On Skills, Attributes, and their intersection with SAs.

[Agamemnon]

There are some out there who champion their own work with a kind of righteous zealotry, unwilling or unable to see any flaws it may have or ways in which it may be improved. Then there are people like me, who are their own worst critics. I can't count how many times I've approached poor Higgins out of the blue with the stark and foreboding message "Something is broken." I think he has grown to loathe those messages because they are usually followed by me attempting to completely upturn or rewrite some element of the book he thought was settled. Oops. It's in the spirit of this unfortunately frequent occurrence I'm going to ramble about something to the forums at large and see if I can get feedback before I do something drastic.

As I've played various test-games and run through the demos we've set up, I've come across something that could either be a bug or a feature depending on how you might look at it. Story Aspects form the heart of 'Bastards in many ways. They guide play, they form the reward mechanism, and they keep your butt out of the fire in the process through the bonus dice. In combat, they integrate seamlessly into the flow of things and make the whole thing feel just a little more exciting. Outside of combat, they cause ... weirdness.

The problem rests in the intersection of the way skills, attributes, and SAs are scored. All three are on a 0 or 1-5 range. The unintended side effect of this is that having 5 dice from an SA is functionally identical to having 5 dice in an attribute or a skill. In practice, this means that a someone who with a mediocre attribute and apprentice level training (2+2) for their skill check stands an even chance of beating someone with a high attribute and grandmastery in their skill (4+5) if they just want it bad enough. In my own games, I've watched PCs who were untrained in a skill beat other PCs who had actually invested in said skill because the former had SAs firing (or more accurately, both had SAs firing but one had 5 dots and the other had 1).

The effect is even more pronounced when making unopposed checks. The way the Obs are structured, +5 dice from an SA are +2.5 successes at TN6. If we start talking about exploding dice, things can get even more goofy. If you set the difficulty scale for skill/attribute checks without SAs factored in, then SAs will make difficult things relatively trivial. If you set the difficulty scale in order to account for the influence of SAs, you make difficult checks impossible when they aren't firing.

Now you can make an argument that this is all a feature, rather than a bug. It takes a very Hollywood kind of logic where what's more important than the skill of the character is their passion/desire/the weight of the narrative. Of course Daniel-san can't realistically train hard enough over the summer to beat out the kids who have been doing this for years, but it's way more dramatic if he does. It makes perfect story-telling sense that the character who completely lacks Oratory can get up on stage and make an impassioned speech with their SAs firing that sways the crowd away from the evil vizier who would mislead them. You can certainly make these arguments and defend the system as it stands now.

But I'm not sure I like those arguments.

My gut feeling is that this short-changes people who have actually spent the time and points investing in those skills. Just like the poor bastards from Cobra Kai who have gone to practice twice a week every week since they were in kindergarten and put in the blood, sweat, and broken bones to earn what they have, it feels a bit cheap when someone else can completely school you because they had a training montage and the story was in their favor.

Oddly enough, I keep using The Karate Kid as an example, but it's probably the worst one I could go for. The SA dice seems to work fine for combat.. So why is that?

It's the nature of the die pool. First, you're comparing far differently-sized pools. CP is Agility+Cunning (already the size of a skill pool), plus your proficiency rating which for most combatants will be anywhere from 0-10. Then add in the potential to get in another 0-4 dice from Reach? You're looking at a potential pool of anywhere from 4 to 24 dice. Where +5 dice from SAs can potentially be a 50% boost in total die pool for skills or attribute checks, it may only be a 20% boost to your combat pool. On top of that, the effect is even further diluted by the way you will split your CP in most cases and that activation costs and maneuvers can further manipulate the dice you have to spend. The outcome of a fight between two opponents is far less influenced by SAs than any skill or attribute contest between opponents would be.

So what can be done?
If it's a thing I/we decide to tackle at all, then the crux of the issue rests at the intersection of skills, attributes, and SAs. To help mitigate the effect, we either need to reduce the amount of SA points you can have, or change the way in which skills and attributes are rated.

The former does not thrill me. You'd have to reduce SAs to a maximum of 3 dots or so in order to achieve the effect. Skills and Attributes would still benefit more from SAs than combat does, and I fear that +3 dice in combat might not be enough to make a significant impact in combat. Of course, if +3 is the max, that means most of the time SAs fire, you will only be getting 1 or 2 dice out of it which makes the whole thing slightly less exciting. More worryingly, reducing the total number of SA points you have to play with makes spending them on anything (whether advancement or narrative effects) slightly more painful and a careful process, as to buy anything expensive you may have to wipe out multiple SAs to afford it.

The latter option is doable but requires a significant overhaul of the current skill system. You'd have to bring attributes and skills to a 1-8 or 1-10 scale to bring their numbers more in line with Proficiencies. While this would actually make the individual skill way more important by comparison, you'd have to rework a bunch of things.

• We'd have to toss the X+Y system of attribute/skill interaction, losing the flexibility of being able to differentiate between, say, an acumen/larceny check for knowledge of burglary practices and agility/larceny for the physical act of trying to gain entry. On the other hand, from a GM perspective, it would at least require a tiny bit less computation in play (you wouldn't need to think through "what is the right attribute to pair with this.")
• We'd be required to find a different way to make attributes and skills meaningfully interact since the above is no longer possible.
• Strength and Stamina would need to be balanced as an additional step for the purposes of DR and AV in combat. You'd need to halve them when you put them on your sheet. Dragon Tap and Naked Dwarf are known issues in TROS that stemmed from the difference between Strength and Stamina (or whatever the stat was in TROS) being too swingy.
• We'd have to get rid of the dots, which would probably make Higgins cry. He likes dots.

There are probably additional wrinkles you'd have to iron out as well, but the latter is at least doable even if it would require a significant rewrite of some material.

Give me some feedback. It this all a non-issue? Is it as broken as my gut says? Which fix do you prefer? Do you have any alternate fixes to suggest? Are dots as important as Higgins says?

Let me know.

[higgins]

I can't count how many times I've approached poor Higgins out of the blue with the stark and foreboding message "Something is broken." I think he has grown to loathe those messages because they are usually followed by me attempting to completely upturn or rewrite some element of the book he thought was settled.

We got the new variable TN system out of that, which I'm really happy with. So, no complaints.

On the other hand, from a GM perspective, it would at least require a tiny bit less computation in play (you wouldn't need to think through "what is the right attribute to pair with this.")

Counterpoint could be argued from a player perspective where someone could argue for another attribute to be relevant.

Are dots as important as Higgins says?

My two main likes about them are:
a) flexibility of combinations
b) such visual representation goes along way in taking away that excel spreadsheet look most detail oriented games have

[Korbel]

Yeah, I've noticed this when I was rolling Oratory in our game. It was Social 2, Oratory 0 and SA 3. More dice from the Story Aspect, than from Attribute and Skill. Feels a bit awkward.
And now it would be 4 from the SA and still 2 from Att/Skill.
So, my gut feeling is the same as yours.

I've been thinking about the range 1-10 for Attributes before...

We'd have to toss the X+Y system of attribute/skill interaction

Actually, why? Just "double" the table, where the Obs are explained, and you're done. You know, so the most difficult task is at Ob 20. I know it would not be handy in use, but...
(well, you could rule that we only use odd numbers for Obs, range 1-19, so you still have 10 "steps" of difficulty)

Strength and Stamina would need to be balanced as an additional step for the purposes of DR and AV in combat. You'd need to halve them when you put them on your sheet. Dragon Tap and Naked Dwarf are known issues in TROS that stemmed from the difference between Strength and Stamina (or whatever the stat was in TROS) being too swingy.

As for the Naked Dwarf Syndrome, I think there's no better solution, than halving St/Sm in application for combat. Or take the drastic way and just stop calculating wounds from Strength and Stamina. Just the MoS vs AV. I know it's not what you are trying to model, but... well, it's very simple. Than you could make Strength more important in other aspects, grapplin or something, to compensate.

PS. Yeah, I love the dots, too.

[Agamemnon]

Actually, why? Just "double" the table, where the Obs are explained, and you're done. You know, so the most difficult task is at Ob 20. I know it would not be handy in use, but...
(well, you could rule that we only use odd numbers for Obs, range 1-19, so you still have 10 "steps" of difficulty)

The biggest reason is the purely practical one - number of dice. If we double the ranges but keep the X+Y arrangement your average pool is going to be 12 dice. The high end is 25. That seems like far too many d10s to comfortably throw around in hand. There's a reason the existing rules start just granting auto successes after you hit 10 dice.

[thirtythr33]

In my opinion this is feature, not a bug... as long as it doesn't get out of control. I love the idea of a mother being able to lift a car off her child, strength score be damned! Really, the way to keep it under control isn't in limiting how powerful the SAs are, but in limiting how often you get to apply them. If firing SAs for skills is easy then it's going to make it feel like a super-heroes game when people are firing left and right at anything remotely related to the main quest. If it is very difficult to fire, you only get the SAs when your life is on the line, it feels much more believable.

Outside of combat, they cause ... weirdness.

I tend to agree, but for another reason. Already in the Floating City game I have found it hard to justify allowing the players to apply SAs to intellectual skills like knowledge and memory tests.

My gut feeling is that this short-changes people who have actually spent the time and points investing in those skills.

Hmm.... Kind of. Someone who has invested will still likely beat someone who hasn't any time they either both or neither have SAs firing. These cases you bring up are quite specific; how often are two PCs going to be using opposed rolls on each other, where one has a significant SA advantage and the other has a significant Skill advantage? In the interest of narrative flow, I think the SA having power in these situations makes sense. In 95% of cases SAs are firing against NPCs (who don't use SAs) and SAs are inherently temporary; making them more about "spotlighting" a character than about "overshadowing" other players. Whoever has the highest SA will naturally drive play, until it ebbs and flows to another player. Sure, the fighter might be as good as the rogue at sneaking right now when he needs to get past the dragon to save his darling, but next week hes going to back to his oafish self. Usually the rogue is going to be happy the fighter can come sneaking with him for once. Again the problem would seem to be when someone has an SA so broad that the damned thing never turns off.

Another thing you may need to watch out for is someone who intentionally dumps skill points in character creation and then makes SAs like "I want to be the best sneak in the land". That's abusing the system and bad form.

It's the nature of the die pool...So what can be done?

So there is a lot of things going on here. One thing I like about the way the system is currently set up is that you have skills and combat using the same Ob levels. Since combat uses the split dice pool mechanic, this means that the skills dice pools have to be half the size of a CP on average. You correctly point out that makes firing SAs twice as powerful on skills as in a combat pool by ratio. To bring these to parity you can either:
a) Half the effect of SAs on Skills. I would do this by adding 'soft caps', for example An SA cannot increase any pool to more than twice it's starting size. This leaves CP untouched and neuters the cases of someone with 2 Agility and 1 Sneak firing for a full 5 SAs. OR
b) DOUBLE the effect of SAs on Combat Pools. This is probably best done by simply saying the SA applies to EACH pool after it is split or by instead comprising the pool by your TWO highest pools, just as it uses your TWO attributes Cunning and Agility. Of course, if you think that SAs already have too much effect on Skills this is just "breaking" combat to match.

Don't throw out the Attribute + Skill system. It is intuitive and useful. It would also break symmetry with how combat pools are calculated, which use Attribute + Proficiency.

If there is a problem in all of this, I think it has more to with how to write good SAs and the frequencies they should apply and less to do with the balance of PVP conflicts. Assuming players are working towards a common cause and SAs are not firing 24/7, I don't see the skewed numbers being much of a problem. Maybe a clause somewhere that says "An SA always applies in Combat, because it is inherently potentially deadly. SAs only apply to Skills when something as important as life and death to the character is on the line."

[Korbel]

The biggest reason is the purely practical one - number of dice. If we double the ranges but keep the X+Y arrangement your average pool is going to be 12 dice. The high end is 25. That seems like far too many d10s to comfortably throw around in hand. There's a reason the existing rules start just granting auto successes after you hit 10 dice.

Clearly, I got used to rolling on Roll20, where you just type in the number of dice - and forgot about this issue
Anyway... 25 is extreme. 10-15 would be common, sometimes reaching 20. I just rolled 15 dice, using a bowl, and it was not really troublesome. And it's still takes less time, than doing one Phrase of melee combat.

[Agamemnon]

The biggest reason is the purely practical one - number of dice. If we double the ranges but keep the X+Y arrangement your average pool is going to be 12 dice. The high end is 25. That seems like far too many d10s to comfortably throw around in hand. There's a reason the existing rules start just granting auto successes after you hit 10 dice.

Clearly, I got used to rolling on Roll20, where you just type in the number of dice - and forgot about this issue

My group being mostly online, I can be guilty of the same. One of the tragedies of my current situation is that I've never actually gotten to play this game in person at a table. A shame, truly.

I love the idea of a mother being able to lift a car off her child, strength score be damned

Agreed! Though that isn't necessarily lost in tinkering with the attributes and skill ranges. It wouldn't actually lessen the impact of SAs so much as make the individual skills and attributes more important, as they couldn't then be compensated for with attributes, etc.

Really, the way to keep it under control isn't in limiting how powerful the SAs are, but in limiting how often you get to apply them.

Henri and I discussed the same idea when we were talking about this in private, but the counterpoint was that you weren't then necessarily fixing the issue so much as making it come up more rarely. When it did come up, however, it was still a thing.

Already in the Floating City game I have found it hard to justify allowing the players to apply SAs to intellectual skills like knowledge and memory tests.

Most SAs work out like some kind of goal, or at the very least, when they are employed they are used in the pursuit of a goal related to the SA. At least as I had originally conceived them, you get your SA dice whenever you're doing something that furthers the goal.

Let's say your SA is "For Love of Gwen." There's no direct goal there... but on the other hand, if you were trying to get her to marry you... or defending her honor from a base scoundrel who had besmirched it.. or she was kidnapped.. Suddenly there are goals implied. Get her to marry you, beat the snot out of the offending cur, rescue her from her captors. On the face of it, any roll that will push you forward to the implied goal would be fair game for an SA to apply. Certainly the seduction roll in the first and the combat pool in the second. But what about the third? Does a connections roll to find a contact in the underworld count, if they might have information on her whereabouts? What if I wind up in jail? I can't free her from jail. Do I get the dice for trying to escape in order to free her? What if I'm only in jail as part of an elaborate plan to free her? This whole thing gets very murky.

Hmm.... Kind of. Someone who has invested will still likely beat someone who hasn't any time they either both or neither have SAs firing. These cases you bring up are quite specific; how often are two PCs going to be using opposed rolls on each other, where one has a significant SA advantage and the other has a significant Skill advantage? In the interest of narrative flow, I think the SA having power in these situations makes sense. In 95% of cases SAs are firing against NPCs (who don't use SAs) and SAs are inherently temporary; making them more about "spotlighting" a character than about "overshadowing" other players. Whoever has the highest SA will naturally drive play, until it ebbs and flows to another player. Sure, the fighter might be as good as the rogue at sneaking right now when he needs to get past the dragon to save his darling, but next week hes going to back to his oafish self. Usually the rogue is going to be happy the fighter can come sneaking with him for once. Again the problem would seem to be when someone has an SA so broad that the damned thing never turns off.

Ha.. That example was actually from the demo adventure The Traitor. It featured a pretty decent number of PvP elements, with the player characters all having slightly different motivations and desired outcomes. The bit that makes this really tricky is the assumption about whether or not SAs are firing. In the games I've played, I've found that I very rarely have a reason to call for a roll on something where no SAs apply. When I do, it's usually Perception or some form of knowledge and that's just to tell how much information I should give the players about something. It's less that the SAs are so broad that they always apply, so much as that if something has stakes sufficient enough to warrant a die roll, there are usually SAs on the table driving it.

a) Half the effect of SAs on Skills. I would do this by adding 'soft caps', for example An SA cannot increase any pool to more than twice it's starting size. This leaves CP untouched and neuters the cases of someone with 2 Agility and 1 Sneak firing for a full 5 SAs. OR
b) DOUBLE the effect of SAs on Combat Pools. This is probably best done by simply saying the SA applies to EACH pool after it is split or by instead comprising the pool by your TWO highest pools, just as it uses your TWO attributes Cunning and Agility. Of course, if you think that SAs already have too much effect on Skills this is just "breaking" combat to match.

Thought about this, but multiplying/dividing SAs on the fly seems messy and inelegant.

Don't throw out the Attribute + Skill system. It is intuitive and useful.

I enjoy it, as a whole. It does pose a tradeoff, though. Both in the current issue of SAs being able to replace 50% or more of your die pool and in that combining attributes on the fly requires more mental horsepower. As higgins pointed out, it could also invite some debate at the table when the player thinks attribute X is more applicable to the situation than attribute Y (further complicated if their character is better at attribue Y).

Assuming players are working towards a common cause

Heh. heheh. Heheheheheh...

Maybe a clause somewhere that says "An SA always applies in Combat, because it is inherently potentially deadly. SAs only apply to Skills when something as important as life and death to the character is on the line."

This is much harder than it appears at first blush. Look at the example above with the kidnapped lover. Something as important as life and death is on the line, the problem is judging how far in the service of that task you can still use its dice. Of course, you could look at this in reverse as well. If someone's SA was not necessarily life and death on its own, but still important to the character "Make peace with my father" how do you define when it can come up?

[thirtythr33]

Really, the way to keep it under control isn't in limiting how powerful the SAs are, but in limiting how often you get to apply them.

Henri and I discussed the same idea when we were talking about this in private, but the counterpoint was that you weren't then necessarily fixing the issue so much as making it come up more rarely. When it did come up, however, it was still a thing.

Here I disagree. If we both agree that 1) a weak person should not normally be able to lift a car normally and 2) a weak person should be able to life a car off their child, then the very issue itself is the frequency we let SAs apply.

In the games I've played, I've found that I very rarely have a reason to call for a roll on something where no SAs apply. When I do, it's usually Perception or some form of knowledge and that's just to tell how much information I should give the players about something. It's less that the SAs are so broad that they always apply, so much as that if something has stakes sufficient enough to warrant a die roll, there are usually SAs on the table driving it.

This is a very good point, especially since you make it a rule not to call for unimportant rolls. Maybe we should assume the player ALWAYS has an SA firing and set up the Obs accordingly? If they don't have the SA, they probably need to seriously consider just running the hell away because they are going to fail.

a) Half the effect of SAs on Skills. I would do this by adding 'soft caps', for example An SA cannot increase any pool to more than twice it's starting size. This leaves CP untouched and neuters the cases of someone with 2 Agility and 1 Sneak firing for a full 5 SAs. OR
b) DOUBLE the effect of SAs on Combat Pools. This is probably best done by simply saying the SA applies to EACH pool after it is split or by instead comprising the pool by your TWO highest pools, just as it uses your TWO attributes Cunning and Agility. Of course, if you think that SAs already have too much effect on Skills this is just "breaking" combat to match.

Thought about this, but multiplying/dividing SAs on the fly seems messy and inelegant.

What about my suggestion of capping SAs to no more than doubling the pool? This doesn't require calculation, only comparing sizes.

Already in the Floating City game I have found it hard to justify allowing the players to apply SAs to intellectual skills like knowledge and memory tests.

Most SAs work out like some kind of goal, or at the very least, when they are employed they are used in the pursuit of a goal related to the SA. At least as I had originally conceived them, you get your SA dice whenever you're doing something that furthers the goal.

Let's say your SA is "For Love of Gwen." There's no direct goal there... but on the other hand, if you were trying to get her to marry you... or defending her honor from a base scoundrel who had besmirched it.. or she was kidnapped.. Suddenly there are goals implied. Get her to marry you, beat the snot out of the offending cur, rescue her from her captors. On the face of it, any roll that will push you forward to the implied goal would be fair game for an SA to apply. Certainly the seduction roll in the first and the combat pool in the second. But what about the third? Does a connections roll to find a contact in the underworld count, if they might have information on her whereabouts? What if I wind up in jail? I can't free her from jail. Do I get the dice for trying to escape in order to free her? What if I'm only in jail as part of an elaborate plan to free her? This whole thing gets very murky.

Maybe a clause somewhere that says "An SA always applies in Combat, because it is inherently potentially deadly. SAs only apply to Skills when something as important as life and death to the character is on the line."

This is much harder than it appears at first blush. Look at the example above with the kidnapped lover. Something as important as life and death is on the line, the problem is judging how far in the service of that task you can still use its dice. Of course, you could look at this in reverse as well. If someone's SA was not necessarily life and death on its own, but still important to the character "Make peace with my father" how do you define when it can come up?

As it stands now, it is very murkey where to apply the SA and where not. But it could be made explicit and simple, using the systems you already have in place:

A Story Aspect will fire only if there is something at Stake in the Conflict that relates to that Story Aspect. If a player invokes an SA and fails, any complications introduced will be related to that Story Aspect.

By linking it directly to the Stakes instead of the wishy-washy "directly further, uphold or defend" it makes it more firm. If it doesn't hurt losing, it doesn't fire. So instead of looking at the SAs as "does accomplishing this get me closer to my goal?" we look at it as "does FAILING this make it HARDER for me to accomplish my goal?". So now we see that if I have the SA "For Love of Gwen" and I am locked up in jail and the Stakes are "do I bust out without being seen?", then the SA doesn't fire because the Stakes have nothing to do with Gwen. If I am locked up and the Stakes are "do I bust out before the bandits take Gwen to another city?" then yes, the SA fires. If you fail with the SA "For Love of Gwen" firing, a complication involving that SA should be introduced... When you get to the rendezvous, Gwen is nowhere to be found. You could then even leave it up to the player to decide whether or not they want to turn on their SAs, since it is now a hard choice to make whether they want to risk those shitty complications for some bonus dice. The idea would be to make SAs firing less often and more important when they do.

[Agamemnon]

If we both agree that 1) a weak person should not normally be able to lift a car normally and 2) a weak person should be able to life a car off their child, then the very issue itself is the frequency we let SAs apply.

While it makes for a dramatic and interesting example, consider this in other circumstances. A novice with a single dot and SAs firing has the same chance of playing the most difficult Bach compositions as a master does without - in fact, a better chance. At a certain point, it begins to strain the credibility of the fiction.

The problem isn't that SAs improve your chances of doing something beyond what they are capable of ordinarily doing. It's that due to the 0-5 scale, the dice from SAs can make your actual skill in the thing effectively irrelevant. More dice is always better than less, but with +5 dice from an SA your actual skill can be 0 and you can still pull off any ob that you would have been able to pull off without SAs.

For the sake of argument, compare to the effect if we put it on a 0-10. A +5 from SAs still grants you a massive improvement, but if you have nothing in the skill you're still going to have a hard time beating someone who has mastered it.

Maybe we should assume the player ALWAYS has an SA firing and set up the Obs accordingly? If they don't have the SA, they probably need to seriously consider just running the hell away because they are going to fail. /quote]
This has the potential to help deal with the Ob curves. Make anything that isn't a completely routine task have an Ob sufficiently high that it practically demands an SA firing. This won't help PvP scenarios though, unfortunately. Still. Part of a potential solution.

What about my suggestion of capping SAs to no more than doubling the pool? This doesn't require calculation, only comparing sizes.

A possibility. Still a bit fiddly. Every additional operation you tack onto something is a thing that slows down gameplay. But it's something to kick around.

A Story Aspect will fire only if there is something at Stake in the Conflict that relates to that Story Aspect. If a player invokes an SA and fails, any complications introduced will be related to that Story Aspect.

This actually gives me a bit of an idea. It's half-formed, but I'm chewing on it:

What if we scrapped the current complications system as a default. Made it that instead, by default, when you make a roll and failed, you don't provoke complications or anything. Things just are what they are. Then we declare that players can choose when they want to bring an SA into it.. BUT if they do, and it fails the current complications chart is in play.

This combined with an adjusted Ob chart might go some direction towards addressing the balance, as players will be invoking SAs less often. It might also mean in PvP scenarios (particularly non-lethal ones), there will be occasions in which one or both players might decide the circumstances are not dire enough to want to call on SAs and risk the fallout.

(It occurs to me that this might actually be what you intended in your response. If so, then credit for the idea goes to you and I'm just obtuse.)

[thirtythr33]

While it makes for a dramatic and interesting example, consider this in other circumstances. A novice with a single dot and SAs firing has the same chance of playing the most difficult Bach compositions as a master does without - in fact, a better chance. At a certain point, it begins to strain the credibility of the fiction.

Again, I think this is more to do with the situation the SA is applying and not the magnitude of the bonus. Is it reasonable for Strength to be boosted by a persons determination and resolve? Yes. Is an impassioned speech going to be boosted by a persons true convictions to the cause they are speaking? Yes. Is someone's ability to perform a complicated symphony increased because this is the "big night"? Not so much. Arguably, pressure would often have the opposite effect on such performances. Is someone's memory suddenly made sharper when a critical piece of information will save a loved one? Obviously not.

Perhaps the skills and attributes need to be broken down into separate lists where some can be boosted by SAs and others cant? Or you just give the GM veto power on when SAs can apply.

It occurs to me that this might actually be what you intended in your response. If so, then credit for the idea goes to you and I'm just obtuse.

It wasn't exactly, I was being even more specific in saying the Stakes and Complications have to relate specifically to that Story Aspect for it to be allowed to be used in play. But, I beat you to that idea by a couple of months anyway Here's a conversation me and Korbel had 2 months ago in the Floating City:

Memory test with or without my first SA (to do, what my father wants, which is to find the rapist) firing?

First, lets look at the rules.
p87 First, whenever your character is in a conflict that would directly further, uphold, or defend an SA they have, it is said to fire.
p14 Complications are the consequences of a conflict.
So I'll leave the choice as to whether this is a Conflict or not up to you.
If a Memory Attribute Check is a Conflict, then you get to apply the SA but you run the risk of creating a Complication. Otherwise, no SA and no Complication.
I would lean towards it not being a Conflict because p10 "Every conflict has two parts: task and intent." and I am having trouble separating the intent from the task when it comes to memory.
I also don't see how remembering something is "directly furthering, upholding or defending Marco's obedience to his father". To me that describes a situation where Marco has been directly tested or tempted to do something against his father's wishes, not every incidental conflict that comes up in carrying them out. If I allow the SA for memory, there is basically never going to be a single roll in the entire campaign that it wouldn't also apply to.
But basically, I'll leave it up to you to decide. Any roll you choose to apply your SA to will generate Complications on failures though.

Yup, I was leaning towards "no SA firing", too. But I was a bit confused by this:
"A character with “Win Circe’s Heart” at 2 dots would get 2 bonus dice to any pool that directly involved wooing Circe but they could also apply those bonus dice to any conflict that was meant to impress her"
It sounds like SAs fire not only in conflict.
But you're right: "f I allow the SA for memory, there is basically never going to be a single roll in the entire campaign that it wouldn't also apply to.".
Anyway, the Ob is quite high for my Memory score (even with SA firing... 4 dice, Ob 3 - not good), so I won't roll. It's too risky, my precious! And I'm not in dire need of this info.

Korbel, why not roll the dice? Without firing SAs, there are no consequences for failing a roll (I assume).

My memory is 2, the Ob - 3. And knowing roll20 I will get 0 successes and 33 will give me some false info
As I said, it's not very important, so I don't won't to touch roll20. There is some foul magic. I'm scared.

[Agamemnon]

While it makes for a dramatic and interesting example, consider this in other circumstances. A novice with a single dot and SAs firing has the same chance of playing the most difficult Bach compositions as a master does without - in fact, a better chance. At a certain point, it begins to strain the credibility of the fiction.

Again, I think this is more to do with the situation the SA is applying and not the magnitude of the bonus. Is it reasonable for Strength to be boosted by a persons determination and resolve? Yes. Is an impassioned speech going to be boosted by a persons true convictions to the cause they are speaking? Yes. Is someone's ability to perform a complicated symphony increased because this is the "big night"? Not so much. Arguably, pressure would often have the opposite effect on such performances. Is someone's memory suddenly made sharper when a critical piece of information will save a loved one? Obviously not.

Perhaps the skills and attributes need to be broken down into separate lists where some can be boosted by SAs and others cant? Or you just give the GM veto power on when SAs can apply.

Now we're into the nebulous territory of defining what the bonus from SAs actually represent. The only way we can argue that "woman lifts car," "person fights harder," or "speech is impassioned" makes sense but "perform symphony" and "boost memory" don't is if one tries to argue that SAs are supposed to represent some kind of in-world physiological response. That the increased strength or heightened fighting ability are adrenaline, that the impassioned speech was simply one's emotions getting involved -- and that boosting memory or symphonic abilities don't work because that's not how the brain responds to pressure.

If on the other hand we take the position that SAs are a purely metagame mechanic (as the current writeup would heavily imply -- see pg. 86, under Assigning SAs) then SAs have nothing to do with physiological responses at all. They are a measure of narrative weight - how much the flow of the plot has your back. On the face of it, an SA about beating your musical rival in a symphonic duel is no less valid than defeating your rival with a blade. From a story-telling perspective, why would one be deserving of bonus dice but the other not?

[thirtythr33]

You're absolutely right, I was mistakenly thinking of SAs as Passions or Beliefs are used in other games. But if we are thinking of SAs as purely narrative (and not a simulationist "adrenaline rush"), then their purpose is in replicating the kinds of inexplicable overcoming of insurmountable odds we see in fiction. If that is the case, then why would we have a problem with a novice beating a professional in a Bach performance? We see that happen in fiction all the time, even if it is utterly unbelievable. The difference, I think, is that in fiction we usually see that kind of overcoming at the very crescendo of a story arc that focuses on the skill in question, not routinely overcoming every challenge that presents itself along the way as can be achieved if SAs are firing for every roll.

[Agamemnon]

You're absolutely right, I was mistakenly thinking of SAs as Passions or Beliefs are used in other games. But if we are thinking of SAs as purely narrative (and not a simulationist "adrenaline rush"), then their purpose is in replicating the kinds of inexplicable overcoming of insurmountable odds we see in fiction. If that is the case, then why would we have a problem with a novice beating a professional in a Bach performance? We see that happen in fiction all the time, even if it is utterly unbelievable. The difference, I think, is that in fiction we usually see that kind of overcoming at the very crescendo of a story arc that focuses on the skill in question, not routinely overcoming every challenge that presents itself along the way as can be achieved if SAs are firing for every roll.

Emphasis, mine.

The problem of one of scale. 1 dot is described as "Novice - Basic familiarity, but no practical experience. An enthusiast."

It's easy to imagine a story in which the plot is a youthful musican trying to take on a grandmaster violinist to win the heart of some girl. Sounds plausible. If it was a modern movie, a good twenty minutes of it would be dedicated to some kind of musical training montage to build credibility to the idea that he could, in fact, try to hold his own against said violinist. You'd have the climax in which the competition is held, the SAs fire, and he wins! huzzah!

The problem is, due to the scale, he doesn't actually need a training montage. He doesn't need even the "beginning of an education in a given craft." He can literally have bought the student violin a couple days before and have learned things like "how to hold the bow," and then beat the master.. because he only really needed one dot in the skill, and that was just to avoid the disadvantage for using the skill untrained. Would you, as a reader, be able to suspend your disbelief so far that the person who put a couple days into learning the violin could beat the grandmaster with years (if not decades) of experience?

Most wouldn't.

The space between "novice who knows only the basics with no practical experience" and "grandmaster with the highest dedication to their craft" can be completely bridged with SA dice alone. Technically, a person with a novice rank and 5SA dice is mathematically the equivalent of a Savant for that roll.

Compare with the results if it were on a 0-10 range. At a 1 or 2, he's still not going to be able to beat the grandmaster. The skill gap is just too far to overcome. On the other hand, if he's closer to 5 or 6? well, now he's at least a journeyman. Per the current rules he'd be "competent bordering on professional. You understand what to do, but lack the hands-on experience and tricks of the trade." It's still not necessarily realistic, but at least now we can squint and assume that factors were at play and the heat of the moment and blah blah.

See where I'm coming from about scale, at least?

[thirtythr33]

The problem is, due to the scale, he doesn't actually need a training montage.

So if the character is a novice violinist and takes the SA "beat the champion violinist at the big competition" he is going to start with 1 SA in it. He has to go through his training montage and other violin-related sidequests in order to build his SA up to 5 to be able to beat the master. This is working as intended, I think.

The problem I think is when we have an adventurer with "Win Gwen's love" SA and he does a bunch of things to impress her, building his SA up to 5. Then, at the end of the story arc the rival challenges him to a impromptu fiddle competition and despite building the SA with nothing fiddle related, the novice beats the master at his own game.

Isn't the fix here in restricting what SAs can be boosted and applied under what circumstances?

See where I'm coming from about scale, at least?

Yes. It's because Skill only contributes to 33% of the strength of a Skill test. Attribute and SA also contribute 33% each. It is related to the problem DND 5e has with novice vs master skills. A novice has a +1 on a d20 and a master has a +5. The novice still has a 35% chance to best the master because the randomness of the dice contributes to 75% of the result. a 100% skill contributing test would be one where "the higher skill wins". Also remember that the novice with full SA versus the master is still a 50% - 50% contest. That is actually pretty terrible odds to win considering the medium we are talk about. Imagine if a dragon boss fight was actually a 50% chance to lose. In most role-playing games even a big boss fight is probably only 20% to lose. Players are SUPPOSED to have the deck completely stacked in their favour.

Here's another thought. Maybe Skills and Attributes SHOULD get a bigger bonus from SAs than combat?

A skill test or attribute check is usually a single roll ordeal. You either win your opposed Bach performance or you don't. If you flub one roll, that's it, its over. You lose.

A combat comprises possibly dozens of rolls. If you flub the first roll, you are disadvantaged but you still have the ability to make a comeback. You get to use your SA bonus numerous times and the more rounds the combat takes, the more the SA bonus compounds making it more likely you win.

If we imagine a skill contest where two characters are 60% and 40% to win on an opposed roll, then they are 60% and 40% to win the contest as a whole.
If we imagine a combat where two characters are 60% and 40% to win an opposed roll and assume that it will consist of 5 opposed rolls, then they are 68.2% and 31.7% to win atleast 3/5 opposed rolls. The snowballing effect makes it more one sided but the math for that is too complicated for me to do off hand.

To keep the actual win rates at parity, the skill tests need a larger modifier.

[taelor]

The effect is even more pronounced when making unopposed checks. The way the Obs are structured, +5 dice from an SA are +2.5 successes at TN6. If we start talking about exploding dice, things can get even more goofy. If you set the difficulty scale for skill/attribute checks without SAs factored in, then SAs will make difficult things relatively trivial. If you set the difficulty scale in order to account for the influence of SAs, you make difficult checks impossible when they aren't firing.

This latter option is what is what Burning Wheel does with it's Artha system, along with the expectation that the players will fail around half of all their rolls (but that's okay because failure is fun).