Linear Wizards
Magic Dice are weird. GLOG wizards have almost no reason to cast spells with more than 1 MD!
Assuming the common standard of "deal [dice]+[sum] damage to any target", that's exactly the same total damage if you cast 1 4MD spell vs 4 1MD spells, but the 1MD spells have no chance of doubles or triples, whereas the 4MD spell has a ~72% chance of rolling doubles for a mishap and ~10% chance of rolling triples for a doom!
Sure you get 4x the DPS but that benefit is nowhere near as fast scaling as the combinatorics involved in the drawbacks.
And if time isn't a concern, you can expect to roll each MD twice in a day since it exhausts when it rolls in the upper half of its range. (That's the expected number of tails you'll flip in a row.) It doesn't matter if you roll the dice together or for individual spells, it's gonna be around 2 rolls per die. So why not take it one at a time? It lets you split up targets.
Plus the actual standard is "deal [dice]+[sum] damage to any target plus a small benefit," and if that benefit doesn't scale off of [dice] or [sum] then it's not a non-scaling effect, it actually secretly scales based on number of casts.
There are absolutely spells that avoid this problem don't get me wrong. Increasing areas of effect based on [dice], for instance. I'm just saying [dice]+[sum] on its own is a bad standard. [dice]x[sum] is really fast scaling though so that's scary.
What are some ways to fix this?
- Maybe a maximum of one die can be exhausted per spell cast? I haven't worked out the exact math but it would definitely make it more efficient to cast bigger spells, protecting your MD pool from exhaustion.
- [sum]+[dice²] is quadratic but a gentler curve than [dice]x[sum]. It makes [sum] irrelevant when you get to really high MD casts though.
- The [doubles] keyword I've seen some people propose is elegant in some ways, but doesn't fit every spell since it's not guaranteed to be nonzero. It's only useful for very particular kinds of effects.
- I had been fond of the [choose] keyword. A spell using it comes with a d6 table, and the caster picks the result of one of their rolled dice and goes with that entry in the table. So a 1MD spell is leaving it to chance, while increasing your MD gives you more control. But... that doesn't actually work as an incentive, does it, looking at the 4x1MD case? 4MD lets you pick one of their results from the table, but if you just cast it 4 times with 1MD you will get the one result you would have picked anyway plus 3 bonus effects that wouldn't have resolved at all had you cast it as a 4MD.
- Add back Vancian casting. (I think some people still play with it in GLOG and maybe this whole debacle is just me finding out why.) Once you cast a spell, you can't cast the same one until you prepare it again, independent of MD exhaustion.
- New keyword idea [buy]. It's like a little shop tacked on the end of a spell where you can spend points from [sum] and [dice] for more effects. 1 point: spell smells like roses. 3 points: spell is invisible to one of the five senses. 7 points: 1 additional target. 15 points: target shits their pants. I don't know. Seems like a lot of nonsense to add to a spell, bad game design probably to slap it on everything. But I think it is good game design to add a bunch of nonsense to maybe 2 or 3 spells.
- Mishaps and dooms in my mind are kind of iconic GLOG wizard features but maybe it's best just to ditch them if they are purely negative and the only part of the wizard that scales superlinearly? That's pretty much the crux of the issue IMO. Or maybe they can stay but they shouldn't be purely negative? What if rolling a doom gives you an extra MD forever?
Ok that's all I wrote for this blog post thank you.