D&D 5E Looking for inspiration for D&D data projects

[Guest 6801328]

I had fun creating my Rogue Elven Accuracy data visualization, and want another cool project. Any ideas?

I can either work with generating data from the rules, or if you know of interesting data sets (I really wish D&DB had an API for extracting usage data) I can work with that.

 

[CodeFlayer]

Given that the problem space is gigantic, perhaps give us more criteria? What makes a data set interesting? A natural next step might be damage per second extended to other classes against different targets. Of course, this has probably been exhaustively done elsewhere, which might or might not dissuade some.

Good luck!

 

[Fenris-77]

Something to track the effectiveness of the GWM -5/+10 versus ascending AC might be a simple but interesting place to start. I'm sure it's been done before, but that's no reason not to give your own go. It's one of those statistical things in the game that a lot of people don't really grok.

Or maybe something to rate the value of each AC bump versus various enemies? IDK, I'm just throwing monkey poop at a wall here.

 

[Guest 6801328]

Something to track the effectiveness of the GWM -5/+10 versus ascending AC might be a simple but interesting place to start. I'm sure it's been done before, but that's no reason not to give your own go. It's one of those statistical things in the game that a lot of people don't really grok.

That might be interesting. To really examine it's efficacy you'd want to include:

1. Base damage (that is, how much you would do without GWM)
2. Attack bonus
3. Disadvantage, Normal, Advantage, and Trivantage
4. Target AC

 

[Fenris-77]

Would taking multi-attack into account mean anything one way or the other? I know the really savage examples of GWM I have seen have been on high attacks per round. That might just be a matter of addition though. My stats repertoire definitely fits on the back of napkin, and I always get a little lost past rocking a bell curve or piles of d6s for 40k.

Since SS uses the same mechanic you could do both. I'm thinking there of a an actual statistical understanding of the efficacy of the archery fighting style, and thus the worth of one-level fighter dip for a rogue.

 

[Guest 6801328]

Given that the problem space is gigantic, perhaps give us more criteria? What makes a data set interesting? A natural next step might be damage per second extended to other classes against different targets. Of course, this has probably been exhaustively done elsewhere, which might or might not dissuade some.

Good luck!

I don't know...what makes a data set interesting?

Having more than one dimension of data is good.

Also, I'd much rather have raw data than aggregated data. A list of coronavirus cases by state (for example) is ok. But every single coronavirus case with latitude, longitude, and date would be way cooler.

 

[Guest 6801328]

Would taking multi-attack into account mean anything one way or the other? I know the really savage examples of GWM I have seen have been on high attacks per round. That might just be a matter of addition though. My stats repertoire definitely fits on the back of napkin, and I always get a little lost past rocking a bell curve or piles of d6s for 40k.

Since SS uses the same mechanic you could do both. I'm thinking there of a an actual statistical understanding of the efficacy of the archery fighting style, and thus the worth of one-level fighter dip for a rogue.

Yeah I think if you have 3 attacks its the same math as for one attack, but you use it three times.

For rogues with sharpshooter and multiple attacks (e.g. 5 levels in Ranger or Fighter) there might be some interesting considerations for sneak attack. I.e., does it make sense to make a normal shot first, and if you hit (and trigger sneak attack) then go ahead and throw the hail mary? Or maybe you should try sharpshooter on your first attack, then switch to normal if you miss?

 

[Fenris-77]

See, now we're cooking with gas. That sounds like information I'd want to have.

 

[TwoSix]

Yeah I think if you have 3 attacks its the same math as for one attack, but you use it three times.

For rogues with sharpshooter and multiple attacks (e.g. 5 levels in Ranger or Fighter) there might be some interesting considerations for sneak attack. I.e., does it make sense to make a normal shot first, and if you hit (and trigger sneak attack) then go ahead and throw the hail mary? Or maybe you should try sharpshooter on your first attack, then switch to normal if you miss?

That would be an interesting one, for sure. It would need some boundary conditions, though. It's only interesting when accuracy is high enough that a normal SS shot with no sneak attack is better than a normal shot, otherwise not using Sharpshooter for both shots is always the better option. The really interesting point is in the situations where the first shot has advantage (due to Rogue Aim or some other bonus, like Help or Guiding Bolt) and the second doesn't, that's going to be really sensitive to any bonuses on the normal damage as well as the amount of Sneak Attack dice. The increased chance to crit will definitely play a role, as well.

 

[Eyes of Nine]

I'm interested in Sorcerer Metamagic and what spells are best at what levels to do most damage to 1, 2, 5, and 10 opponents within a 20ft square.