incorporates the Sickened and Nauseated conditions into the progression of strain/mana expendatures; and throws in something alse to handle those player's inquisitive enough to see what happens if they try again, even when they really should know better by this point.

:Basically my idea for pushing yourself even further is that you make yourself physically ill, coughing blood, bleeding from the ears/eyes, that sort of thing. The recovery period for the nausea is based on that, and you are assumed to be pretty much bedridden anyways during that time. If for some reason you do end up exerting yourself during the week following the day you become nausaeted due to strain, that day does not count towards your recovery, you immediately gain twice your strain tolerance worth in strain, becoming exhaused and sickened, and your recovery period increases by 1 day.

== Anyone Modify This for Partial Spellcasters / Multiclass Spellcasters? ==

I have never gotten around to it, and truth be told I'm a bit stuck. Anyone want to take a whack at this challenge? [[User:Surgo|Surgo]] 13:40, 22 June 2012 (UTC)

:Well, for partial spellcasters, you just have to use a slightly different strain cost table. Multiclass spellcasters are more difficult. I guess you could make them function like Theurges, where the number of spell levels lost is capped at some small number, but then you might get odd things where people go Fighter 19/Wizard 1 for the 8th level spells. Hmm...
:Perhaps the Fighter 19/Wizard 1 could just cast 1st level spells like candy &mdash; that would even be easy to implement &mdash; just use the strain cost row based on your character level. --[[User:Foxwarrior|Foxwarrior]] 16:40, 22 June 2012 (UTC)

::If you're willing to sacrifice the relatively minor even level strain cost adjustments, you could redo the table in a "highest spell level" format instead. That would basically function for full, partial, and multiclass casters without any additional work since they could just look up their highest spell level from their class progression and use the costs listed on the table, though it still shafts multiclass casters a bit. Strain tolerance would be determined based on casting progression; it basically works for multi-class casters already. - [[User:Tarkisflux|Tarkisflux]] <sup>[[User talk:Tarkisflux|Talk]]</sup>  16:52, 22 June 2012 (UTC)

== Upcoming Modifications ==

I plan to get to this tonight, for Bards and other lesser spellcasters. [[User:Surgo|Surgo]] 19:25, 2 July 2012 (UTC)

:Complete. Tell me what you think. [[User:Surgo|Surgo]] 18:38, 22 July 2012 (UTC)


Did you do it, Surgo? I can't find it.  --coins

== Requesting Clarification ==

I used an older version of these rules briefly in a campaign I ran, they have been updated significantly since.  I am left wondering, I know that a high ability score grants a higher strain tolerance, so a caster can cast more spells.  What I'm wondering is, does a higher ability score (intelligence for example) still give a prepared spell caster, bonus spell slots to prepare additional spells for the day? Or would a wizard in this example be capped at 4 different spells prepared per spell level that he/she can cast within the limits of their strain tolerance? Tunganation ([[User talk:Tunganation|talk]]) 09:06, 20 February 2014 (UTC)

What are your questions? Because I am really confused by your sentences.  --coins
: Hmm let me put it another way.  A wizard gets up to 4 (depending on caster level) spells at each level they prepare each day from their wizard levels. Under the normal rules they get extra spells based on having a high intelligence.  In the Mana-Based Spellcasting variant, their intelligence sets their strain tolerance pool.  A higher intelligence lets them cast the spells they have memorized for the day more times, because they can take more strain than a low intelligence wizard.  My question is do they still get the extra spell slots from high intelligence, so that they have a wider selection of spells? Or are they limited to the max of 4 at each level they get from Wizard levels alone?  Wizard is just an example here as it would apply equally to Clerics, Druids, Archivists and so on.

:To provide an example, say I am playing a wizard who is level 5 with an intelligence of 20.  So this Wizard would have a Strain Tolerance of 22.  Say I prepare his spells for the day, his first level spells will be Mage Armor, Shield, Sleep, and Color Spray.  From his high intelligence he would normally get bonus spells, so does he also get to prepare Grease and Sleep? Or not? Tunganation ([[User talk:Tunganation|talk]]) 08:17, 23 December 2014 (UTC)

::Great question. I am tempted to say "yes", but that makes them more powerful than if you said "no". (this is surgo, just not logged in)

== Are any other casters missing? ==

This now (after too many years) covers normal wizards/sorcerers/full casters, as well as specialist wizards and partial casters (Bard, 4-level jokers, etc). Is there anything that remains missing from this? [[User:Surgo|Surgo]] ([[User talk:Surgo|talk]]) 06:14, 14 February 2016 (UTC)

:Actually, I think this would be very fitting for maneuver initiators as well; just base it off of constitution or strength or dexterity (or some combination). I think it is more fitting than the current system for power users, representing mental fatigue. I can also see a variant feat or skill allowing the caster to use strain in the creations of alchemy, with the more highly skilled caster requiring less time, equipment and materials to make his potions. A high level caster needs a potion of remove disease? He gets out his vial of pure water, funnels enough mana into it to overcome the lack of materials and time, and a standard action later has his potion. He could also take a previously created healing potion and supercharge it with some mana infusion, applying meta-magic effects to it. Enchanting could be altered as well to not only require crafting feats, but also using strain to catalyze the effect; strain could also be used to modify DC rolls in the process. In short I like the concept, and I think it could easily be expanded to become the base for most classes, and a great many specific endeavors (such as alchemy and enchanting).[[User:Elohim|Elohim]] ([[User talk:Elohim|talk]]) 18:57, 12 May 2016 (UTC)

::I've read this a few times now and I really like where you're going with this, but I haven't been able to put pen to paper for it just yet. [[User:Surgo|Surgo]] ([[User talk:Surgo|talk]]) 01:23, 16 August 2016 (UTC)

== This is not the best way to do this. ==

There's a better way to do this.  I've worked a perfect and complete system of [[User:Cedric/Mana|mana]] for D&D.  All classes accrue mana each round 100 mana for each 5ft of speed.  Most classes use this life force in their own way (see [[user:cedric/D&D 6 Homebrew|cedric/D&D 6 Homebrew]]).  Fighter's mana, for example, turns into muscle (if their muscles are used).  Clerics can store it with their gods and draw on it for epic healing abilities.  [[User:Cedric/Wizard|Mages]], obviously use it to case spells.  The power of the mana grows with level, so while they don't get more mana per round at higher levels, that mana is more packed with power.  This removes a lot of bogus spells that get "tougher" with levels.

The other sweet thing about this mana concept is to integrate it with other parts of D&D mechanics:  food value, corpse's mana (value if sacrificial or eaten), wars with mana, value at death, conversion with HP, etc.

Anyway, I'm just putting it out there as a possible collaboration.  I have to say it's pretty sweet.  [[User:Cedric|Cedric]] ([[User talk:Cedric|talk]]) 20:29, 27 February 2018 (MST)

:That is not better of perfect.  Only vastly more complicated, time consuming, and less fun.  Your writeup actually keeps track of breathing as an action that costs points to do! You regain mana for bathing?! and you die if it reaches zero. Someone for got K.I.S.S. (Keep It Simple, Stupid). [[User:Grog toad|Grog toad]] ([[User talk:Grog toad|talk]]) 05:04, 28 February 2018 (MST)

:: No, you don't get it.  The detail is there if the DM wants or needs it, and it especially critical at boundary areas where it can make a difference, not to bog anyone down when it's not needed.  But I like the idea of gaining mana for bathing!

:::Going to have to say "no". One of these has playable and complete rules (mine). One of them does not (yours). I do not see what sort of value could be gained by any kind of collaboration, sorry. [[User:Surgo|Surgo]] ([[User talk:Surgo|talk]]) 08:19, 28 February 2018 (MST)

== Table Error ==

Table: Mage Strain Costs is incorrect for spells of level 5 and above. It looks like you forgot to account for the fact that strain cost fall for 2.5 rather than 2 on the level when a new level of spells are gained. Curiously spells of level 2 and 4 are correct despite also having a starting cost ending in .5 and correctly reduce costs on each of their first four levels.