Tag Archives: Simulation

On Random Number Generation, Continued

One of the things I like about d20 systems is that doing math relating to a d20 is really easy. There’s a 5% chance to roll any given number, so probability calculations can be done quickly without having to get out a calculator. That said, D&D does have a fair bit of math for which a calculator can be helpful. Usually I resort to anydice for complicated math, because I believe in making informed decisions. The most obvious thing here is the Advantage/Disadvantage system 5e adds.

More Like “This Advantage”

Advantage and disadvantage were among the first new mechanics revealed in playtesting for what was once called D&D Next. If you’re unaware, having advantage (or disadvantage) on a d20 roll allows you to roll two d20s and take the highest (or lowest) result. Mathematically, how many pluses or minuses this works out to depends on what number you needed to hit; it’s exactly equal to +5 or -5 if you’re looking at a target number of 11, decreasing as you need rolls that are farther from “average”.

advantage graph
The nature of this bonus makes sense if you stop and think about it. If you only need a low number, then your already high chances of success aren’t increased much by rolling a second die. If you need an extremely high number, your chances of success still aren’t very large. The system math keeps things mostly in the middle, so estimating it as +/-5 works in most cases.

Let Justice Be Done

I looked at it when examining fighter features, but paladins also get to pick a fighting style (but from a reduced list consisting of Defense, Protection, Dueling, and Great Weapon Fighting). It turns out that Paladins don’t value these equally, as great weapon fighting would appear to apply to smite damage, making it way out of line.

Normally, GW Fighting is an average boost of 1.33 damage per hit normally, but if you start adding smite damage this changes, getting better and better as you roll more dice. Jury’s still out on whether or not this works, so check with your DM. (Speaking of checking with your DM, it turns out the dueling style is intended to apply to sword and board. Point them at this if they say otherwise, but remember rule 0.)

On Random Number Generation

Blaugust is over, but that doesn’t mean my posting is going to stop. I’m aiming for Tuesday/Thursday, with maybe an extra post on the weekends at times. This week is a bit weird due to Labor Day.

This past Monday was the first D&D session with the new group, starting with being captured by a very powerful wizard and then escaping from the bandits and trapped dungeon she left to guard us. The party is now headed to Watchwall for various reasons, united primarily in the goal of figuring out why we were captured in the first place.

My virtual dice primarily failed me: I did not manage to hit anything with my Paladin’s giant stick at any point and rolled a 2 on survival (a skill I’m allegedly proficient in) in trying to figure out where we’d been taken. My track record with physical dice isn’t much better, I still remember rolling a 9 on a 6d6 Flame Strike. There were some successes: I rolled max damage on dragon breath and a very high persuade roll, earning us a break in the fighting until someone else put a dagger in the back of the person we were attempting to get information out of. (Even as a Paladin, you can’t save everyone.)

Like the Mountains of Illinois

D&D 5 is very flat, and the difference between being good at something and not good at it is not very large, especially at low levels. Previously (in 3.5) full ranks in a skill meant level+3, or +4 at level 1, going all the way up to +23 at level 20 (but probably much higher than that from synergies and magic) which meant that some skills had unusual rules interactions at high levels, and the difference between trained and untrained was impossible to overcome on any dice roll. By contrast, proficiency bonus is +2 at level 1 and only goes up to +6 at level 17, which means that even an untrained character can make certain rolls (albeit with a flat 30% lower chance of success). Rogues and bards can get double proficiency to certain rolls, so they can get up to +12, but they’re intended to have outlandish values on skill rolls as part of their class design.

In general, I like this approach, but I can see how some people would not. It does lead to a lot of chance inherent in most actions, as it’s next to impossible to have anything be a “sure thing” (although you can get close in some special cases). The trade-off for this is that untrained actions have an actual chance of success without rolling 20. (As a side note, this edition doesn’t give 20 or 1 special effects on skill checks or saves. They’re just numbers.) The highest suggested DC for anything in the player’s handbook is 30 for tasks that are “nearly impossible”. A non-bard/rogue with a 20 in the relevant stat at level 17 or higher would only hit such a target number when rolling 19 or better, so that description is fairly accurate.

I was going to discuss how I use anydice to sort out numeric things, but I think I’ve rambled on long enough. Maybe I’ll get to that next week.