MrStabby
2021-03-16, 06:45 PM
So I was having some ideas about a system. It is obviously early days and I am not yet looking too much for feedback on specifics but would love some comentary on the overall themes, especially anything you think I have not thought about. If nothing else I might use this thread as a repository for my own thoughts.
And no, I don't think that I am going to publish and make a fortune or anything like that, but rather just enjoy it as a) an interesting thought excercise, b) a chance to discuss such things here.
I suspect a lot of ideas won't be original - having played too few other systems I can't really avoid that - but better to discuss the actual mechanics than to say "take a look at XYZ system that does this"; after all going away and reading the answer elsewhere neither provides the stimulation of working things through myself nor discussing them. Mostly my experience is D&D (a bit of 2nd, bit of 3rd lot of 5th) and I suspect it will show (if only when using it as a point of reference).
I have a few guiding principles that I am trying to follow:
1) Simple is better but it trades off against flexability and customsiation and a feeling off differentiation. Where it doesn't always go simple. My broad target is a little more complex than D&D 5E but significantly less complicated than 3E.
2) Complexity at the table is worse than complexity away from the table. Is someone needs a spreadsheet to build their character thats... OK I guess. If someone needs a spreadsheet to play their character - that's a hard no.
3) Balance is important. Flavour is easier to add on the bones of a well balanced set of mechanics.
4) Diversity in character/playstyle is a good measure of success.
5) Build combat first, not because it is more important than other pillars but because it determines the structure/balance that other elements will have to fit around.
So I am not preenting a set of game rules here, but more trying to get feedback on some concepts.
A lot of effects on sliding scales. 5th edition has a range of conditions that could make sense as a scale: grappled, slowed, restrained, stunned, paralyzed, petrified... and then the effect of haste on the other side. Rather than having a lot of spells that don't interact have a smaller number of spells that instead are incremental. So a "slow" spell that asks for 5 saves and then advances an enemy a number of points up the track equal to the number of failures. This would allow haste to offset the spell somewhat (and vice versa) but would also allow for less all or nothing type effects that pass/fail/legendary saves produce. Saving well against a powerful spell can be more likely but can still give a muted effect.
So scales can be: a slowed/restrained scale, a wounded scale, a madness scale... I wouldn't want to go overboard on this though.
Speaking of which, I am seing a number of spells be multiple saves - so a fireball might be make X saves, take Y d6 damage where Y is the number of failed saves.
Reduce stats to 4
Athleticism - roughly analagous to D&D's dexterity and strength
Resilience - analagous to constitution, also will determine stamina
Wit - quick thinking, a combination of charisma and intelligence
Wisdom - slower, deeper thinking and analysis, a mixture of D&D's intelligence and wisdom
I think that this bare bones approach still offers enough scope for customisation.
Spells will now have a casting stat rather than a class having a casting stat. So if you want to play a blasting wizard your spells will predominantly use wisdom and you are likely to prioritise wisdom. As a wizard you will still have access to illusion spells, but they will be intelligence based so you are likely to be less good at them. The aim is to avoid any character being able to cover all the bases better than any other through the right choice of spells and to encourage more specialist/narrower themed casters than D&D's omnimancers. Each stat will have spells associted with them.
Athleticism will be spells that need attack rolls. Throwing firebolts for wizards, divine smites for clerics/paladins, lashing vines for druid type characters.
Resilience will be drawing on your own suffering to protect others. This will empower healing and "abjuration" type spells - wards and protective magics.
Wit will be for those spells where it makes a difference whether a target is living or inanimate: enchantment, illusion but also a lot of necromancy will sit here
Wisdom will be for those spells where there is no mind/spirit in play - your fireballs, your walls of stone, your control weather spells.
I am hoping that this structure will enable fewer classes, but each to be built in more ways. So for example you can play a cleric - if you want a classic paladin type character you can grab some fighter multiclass but put your good stats into athleticism and resilieince to make you good at smiting and good at protecting those around you. If on the other hand you put your strong stats into wit and wisdom you would be calming emotions and blasting with holy fire. Hopefully if every stat has something to offer then each combination should also be viable.
I am debating eliminating numerical stats and just replacing them with levels: Superb, high, mediocre, poor, and at character creation you just rank them.
Access to spells will be through something I am calling traits. So for example "raise dead" as an effect might have a level of 7 and the traits of divine and necromancy. This would mean that to have access to the spell a character would need to have some combination of divine and necromancy traits that would add up to 7. So a single class cleric would get access to the spell at the same time as a multiclass cleric/necromancer. Multiclasing would therefore limit the range of higher level spells you can access without eliminating access to higher level magic altogether. Traits could be gained from class, race, background and feats - sometimes by default and sometimes as a choice (say taking a wizard level gives you an aracane point ans one point from a school of magic of your choice)
I would intend for spells to interact with, rather than replace skills. A "Bigby's Hand" type spell would create a diesembodied hand that would have strength equal to the caster's (+/- an amount depending on the spell level). "Invisibility" as a spell would boost the stealth skill rather than directly making the target invisible. Skills themselves would have scope for criical success to achieve the truely remarkable - i.e. a sufficiently high stelth check would actually be invisible, a sufficiently high strength check could shatter stone, a sufficiently high perception check would identify the location of everyone in the next room from their heartbeat and so on.
For spells known I am thinking to divide them up amongst Major and Minor themes. A Major spell is a broadly useful spell that you will use a lot and that will scale well as you level up - something like fireball. A minor spell might be better though of as a niche spell - invisibility or raise dead effects would fit this bill. As each class would get a certain number of each of these that would scale with level you will build up a collection of big theme spells and a more extensive library of spcialist solutions.
Martial characters will have some similar elements to spellcasters. Some weapons, skills, stances and strikes will need certain traits as well. Some traits will be relatively neutral such as Law or War so there is no need to have a magical character to pick these up. Others might be things like Deciet that would enable both rogue-like abilities and illusion type magic.
Likewise these elements of martial prowess may sometimes be dependant upon a range of ability scores - I see monk type stance mechanics being dependant upon wisdom, ability to launch anatomically devastating strikes against an enemy being tied to intelligence and so on. Some abilities will be passive, some limited and some expending stamina...
I want stamina as a resource. I am thinking you have a pool of stamina (and a maximum level) and it tops up an amount each turn depending on your resiliance. Some abilities will cost resilience to use - representing a particular exertion. When you take damage you can also block some of it using resilience left in your pool. In this way you can be beaten down till you are vulnerable and you need to trade off using your stamina in offence to end a fight quickly vs holding back to block more effectively. The relative size and impact of this is still something I am thinking about - it will have some major balance implications.
More martial focussed characters will also have more flexability through weapon choices and decisions about what to specialise in. More weapons will have (simple) special rules to them such that a differently equipped character will play differently. Also there will be more strengths and weaknesses to them to make fighting with the wrong weapon somewhat disadvantagious.
I have a few open questions/challenges.
1) How to handle healing? So healing is often a bit of a dull mechanic and healing isn't as fun as it could be. I think to be fun, it shouldn't just matter how much you do but also how you do it and who you heal. I would like it if there were some limits that came into play at different damage levels so you need to avoid pop up healing as people near death are more limited, but this adds an extra element of bookkeeping and monitoring if you constantly are rocking back and forth accross different thresholds.
2) How does the game handle money/treasure. So there is a trade off again here. You want treasure to be exciting - i.e. there is a high marginal value to treasure. I also want most character strength to come form the character being awesome rather than just owning awesome stuff. How should consumables be embedded into classes? Should there be abilities such as getting extra benefits from healing potions etc. to encourage expenditure? How do we ensure that the rules work in both low and high magic item settings?
3) What is the optimal length of combat? In 5th edition length tends to be about 4 rounds. How do we have a system where it isn't just a race to do a load of damage as quickly as possible but that pacing can also be slower and more tactical?
4) Any particular tips on how to ensure that each character has multiple different things that it is optimal to do (under different circumstances), so that specialism is possible, even good, but you have to sometimes do things other than to use you single best ability over and over?
5) How much of the setting should be baked into the rules? So for example in D&D there are domains for cleric but you can use them with any setting's gods, on the other hand druid wildshape and animal conjuration presumes that there are certain beasts in the world to be summoned and the paladin assumed there are such things as undead and fiends such that divine smite bonus damage works.
6) Should there be class spell lists? I am leaning towards No. That the trait system should be enough.
7) How should spellcasting and martial ability differ between classes to help make them distinct?
8) What type of spellcasting should there be? Should this vary class to class? (I touch on this a bit later)
9) How to deal with the equivalent of D&D 5E advantage. So 3rd edition had a bucketfulls of bonuses that got added in different combinations - this was sometimes a bit confusing (not least for the DM looking on in bewilderment) and slowed down play. 5th edition has advantage and disadvantage which is somewhat binary and lacking in nuance (i.e. you cant add +2 or +3) but is quick and does create some interaction with things like critical hits and is very quick. I am leaning towards a bonus to a roll, the size of which is determined by abilities. So a rogue type character might get more of an advantage attacking from the shadows than a soldier reflected by abilities that give a bigger bonus. Not entirely sure though.
10) What to do about summons, pets etc. What stat should govern spells?
11) How to handle the action economy (again, touched upon later)
Classes
These are the rough classes I have in mind:
Gladiator: A fighter whose special emphasis is flexability. A jack of all trades when it comes to weapons, with abilities for quickly swapping them around and the broadest range of proficiencies in the game including with the most exotic weapons. Has access to a great set of combat tricks and techniques.
Soldier: A disciplined fighter with the best armour related abilities and good combat abilities. Access to some unique support abilities to protect, enchance and direct comrades in arms.
Trickster: The rogue with a supernatural set of abilities to sow chaos and confusion. Master of stealth and the shadows can be exceptionally dangerous making attacks whilst unseen or against a distracted enemy
Priest: Divine spellcaster - in addition to a somewhat narrow divine spell list will have deep access to another trait. Likely to have a good selection of major spells but a narrower selection of minor spells will limit versatility.
Magus: Arcane Caster - Deep access to the arcane tradition or an elemental tradition with broad but shallow access to multiple other traditions. Narrow selection of major spells but many minor spells means that the magus will often have just the right trick to fix a situation.
Ocultist: Witch and warlock wrapped up together. Deep access to a choice of Occult, Infernal, Shadow, Nature and prophesy disciplines, this class will have an emphasis on preparation - preparing the right spell, the right potion, scrying, setting traps. I anticipate this to be designed as a rewarding class for higher system mastery.
Paragon: The exemplar hero of your culture. Bonuses that are scaled versions of racial and cultural abilities, deep access to racial traditions and bonuses to ceremonial weapons and armours. An all round class but also able to provide a useful add-on to other classes in a multiclass character.
Traditions
These represent the spells and abilities that need certain class traits to be able to use. Many of these will have both spells and martial abilities tied to them
Arcane
Divine
Good
Evil
Death
Nature
Light
Occult
Eldritch
Shadow
Trickery
Law
War
Fire
Cold
Storm
Suffering
Infernal
Prophesy
Stone
Ethnicity (i.e. one for human, one for dwarf etc.)
Example Ability
Shadow Strike
Level 8
Traits: Shadow, Trickery, War, Goblin
Cost: 3 stamina
You make a single weapon attack against an enemy. In addition, should that attack hit, they must make 4 resilience saves, and for each one failed they take an additional 2d6 damage. If they fail 3 or more saves advance the target one step on the critical injury track.
So this ability would be accessible to warriors (who would get easy access to the war trait, Rogue type characters (access to trickery), certain casters (if you have access to the shadow trait and selected to invest in athleticism). Goblins would get access to it a bit sooner (assuming your race gives you a point or two in its corresponding trait).
The effect uses multiple saves to determine outcome to avoid an all/nothing scenario too often. It references a track (like the example upthread on the hasted to petrified track) which might provide debilitating effects and it adds a degree of situational usefulness to abilities - it is probably underwhelming against tough enemies but becomes rapidly better as your odds of getting 3+ save failures increases. I also envisage some specialist feats/abilities/things that would give something like "whenever you force one or more saves from a spell of type X, instead force that number of saves +1", which can help scaling.
I wanted to add a bit more about the two big questions I am having difficulty with: spellcating style and action/stamina economy.
For spellcasting I could see:
1) Full vancian casting with prepared spells
2) 5th ed style casting - limited subset of spells but can use spell slots on any of them
3) Spell points rather than spell slots for more versatility
4) At will resource-free casting.
Now there isn't a balance issue per se with any of them, more relaxed casting can have effects toned down so the net effect is everything is inline. My problems with each are:
Full Vancian/5E vancian - I want few enough spells that it feels like selected spells are an important thematic part of the character. Dropping them out feels a bit wierd. Also, if there are spell slots of different levels then either the higher level ones are tightly restricted... or they are not. If they are tightly restricted then it feels like they should be pretty powerful to warrent their rarity - if they are then you run into a problem where one PC can dominate a few encounters per day and leave others underwhelming. If there is a pretty liberal restriction then why bother - if top end resources won't run out then why track them?
There is potentially an option here where ALL spells scale pretty well with level so the top level ones are only a little bit better - enough to get an edge rather than to dominate.
Spell points - more flexible than spell slots but oh so much riskier. With spell slots you maximum efficiency is determined by your most efficient spell of each level. With spell points where you can redistribute power between levels your maximum efficiency is determined by your most efficient spell. One mistake and you can screw a game right up.
At will saves all bookkeeping, and might be very efficient but undermines somewhat the feeling of progress as you level up. Either there is no qualitative change in what you do or you leave behind what you were doing before with no feeling of character continuity.
Potentially it could vary between class but this is all a little bit more complicated as the end result should still multiclass effectively. If spells are more tied to traits than classes then
Multiclass asside I could see something like:
Priest: At will casting of major spells. A pool of points available to add further upgrades to the spell (so the spells grow in power weakly with priest level but can be enhanced with these points to make them more powerful. Minor themed spells (which are the specialist elements) can only be cast by also using this pool of points. At will casting would mean healing spells would need to be pushed to the minor themes if hit-point attrition is ever expected to be a thing.
Magus: At will casting of major spells. Also adds to the pool of points for further upgrades. Minor themed spells are pure vancian (as they will get so many this is a bigger boost in resources but it also encourages their use as there is no opportunity cost to the resource.
Occultist: At will casting of major spells. Minor spells are pre-purchased using spell points, and the palyer choses how much to spens on minor themed spells each rest they take.
Paragon: Slower spell progression and all spells are treated as at-will major spells (but no race should provide a route to healing spells)
I see the martial abilities all using stamana and so with a common resource they should be easy enough to pool; there is a seperate challenge there of ensuring that lower level ones are not completely superseeded.
The other big open question is around action economy. How much, and of what, should you be able to do in a turn and how flexable should this be?
I like 5th edition. The movement, bonus action, action, reaction set of resources is neat, it is clean, it is simple. What it does do is limit some flexability but also some avenues of specialisation.
If there were a budget (say a turn takes 6 seconds and you get 6 seconds to assign) - you could split this between attacks, movement spellcasting and different actions could take a different length of time; it opens up classes to speeding up certain actions with a degree of granularity 5th edition doesn't provide. I like the idea of greater granularity and the tactical opportunities it can provide but I can also see it slowing down play at the table as each person tries to solve their own Knapsack Problem as they divide their time amongst different actions.
And no, I don't think that I am going to publish and make a fortune or anything like that, but rather just enjoy it as a) an interesting thought excercise, b) a chance to discuss such things here.
I suspect a lot of ideas won't be original - having played too few other systems I can't really avoid that - but better to discuss the actual mechanics than to say "take a look at XYZ system that does this"; after all going away and reading the answer elsewhere neither provides the stimulation of working things through myself nor discussing them. Mostly my experience is D&D (a bit of 2nd, bit of 3rd lot of 5th) and I suspect it will show (if only when using it as a point of reference).
I have a few guiding principles that I am trying to follow:
1) Simple is better but it trades off against flexability and customsiation and a feeling off differentiation. Where it doesn't always go simple. My broad target is a little more complex than D&D 5E but significantly less complicated than 3E.
2) Complexity at the table is worse than complexity away from the table. Is someone needs a spreadsheet to build their character thats... OK I guess. If someone needs a spreadsheet to play their character - that's a hard no.
3) Balance is important. Flavour is easier to add on the bones of a well balanced set of mechanics.
4) Diversity in character/playstyle is a good measure of success.
5) Build combat first, not because it is more important than other pillars but because it determines the structure/balance that other elements will have to fit around.
So I am not preenting a set of game rules here, but more trying to get feedback on some concepts.
A lot of effects on sliding scales. 5th edition has a range of conditions that could make sense as a scale: grappled, slowed, restrained, stunned, paralyzed, petrified... and then the effect of haste on the other side. Rather than having a lot of spells that don't interact have a smaller number of spells that instead are incremental. So a "slow" spell that asks for 5 saves and then advances an enemy a number of points up the track equal to the number of failures. This would allow haste to offset the spell somewhat (and vice versa) but would also allow for less all or nothing type effects that pass/fail/legendary saves produce. Saving well against a powerful spell can be more likely but can still give a muted effect.
So scales can be: a slowed/restrained scale, a wounded scale, a madness scale... I wouldn't want to go overboard on this though.
Speaking of which, I am seing a number of spells be multiple saves - so a fireball might be make X saves, take Y d6 damage where Y is the number of failed saves.
Reduce stats to 4
Athleticism - roughly analagous to D&D's dexterity and strength
Resilience - analagous to constitution, also will determine stamina
Wit - quick thinking, a combination of charisma and intelligence
Wisdom - slower, deeper thinking and analysis, a mixture of D&D's intelligence and wisdom
I think that this bare bones approach still offers enough scope for customisation.
Spells will now have a casting stat rather than a class having a casting stat. So if you want to play a blasting wizard your spells will predominantly use wisdom and you are likely to prioritise wisdom. As a wizard you will still have access to illusion spells, but they will be intelligence based so you are likely to be less good at them. The aim is to avoid any character being able to cover all the bases better than any other through the right choice of spells and to encourage more specialist/narrower themed casters than D&D's omnimancers. Each stat will have spells associted with them.
Athleticism will be spells that need attack rolls. Throwing firebolts for wizards, divine smites for clerics/paladins, lashing vines for druid type characters.
Resilience will be drawing on your own suffering to protect others. This will empower healing and "abjuration" type spells - wards and protective magics.
Wit will be for those spells where it makes a difference whether a target is living or inanimate: enchantment, illusion but also a lot of necromancy will sit here
Wisdom will be for those spells where there is no mind/spirit in play - your fireballs, your walls of stone, your control weather spells.
I am hoping that this structure will enable fewer classes, but each to be built in more ways. So for example you can play a cleric - if you want a classic paladin type character you can grab some fighter multiclass but put your good stats into athleticism and resilieince to make you good at smiting and good at protecting those around you. If on the other hand you put your strong stats into wit and wisdom you would be calming emotions and blasting with holy fire. Hopefully if every stat has something to offer then each combination should also be viable.
I am debating eliminating numerical stats and just replacing them with levels: Superb, high, mediocre, poor, and at character creation you just rank them.
Access to spells will be through something I am calling traits. So for example "raise dead" as an effect might have a level of 7 and the traits of divine and necromancy. This would mean that to have access to the spell a character would need to have some combination of divine and necromancy traits that would add up to 7. So a single class cleric would get access to the spell at the same time as a multiclass cleric/necromancer. Multiclasing would therefore limit the range of higher level spells you can access without eliminating access to higher level magic altogether. Traits could be gained from class, race, background and feats - sometimes by default and sometimes as a choice (say taking a wizard level gives you an aracane point ans one point from a school of magic of your choice)
I would intend for spells to interact with, rather than replace skills. A "Bigby's Hand" type spell would create a diesembodied hand that would have strength equal to the caster's (+/- an amount depending on the spell level). "Invisibility" as a spell would boost the stealth skill rather than directly making the target invisible. Skills themselves would have scope for criical success to achieve the truely remarkable - i.e. a sufficiently high stelth check would actually be invisible, a sufficiently high strength check could shatter stone, a sufficiently high perception check would identify the location of everyone in the next room from their heartbeat and so on.
For spells known I am thinking to divide them up amongst Major and Minor themes. A Major spell is a broadly useful spell that you will use a lot and that will scale well as you level up - something like fireball. A minor spell might be better though of as a niche spell - invisibility or raise dead effects would fit this bill. As each class would get a certain number of each of these that would scale with level you will build up a collection of big theme spells and a more extensive library of spcialist solutions.
Martial characters will have some similar elements to spellcasters. Some weapons, skills, stances and strikes will need certain traits as well. Some traits will be relatively neutral such as Law or War so there is no need to have a magical character to pick these up. Others might be things like Deciet that would enable both rogue-like abilities and illusion type magic.
Likewise these elements of martial prowess may sometimes be dependant upon a range of ability scores - I see monk type stance mechanics being dependant upon wisdom, ability to launch anatomically devastating strikes against an enemy being tied to intelligence and so on. Some abilities will be passive, some limited and some expending stamina...
I want stamina as a resource. I am thinking you have a pool of stamina (and a maximum level) and it tops up an amount each turn depending on your resiliance. Some abilities will cost resilience to use - representing a particular exertion. When you take damage you can also block some of it using resilience left in your pool. In this way you can be beaten down till you are vulnerable and you need to trade off using your stamina in offence to end a fight quickly vs holding back to block more effectively. The relative size and impact of this is still something I am thinking about - it will have some major balance implications.
More martial focussed characters will also have more flexability through weapon choices and decisions about what to specialise in. More weapons will have (simple) special rules to them such that a differently equipped character will play differently. Also there will be more strengths and weaknesses to them to make fighting with the wrong weapon somewhat disadvantagious.
I have a few open questions/challenges.
1) How to handle healing? So healing is often a bit of a dull mechanic and healing isn't as fun as it could be. I think to be fun, it shouldn't just matter how much you do but also how you do it and who you heal. I would like it if there were some limits that came into play at different damage levels so you need to avoid pop up healing as people near death are more limited, but this adds an extra element of bookkeeping and monitoring if you constantly are rocking back and forth accross different thresholds.
2) How does the game handle money/treasure. So there is a trade off again here. You want treasure to be exciting - i.e. there is a high marginal value to treasure. I also want most character strength to come form the character being awesome rather than just owning awesome stuff. How should consumables be embedded into classes? Should there be abilities such as getting extra benefits from healing potions etc. to encourage expenditure? How do we ensure that the rules work in both low and high magic item settings?
3) What is the optimal length of combat? In 5th edition length tends to be about 4 rounds. How do we have a system where it isn't just a race to do a load of damage as quickly as possible but that pacing can also be slower and more tactical?
4) Any particular tips on how to ensure that each character has multiple different things that it is optimal to do (under different circumstances), so that specialism is possible, even good, but you have to sometimes do things other than to use you single best ability over and over?
5) How much of the setting should be baked into the rules? So for example in D&D there are domains for cleric but you can use them with any setting's gods, on the other hand druid wildshape and animal conjuration presumes that there are certain beasts in the world to be summoned and the paladin assumed there are such things as undead and fiends such that divine smite bonus damage works.
6) Should there be class spell lists? I am leaning towards No. That the trait system should be enough.
7) How should spellcasting and martial ability differ between classes to help make them distinct?
8) What type of spellcasting should there be? Should this vary class to class? (I touch on this a bit later)
9) How to deal with the equivalent of D&D 5E advantage. So 3rd edition had a bucketfulls of bonuses that got added in different combinations - this was sometimes a bit confusing (not least for the DM looking on in bewilderment) and slowed down play. 5th edition has advantage and disadvantage which is somewhat binary and lacking in nuance (i.e. you cant add +2 or +3) but is quick and does create some interaction with things like critical hits and is very quick. I am leaning towards a bonus to a roll, the size of which is determined by abilities. So a rogue type character might get more of an advantage attacking from the shadows than a soldier reflected by abilities that give a bigger bonus. Not entirely sure though.
10) What to do about summons, pets etc. What stat should govern spells?
11) How to handle the action economy (again, touched upon later)
Classes
These are the rough classes I have in mind:
Gladiator: A fighter whose special emphasis is flexability. A jack of all trades when it comes to weapons, with abilities for quickly swapping them around and the broadest range of proficiencies in the game including with the most exotic weapons. Has access to a great set of combat tricks and techniques.
Soldier: A disciplined fighter with the best armour related abilities and good combat abilities. Access to some unique support abilities to protect, enchance and direct comrades in arms.
Trickster: The rogue with a supernatural set of abilities to sow chaos and confusion. Master of stealth and the shadows can be exceptionally dangerous making attacks whilst unseen or against a distracted enemy
Priest: Divine spellcaster - in addition to a somewhat narrow divine spell list will have deep access to another trait. Likely to have a good selection of major spells but a narrower selection of minor spells will limit versatility.
Magus: Arcane Caster - Deep access to the arcane tradition or an elemental tradition with broad but shallow access to multiple other traditions. Narrow selection of major spells but many minor spells means that the magus will often have just the right trick to fix a situation.
Ocultist: Witch and warlock wrapped up together. Deep access to a choice of Occult, Infernal, Shadow, Nature and prophesy disciplines, this class will have an emphasis on preparation - preparing the right spell, the right potion, scrying, setting traps. I anticipate this to be designed as a rewarding class for higher system mastery.
Paragon: The exemplar hero of your culture. Bonuses that are scaled versions of racial and cultural abilities, deep access to racial traditions and bonuses to ceremonial weapons and armours. An all round class but also able to provide a useful add-on to other classes in a multiclass character.
Traditions
These represent the spells and abilities that need certain class traits to be able to use. Many of these will have both spells and martial abilities tied to them
Arcane
Divine
Good
Evil
Death
Nature
Light
Occult
Eldritch
Shadow
Trickery
Law
War
Fire
Cold
Storm
Suffering
Infernal
Prophesy
Stone
Ethnicity (i.e. one for human, one for dwarf etc.)
Example Ability
Shadow Strike
Level 8
Traits: Shadow, Trickery, War, Goblin
Cost: 3 stamina
You make a single weapon attack against an enemy. In addition, should that attack hit, they must make 4 resilience saves, and for each one failed they take an additional 2d6 damage. If they fail 3 or more saves advance the target one step on the critical injury track.
So this ability would be accessible to warriors (who would get easy access to the war trait, Rogue type characters (access to trickery), certain casters (if you have access to the shadow trait and selected to invest in athleticism). Goblins would get access to it a bit sooner (assuming your race gives you a point or two in its corresponding trait).
The effect uses multiple saves to determine outcome to avoid an all/nothing scenario too often. It references a track (like the example upthread on the hasted to petrified track) which might provide debilitating effects and it adds a degree of situational usefulness to abilities - it is probably underwhelming against tough enemies but becomes rapidly better as your odds of getting 3+ save failures increases. I also envisage some specialist feats/abilities/things that would give something like "whenever you force one or more saves from a spell of type X, instead force that number of saves +1", which can help scaling.
I wanted to add a bit more about the two big questions I am having difficulty with: spellcating style and action/stamina economy.
For spellcasting I could see:
1) Full vancian casting with prepared spells
2) 5th ed style casting - limited subset of spells but can use spell slots on any of them
3) Spell points rather than spell slots for more versatility
4) At will resource-free casting.
Now there isn't a balance issue per se with any of them, more relaxed casting can have effects toned down so the net effect is everything is inline. My problems with each are:
Full Vancian/5E vancian - I want few enough spells that it feels like selected spells are an important thematic part of the character. Dropping them out feels a bit wierd. Also, if there are spell slots of different levels then either the higher level ones are tightly restricted... or they are not. If they are tightly restricted then it feels like they should be pretty powerful to warrent their rarity - if they are then you run into a problem where one PC can dominate a few encounters per day and leave others underwhelming. If there is a pretty liberal restriction then why bother - if top end resources won't run out then why track them?
There is potentially an option here where ALL spells scale pretty well with level so the top level ones are only a little bit better - enough to get an edge rather than to dominate.
Spell points - more flexible than spell slots but oh so much riskier. With spell slots you maximum efficiency is determined by your most efficient spell of each level. With spell points where you can redistribute power between levels your maximum efficiency is determined by your most efficient spell. One mistake and you can screw a game right up.
At will saves all bookkeeping, and might be very efficient but undermines somewhat the feeling of progress as you level up. Either there is no qualitative change in what you do or you leave behind what you were doing before with no feeling of character continuity.
Potentially it could vary between class but this is all a little bit more complicated as the end result should still multiclass effectively. If spells are more tied to traits than classes then
Multiclass asside I could see something like:
Priest: At will casting of major spells. A pool of points available to add further upgrades to the spell (so the spells grow in power weakly with priest level but can be enhanced with these points to make them more powerful. Minor themed spells (which are the specialist elements) can only be cast by also using this pool of points. At will casting would mean healing spells would need to be pushed to the minor themes if hit-point attrition is ever expected to be a thing.
Magus: At will casting of major spells. Also adds to the pool of points for further upgrades. Minor themed spells are pure vancian (as they will get so many this is a bigger boost in resources but it also encourages their use as there is no opportunity cost to the resource.
Occultist: At will casting of major spells. Minor spells are pre-purchased using spell points, and the palyer choses how much to spens on minor themed spells each rest they take.
Paragon: Slower spell progression and all spells are treated as at-will major spells (but no race should provide a route to healing spells)
I see the martial abilities all using stamana and so with a common resource they should be easy enough to pool; there is a seperate challenge there of ensuring that lower level ones are not completely superseeded.
The other big open question is around action economy. How much, and of what, should you be able to do in a turn and how flexable should this be?
I like 5th edition. The movement, bonus action, action, reaction set of resources is neat, it is clean, it is simple. What it does do is limit some flexability but also some avenues of specialisation.
If there were a budget (say a turn takes 6 seconds and you get 6 seconds to assign) - you could split this between attacks, movement spellcasting and different actions could take a different length of time; it opens up classes to speeding up certain actions with a degree of granularity 5th edition doesn't provide. I like the idea of greater granularity and the tactical opportunities it can provide but I can also see it slowing down play at the table as each person tries to solve their own Knapsack Problem as they divide their time amongst different actions.