I must admit I meant my suggestions as alternatives, but sure, you could combine them. Or drop the bonus to +1 per pawn as suggested above, which might be safer for experimenting (by players not characters).
Yes, I had been wondering whether to limit the number of pawns you can use at a time (beyond the MT limit), but the botch deterrent is big. Except for mad Tytali, who are self-limiting anyway, because they explode.
And that’s where the problem is. When you get the rare blowout stress result it doesn’t matter what the bonus you added to it was. So the bonus was pointless, and you needn’t have wasted the 5 pawns. If you don’t get the blowout stress result you’re out 5 pawns to get a result comparable to a really good book… but the book still isn’t wasting vis nor risking Warping/Twilight.
Requiring more vis to be spent just to be on par with the xp studying a book makes the inequity worse because of the swingyness; you’ve now made it a bad gamble for a small but possible % of the time.
I do like @Tugdual’s idea of roll & keep, though.
I never said it wasn’t swingy, just that its not any more swingy – Unlike the roll multiple times, keep best, which because of the nature of stress die is more swingy.
If we really want to make twilight more likely the obvious answer is to roll more stress die to increase the chance of botching, but that would also make it more swingy. Stress die + Stress die … + aura
Functionally, you can’t. The Covenants view of how the Order works is fundamentally at odds with all the published Tribunal books. Which is why people on this forum get mad about the Terram Branch in Normandy being “bad” by Covenants standards.
The basic problem with vis study is that it has a risk of catastrophic botching, while buying a bunch of books doesn’t. Add to that the fact that books are extremely cheap (especially when you consider that they can be used by other members of the covenant) and you’d have to make vis study much, much better than it is to make it make sense.
And the current rules already push magical advancement farther and faster than anything before.
I’m thinking “roll many keep best” would be massively overpowered in the AM system. Consider person A is rolling one die and keeping the results, person B is rolling 4 dice, and person C is rolling 10.
Person A has a 1 in 10 chance of botching, and a 1 in ten chance of exploding
Person B has a 1 in 10,000 chance of botching (all 4 dice must come up zero) and a 34.39% chance of exploding (only 1 needs to be a 1)
Person C has a 1 in 10,000,000,000 chance of botching, and a 65% chance of exploding.
The big question becomes what happens when the dice explode, because if you keep the same number of dice for the reroll (which in principal makes sense) then person B has an 11% chance of double exploding- higher than person A’s chances of a single explosion, while person C has a 42.4% chance of a double explosion, a 27% chance of a triple explosion, an 18% chance of a quadruple explosion, and an 11% chance of 5 sequential 1’s. given that 5 sequential 1’s is 32x the next roll (also likely to be high given 10 dice) this probably goes far beyond simply multiplying the result by 10 for 10 pawns.
“roll many keep best” is each die is seperate. You wouldn’t roll 10 die and if one explodes roll 10 more. If one explodes you roll 1 more, for that die. If two die explode, you’re exploding two seperate die and they will get independent results.
The first thing about the probabilities is absolutely correct, unless you made it expensive to add more die, it does become nonsense very quickly.
There’s more than one way to skin a cat. The trick is finding which one makes sense.
For instance, you could count dropped dice for botch and be both successful and botch. Your person B:
The average result gain for extra dice (5.5, 7, 8, 8.5) quickly tapers off, the botch risk makes it not worth it in principle. As a magus, I wouldn’t roll 4 dice for 4 pawns of vis but I’d certainly use 2 dice.
Well, I wouldn’t waste 4 pawns for less than 20 xp. So keep best ain’t enough be itself to solve Vis study.
Counting dropped dice for botches would seem to contradict the idea of “keep the highest”. Alos “succeed and botching” would be redefining what a botch is…”on a botch the total is always zero”- unless this has changed in the DE means you cannot succeed on a botch.
The point of the OP was to make warping more of a risk. The only way for this to work would be, if at least one zero is rolled (if 2 zeros were rolled, as there are already bonus dice on the botch, making multiple botch tests seems overly harsh), a botch must be tested, and if a botch is rolled, a botch happens regardless of the other dice.
One advantage of study from vis is that you can do it anywhere. Therefore while you can do it in a lab to benefit from all of your safety bonuses, a wandering magus could do it anywhere with a friendly aura and set up camp. Your covenant on a ship or in a wagon train could stop in somewhere with a magical aura for winter, and you can study just fine. Static covenants with libraries are a choice - the most common and most convenient choice, but not the only one.
The perceived issue is that vis study is not rewarding enough compared to books to be worth the risk of big botching and the vis consumption, increasing the risk of botching is going to result in it needing to be even better reward-wise to incentivise people to do it. Just as rolling more and keeping the highest moves you towards inevitably exploding the result, letting botches still occur with unkept die moves you towards inevitably botching, which means no warping occurs because players will just not do it.
To get more warping to happen I think you need to nerf books in some way (knock off a couple of points of level so getting above 15 requires vis study, restrict availability of tractatus, delete book learner, etc) and buff vis study rewards so it ever happens.
I’m not even sure saying that the vis doesn’t get consumed if you botch, so you just get 0xp + warping that season would do it.
But if twilight sucks less than the downsides of vis study are also made less severe. We’ve identified a number of interacting rules; changes need to be considered together.
How about the following change set:
Honestly if you want to encourage studying with vis one of the first things I would get rid of is the “lose xp” outcome. I mean I spend a few pawns of vis, lose a season, and lose experience? Where are the books at?
Is the solution something along the lines of “If you enter twilight while studying from Vis, you automatically get +this or that to understand Twilight” so that, on some level, botching almost becomes a perk of the experience? You’re maybe even AIMING for a botch, to accept a few warping points in exchange for a bunch of xp and maybe some spells/virtues?
Since we want to increase the chance of twilight, we shouldn’t make it harder to dobule botch or avoidance easeir– realisticly we need to make avoidance harder because as is its really easy. Since we need it to be risky, we can’t remove all the risk.
It just needs to be a sane risk, at least for high art scores.
If you botch, you can enter Twilight and reroll to gain xp.
Make twilight not suck so much. Make it much easier to control, or make their always be a good side effect, even if there is a bad side effect. People master spells, rarely use vis to boost spells, avoid casually sponting too much,don’t study with vis, etc, because twilight is bad!
They were the Libri Quaestionum, introduced in ArM4. Basically, they had an audience target level. They were good for a readert at that level, but penalized the reader more and more heavily the further he was from that target level. In general, I think this is a good idea that makes things more realistic, if a little more complicated.
The other side of the story is that knowledge branches out, so a book on some relatively unexplored aspect of Perdo might be useful to both the newly gauntleted magus and the seasoned Flambeau archmage, and that's the rationale behind the ArM5 Tractatus.
For what's worth, Twilight was reworked in 5th edition exactly to this end: to allow magi to have more Twilights, and for the results to be more flavourful and less of a death sentence. I think the authors largely succeeded in this but were not, perhaps, bold enough to go further: most of the problems that people bring up with the current rules were present in previous editions too and were in fact typically harsher.
Ultimately, one has to decide what's the purpose of Warping and Twilight in the game. Note that these were not 'core concepts' of Ars magica and, if I recall correctly, only entered the game in the celebrated 2nd edition supplement The Order of Hermes (so 3rd edition was the first to have Twilight in its core rules). In this sense it's useful to distinguish
a) twilight episodes, the 'vision quests' so to speak, that a magus undergoes when facing particularly powerful/uncontrolled magics (a trope of fantasy literature) and
b) progressive warping, including twilight scars, time skipping, and advancement towards some form of final 'exit' from the game. The idea here is that relying more and more on magic inescapably and irreversibly increases a magus' distance from (more) mundane humans, until he fades away. On a setting level, this limits the sway of older magi on both the world and younger magi. On a character level, it should serve as an 'aging' marker that makes a character neither more powerful nor weaker, but simply shifts him sideways (and eventually out of the picture) giving the player a sense of progression and fun while maintaining balance.
Personally, I think a way to improve the current design of Twilght would be to better decouple a) and b), so that a) is more freeform and does not automatically cause b). This allows much more freedom for the SG to slap a twilight episode on a magus whenever it feels thematically appropriate without feeling unfair, and creates an environment where characters can actively seek twilight episodes (solving a long-standing multi-edition problem with the Criamon, who are supposed to seek Twilight as a mystical experience but ... ).There are so many ways twilight episodes can be implemented that I'll simply stop here as far as they are concerned.
As for progressive warping, I think the core of it should be something like the following:
Depending on what you are trying to encourage there are a number of possible changes. A bonus to comprehend twilight when studying from vis for example would clearly invite more studying form vis (maybe adding the form bonus for the type of vis being studied). A possibility would be some form of otherwise mundane potion (mushroom wine from a rare mushroom?) that would induce twilight when drunk without adding warping points, and perhaps a ritual that could accompany it to boost the chances of comprehention.