D&D General Data from a million DnDBeyond character sheets?

[Clint_L]

With respect, the data set doesn't speak to how long the player has been playing, or show us that after playing a simple fighter, the players go for more complex options. The "new player" thing is a supposition.

I was thinking the same. It seems intuitive that new players would gravitate towards simpler classes, but that doesn't mean it is actually the case - science is full of counterintuitive results.

In looking at my own beginner campaigns, they maybe skew a bit towards simpler classes, but not overwhelmingly so. Speaking from personal experience, I actively chose to play a champion fighter not that long ago precisely because I didn't want to have to think too hard during combat.

 

[Umbran]

I was thinking the same. It seems intuitive that new players would gravitate towards simpler classes, but that doesn't mean it is actually the case - science is full of counterintuitive results.

Science is about replacing intuition with reality.

 

[Ruin Explorer]

Not sure of the legitimacy of the scrape or of the survey. But if it is I thought it was interesting enough to share. And somebody here would pick it apart if it’s bogus.

This seems to say that fighters and rogues are more popular than the online discourse about classes would lead folks to believe.

Didn’t we hear something similar about fighters being the most popular class before, from DnDBeyond?

Player Class Stability from 1M D&D Beyond Sheets

Regular readers will know I do love a good D&D Dataset and Dice-Scroller gives us a good one! Through some Python wizardry they scrapped ~1....

seedofworlds.blogspot.com

I'm much more skeptical of their "forums" data than the DNDBeyond data.

Their methodology for the DNDBeyond stuff is fairly consistent but not as strong as they seem to think.

Their methodology for "forums" seems to be absolute trash to the point of having no value. They make all these wild claims about how forum opinions are changing, but it looks like total and utter bollocks to me, because they've got trash-tier surveys (likely highly biased and selective and error-filled) from a dozens of different places, rather than a whole bunch of consistent surveys from one place.

I can tell you this: 12.6% of the data has either all 8's or lower for their ability scores, at least in terms of point buy. That is, the point buy cost for their abilities is 0 or less. Oddly enough 52% of PCs have point buy score of 27, with 32% of the PCs having the standard array (the 32% is part of the 58%). This suggests racial bonuses were not applied. However, I am not seeing blank races in the data. There are also several point buy values below 27. They go as low as 18 before you start getting 0's. There 70,018 thousand of those, or about 5.8% of the data.

Edit: Duh, the low (but above 0) point buys could be bad rolls on 4d6 drop low.

This is useful - that suggests 12.6% of characters were probably never even taken beyond selecting a class. The fact that they're showing zero blank races despite having six different data-pulls is kind of interesting and makes me a little more skeptical of any findings - are they just eliminating all characters who didn't have a race? Or substituting one? Or does DNDBeyond default to a certain race (it doesn't seem to, for me)?

 

[ichabod]

This is useful - that suggests 12.6% of characters were probably never even taken beyond selecting a class. The fact that they're showing zero blank races despite having six different data-pulls is kind of interesting and makes me a little more skeptical of any findings - are they just eliminating all characters who didn't have a race? Or substituting one? Or does DNDBeyond default to a certain race (it doesn't seem to, for me)?

I am not seeing any evidence that the races for the bad ability PCs are faked. They might but I'm not seeing the evidence. The bad ability PCs have a mix of races, that mix has a different distribution than the other PCs, and there are races in the bad ability PCs that are not in the other PCs (note that there are all sorts of low count weird races in this data, like Werechicken, Dragonborn Goliath, Kenku Mimic, Teethling, Tentamouth, and COPY_OF_Elf).

 

[TerraDave]

Anecdotally I see lots of interest in fighters, some in rogues, less in say land druids--which was finally confirmed to be a powerful option, and even less in wizards. They are hard to recruit for, though people like it when they try it.

On the other hand, multi-class options that try to do everything, yet still have some perceived win button, and of course can both use magic and fight in melee with a sword, as are also very popular, even if they will be sources of disappointment in practice.

 

[Clint_L]

For interest's sake, here is the breakdown of primary character classes from my (mostly) beginner campaigns since I started using DnDBeyond, about three years ago. I would say the breakdown of player experience is roughly 30% absolute newbie, 40% had tried the game or at least had the basic concepts, 30% had played more than a handful of times/owned a PHB:

Artificer: 5
Barbarian: 7
Bard: 6
Cleric: 8
Druid: 3
Fighter: 8
Monk: 7
Paladin: 5
Ranger: 5
Rogue: 9
Sorcerer: 4
Wizard: 5
Warlock: 5

Take it for what it's worth - not a big enough sample to really establish firm trends, though anecdotally it looks fairly in line with other distributions that we've seen, with some extra clerics (but this may be because I kind of pitch clerics as a really cool class that can save the whole party). In my experience, the classes that players have the most trouble "getting" are druid and sorcerer. The few players who have chosen druid mostly just wanted to shapeshift, all the time, and the difference between wizard and sorcerer is tough to explain to a newbie.

 

[Ruin Explorer]

I am not seeing any evidence that the races for the bad ability PCs are faked.

Nobody said they were faked?

What are you talking about?

 

[Shardstone]

Yep, that’s consistent with what every other big analysis has shown.

At this point, forum users have to accept that the vast public very much enjoys the very stripped down fighter and even the no subclass features from 3-9 rogue.

 

[Shardstone]

I think the Warlord should just be made as the complex fighter. Break it from the old 4E definition so that commanding others is just one of many potential things it can do via subclasses. At 11th level+ have it make a decision between being a mythic warrior or the best strategist ever.

 

[EzekielRaiden]

From my first look at the data set after downloading, I've been thinking it needs some cleaning. There are several PCs in the first few rows with either all 8's for abilities or all 0's, which I was thinking should be excluded from any analysis. I would also exclude blank and test names. But what about this Damien guy? And how many people name their test PCs Bob?

Yeah, this sounds pretty much like the concern I raised at the time. That is, this data set actually does deviate in a noteworthy number of ways (for example, dwarf has gone from "not even in the top 5" to third place, while dragonborn, which had been just behind humans, elves, and half-elves, has fallen quite far) from data we've gotten from DDB in the past.

I strongly suspect a significant portion of these scraped characters are either
a) not active/"real" characters, but just tests/mockups/practice
b) DM-made NPCs or other characters that are "real" in the sense of being used, but not truly being played
c) not even remotely serious, e.g. the ones with all 0s for stats, and thus junk data

AIUI, no data filtering was done on the data set. It was taken exactly as is. That's an issue I raised when this was initially brought up, and the creator was both brave and honest enough to make an account and respond. I have no ill will toward them, but I genuinely believe that the data in this set doesn't really tell us much of anything. It certainly doesn't actually tell us things like "nobody actually likes playing Wizards" or "people actually really do love the Fighter exactly the way it is and you shouldn't change ANYTHING about it because they would definitely hate any other alternative." Which...yeah, I've already seen people literally come to that conclusion in different words.

I think the Warlord should just be made as the complex fighter. Break it from the old 4E definition so that commanding others is just one of many potential things it can do via subclasses. At 11th level+ have it make a decision between being a mythic warrior or the best strategist ever.

Given they already tried that with the "complex" Fighter they gave us and it sucked and wasn't even really a Warlord, I'm going to have to ask your forgiveness for having a piss-poor opinion of that particular "compromise."

At this point, forum users have to accept that the vast public very much enjoys the very stripped down fighter and even the no subclass features from 3-9 rogue.

Nope. Unless and until we actually get a real, well-constructed survey (something WotC is not willing to pay the money to do), we don't at all know that. Indeed, we know nothing of the kind, and the only people who might potentially know--WotC themselves--are absolutely not ever going to give us the data we would need to verify any claims they themselves make.