Player Data Doesn't Lie | D&D vs Baldur's Gate 3 

💥 My book, Delve, is LIVE on Kickstarter! www.kickstarter.com/projects/bobworldbuilder/delve-5e-shadowdark?ref=apoi0j

As a data scientist, altering the vertical axis scale like that is a war crime.

Important considerations for BG3’s data: player choices exist within the context of party balance and which companions you decide to use. Paladin and Sorcerer have high representation because they don’t exist in the core 6 (or really at all) and cleric has low representation because your auto-assigned companion Shadowheart serves the role. My first character was a cleric until I realized I didn’t need him to be, and switched to paladin. Then my next was a sorcerer.

One thing worth noting about the BG3 data is that because you get characters of specific classes right at the start of the game. there is a major disincentive to play those classes. In particular, the first two classes you run into are a fighter and a cleric, so that likely affects their popularity. The D&D Beyond data is also skewed because human and fighter are the two defaults, meaning that a lot of basically blank characters are "human fighters" because they didn't actually get made beyond being "made".

For the player data on the dnd beyond app: I use it to play with dice when I'm bored, and create characters I'm not planning on using anytime soon bc I had an idea and wanted to see how it would work with the mechanics, i usually delete those pretty quickly too. So I've created probably dozens of characters and only actually played with two!

For the data on Cleric being the least popular class: It makes sense since the most popular companion is shadowhearth, which means you already have a cleric in your party in most cases.

I think another big reason Paladin, Sorcerer, and Warlock (and sort of bard) are so high in BG3, is because they are all Charisma based characters. When you play BG3 in single player, your character acts as the face of the party, so it makes sense to build a highly charismatic face.

Popularity of Paladin and Sorcerer for new characters in BG3 may have something to do with those classes not being represented in the canon companions you get, at least near the beginning of the game. The players may have wanted to avoid repeating classes in their party, or see how that class would work in the game (I know I played a Pally partly because of those reasons).

The data about the number of characters created on the app (6 million) vs the number of players rolling dice (1.9 million) makes sense to me because while I love the DnD Beyond for character creation, tracking everything, and in every way simplifying the game, I also love to role my own dice. I play every week with 6 other people and all of us use real dice on the table. The only time I've rolled in the app was an errant key stroke.

11:04 to this point, my group all shares one dndbeyond account so that only one of us has to buy everything, and then everything gets shared in every campaign any of us are in. So we’ve made 30 characters this year on just one account, as most of us are involved in multiple campaigns/oneshots. That sort of thing may account for those numbers, although this is obviously just my experience.

Interesting. I think the source material makes a huge difference. Of my 5 ppl. 2 of us play BG3. One of those 2 has 3 characters. I have over 10 myself. Our DnD group however all play on paper sheets. You can (and many do) play DnD without Beyond. But no one can play BG3 without BG3. . .one has an off option the other does not. =)

the class disparity is because - in actual D&D - vertical space and positioning is not usually considered by a lot of DMs. So something like Storm Sorcerer's 'Flight after a spell cast' bonus action doesn't see a lot of traction or value by most players. There's also all the extra dice and math that have to go into Sorcery points, and compounding damage, etc. In a video game, all that math gets done on the backend and you can see the value of your choices more easily visualized - even if limited by the imagination of the developers.

In BG3 the default selection, if you just change nothing about class in the character creation menu, is a Dragonborn Sorceror if you pick the Dark Urge origin, so I guess a lot of people picking that origin (which is both meant to be geared toward an evil playthrough but also pitched as sort of a "Custom Character +" campaign since it adds more uniqueness to your create-a-character) may have felt that those were good choices (they do synergize well, to be fair, with the Dragonic Bloodline Sorceror). And Paladin is one of the few classes not represented by a Larian created companion (well, there's one, but they're hidden and basically a bonus character meant more for an evil playthrough, as you're led to kill them pretty early on if you're going traditional Chaotic Good style), so it makes a lot of sense that they'd be a popular player choice as it lets you fill in a role the others won't. The other one that doesn't have a companion character in the game is Monk, but I think most people shy away from Monks generally compared to Fighters/Paladins/Barbarians because they're more for fitting an Eastern aesthetic and more of the Forgotten Realms is definitely Western aesthetics.

One interesting thing about characters and classes in Baldur's Gate is "the other characters." There is a whole party of other characters available, and people pick and choose which of those characters they want on their team, and then often build their own character to complement that team. They might be less likely to choose Cleric, because they intend to have Shadowheart on their team.

Man, people are really uncreative naming so many NPCs Bob. I never reuse Bob. The second one is always Bob Bobkins.

In several years playing DnD with several groups, I’ve met one person who uses dndbeyond

The reveal of the corrected graph made me unreasonably happy. Thank you .

The 3:1 player to character ratio makes sense because I think I made in total like 50 DnDBeyond characters in 2023, whether it be running through different concepts to very short one-shot characters to making NPC concepts. Most people don't do this and might just make one or two but 3:1 is certainly normal.

No coincidence that there are so many human fighters named Bob. They were inspired by the legendary human warrior of the same name that fought in the OGL wars.

No one can stand against the supreme might of the Bobs! 💪🏽

when i tell you DDB 's data presentation hurt my soul.... thank you for this breakdown. as a data nerd, i love it.

I believe the difference in character creation numbers may be driven by the D&D Beyond limit of 5 characters...unless you have a subscription.

5:41 oh god when you pointed out the weird data squashing that made me feel sick. Arbitrary interval jumps are already crazy, but having bigger jumps and then going back and doing a smaller one is something I’ve never heard of. Truly a “graphic design is my passion” representation.

This was really interesting, Bob! Thank you!

Another factor in cleric unpopularity is the story character most likely to automatically join your party is a cleric, which you often don't want 2 of. (You can respec Shadowheart, but that requires extra time/effort, and the desire to respec, which probably most players don't have.) Personally I went for Cleric in my first coop campaign, and felt it was pretty amazing. Underwhelming heals, but stuff like spirit guardians was absolutely overpowered in some fights. Would've been interesting to present the data side-by-side more. So like * do a graph for class distribution, sorted by most popular D&DB class * for each class, have three bars: D&DB, BG3-launch, and BG3-postlaunch (color-coded, with the two BG3 bars being somewhat similar colors) * basically this would've made it really easy to do what you did briefly in the vid of flipping between launch/postlaunch BG3 to compare shifts in popularity

The baulders Gate popular Classes make sense, there are no Paladin nor Sorcerer NPCs that join your party. (Well none that were originally announced) so makes sense that people wanted to play a class that wasn't already on the Roster. Also CHA is huge for dialog options. People like to "win" the conversation options they choose.

9:57 More importantly, in a video game you tend to want your character to do the talking. Paladin is "the talky Fighter" and Sorcerer is "the talky Wizard." You also have Shadowheart as your dedicated healer already in the story. I feel like most people wanted Shadowheart in their party for many reasons 😂

you also have people who use the app to keep track of the data sheets.. Gold.. extra items... etc. stats that change like HP... this allows for an easy to manage system while also having the ability to keep track of certain things within home games. Our group of 6 - 8 use it for this reason but wont roll on the app because we have our own dice, But we did create our characters on the app with the character sheet but don't actively use the app / website..

Great video! As someone who creates characters in D&D Beyond just for fun, to sketch out an idea that will never get played in a campaign, or to help my players with their stats or choices (as a forever DM), I have a couple points: 1. It's me. I'm spamming all these characters. 9:17 2. If there are lots of people like me who just love character creation, this is even more of a vast comparison to make to BG3, where my assumption is that it's not likely that people do this (it's not as much of a throwaway act to make a character there).

Yeah, video games are more popular than tabletop games, online or not. But now I want an entire party of human fighters named Bob to try to run through a dungeon.

Ok no as someone WITHOUT a strong background in data science or anything that heavily relies on statistics, the graph drove me nuts too. I wouldn't have noticed if Bob didn't take a second to point it out but now it's seared into my brain like a hot poker through the frontal lobe

Love a bit of analysis. Great stuff Bob.

Man, I started going crazy when I saw that vertical axis. I'm glad you also picked it up.

One thing i feel that scews the data on bg3 is the companions. Why play as a cleric when you get shadow heart (one of the mor interestinf companions) so and so forth. So looking at the most popular classes and seeing paladin and sorcerer (durge influenced likley) makes a lot of sense.

The classes in BG3 are slanted towards working with the known starting npc's: Rogue, Fighter, Cleric, and wizard.

5:28 As someone whos taken multiple classes on research and graph making I care! I might show this to one of my professors actually if I get the chance lol

For what it's worth, this is the first time I've heard of DnD beyond. Somehow I've managed to play DnD for years, and consume plenty of DnD youtube content without actually hearing of it before. In our player group we use an excel sheet to track our character stats, and have a PDF of the players handbook: I don't think there's any way for us as players to be captures in metrics. I wonder how much of an anomaly we are in the player-base, or if there's some large silent+invisible minority/majority.

One thing to consider is that many people played Baldur's Gate who are not D&D players but are just avid video gamers or play TTRPGs other than D&D.

Cleric is in the end because we have Shadowheart ❤

Always nice to have someone talk about the scale of a graph, and confounding variables such as whether something is free and how long it has been available. Kudos, Bob! It might be interesting to separate out classes etc by whether they are free or not, and maybe also to look at how long they have been available?

As a member of the 'almost another Bob' club (tm) and being directly blood related to 6 Bobs. It doesn't surprise me its the most common name, I was almost expecting it. As a long time player and advocate for the Human Fighter being the #1 class and species. I am pleased.

One thing might be interesting - as DM I have a lot of characters made to test things, sometimes even play characters (NPC) from a character sheet and not statblock. Effectively I have currently 89 characters made, and since I tend to erase some of them, I'm sure I've created over 200 within the last 3 years. But it doesn't mean that many people play that. So it would be interesting to see how many accounts have average amount of characters.

Something I love to do on DnD Beyond is just hypothesize characters for fun. I also continuously update my back-up characters along with my active ones.

Thanks for the content!

Idk if they have it yet but if they dont they definately need to add a tool that can create a character sheet in dnd beyond using your character from bg3. They could also include a screenshot printable token or character cutout for battlemap play, bonus points if they made a way to convert the 3d model of your character from bg3 into an stl so you can 3d print your character for your table top games. I think that would help push people to explore dnd beyond and dnd more and convert some bg3 sales into dnd/dnd beyond sales

The first couple years of my current campaign, it was 5e and I played completely on paper. When we converted to Pathfinder last year, I sadly had to switch to digital... there's just so much more to memorize and I haven't achieved that level of proficiency with the system yet. But even so, Pathfinder is fantastic. Three action system for the win!

I make characters a lot in Dnd Beyond that I never play, just to work out what I want to do. And the characters I do play on the app, I roll physical dice for, so that's something to consider. I'm sure there are others out there that use the app at the table, but roll their own dice.

I made 3 D&D Beyond characters. That was new for me as I have been (as most of my friends) a mostly analog player. I also dumped 120 hours into BG3 and made a few characters there. Stats seemed right on point along with your assumptions of analog play.

I have Dmed in tons of Games over the past year and a good thing to keep in mind is the content sharing.

"But we don't know how many Bobs are out there..." Does that ever sound like an epic quest 😏

Something else to say about that Data - not just that Paladin and Sorcerer don't exist in default companions - but your main character speaking to most people is a big deal. So charisma is a huge deal for a main character. Where in table top games, you might leave that roles open to someone else.

thanks for pointing out that scale slaugther, because that's the first thing I've noticed and it was killing me...

First thought is 180million players Vs 88 million enemies - either they didn't write this for people who play DnD or they didn't think about that ratio. The majority of fights (in my games at least( tend to be at least 1:1) but here we've got at least 2:1 players to monsters. Paladin and sorcerer are ones that we didn't get as companions in BG3 so maybe that's a reason for their popularity.

For the most part my group shares one account to roll characters so we can downlad them or import them into foundry. Which may explaun some of the data. Normally one person buys the contenet and shares it but thats been my experiance.

one thing i think to note about the BG3 data is you start with 3 pre mades that you dont initally have class control over. as thus could explain why some did not create a cleric as the "healer" was already done.

Love your videos, Bob!

I never considered playing a Cleric in my playthroughs bc I enjoyed having Shadow heart in my team. I think things like that do effect the data.

I do think that part of the Paladin being played more than the Fighter in BG3 could be the visuals of the armor in character selection. And yes, Cleric/Support tends to have a bad rep but also there's Shadowheart who is a very popular companion/romantic interest. No need to play a Cleric if you'll get one.

A note on the BG3 data: they didn’t release the full character creator till launch so that first week all the dnd nerds were rolling character after character exploring the options while waiting for the weekend to play with their friends. At least I knew several. And the paladin being top class I think was split between larian doing a great job marketing the oathbreaker and fitting it inside the game in an RP way, and there being many build guides recommending the paladin/lightning domain cleric for the thunderous smite bug.

Foundry does have a character creation tool, but we usually create the character on DnD Beyond and then migrate it to Foundry VTT, or directly link the character from DnD Beyond to Foundry, so likely Foundry entries are included

You hit on one of the most important aspect of D&DBeyond is the restricted licensing, and the ability to create and make your own. Even a DM owns a core paper product, to put things into DNDB you must have a higher license to share. This is very different from other VTTs. To answer your questions Foundry doesnt really have a character creator with licensable content. But the DM is free to create content like that to fit into their game and share it. It is one of the strengths of Foundry in for third party conflict its great.

I definitely agree with you on the silent analogue majority. In my play group of 6 + DM only 1 uses DnD Beyond. It is also very important to remember that BG3 is a video game. Popular games regularly hit the numbers that you read out there. I bet a lot of people who played BG3 probably don't even know it's based on DnD. With video games, the amount of sales also slowly dries up, and the amount of players decreases as people finish the game. With DnD, once a campaign is done, you can always come back for an entirely new experience.

My personal DndBeyond use: I make characters on that site to get the rules right, and then I copy them over to paper, or to Roll20 to play with them. Never roll dice in the app, because i like my click-clack math rocks! When I set up an intro game for some of my friends, I made all their characters with them in one character slot on DndBeyond, because it's the slot enrolled in a campaign with one of my main players, who bought all the books :P

A note on the D&D Beyond app. I actively play with a large group of people divided among 4 tables, the DMs have decided that everyone must create a character on the app and join one of four campaigns, mostly to keep track of who plays and what their characters are, however, most of the group doesn't actually roll in the app, and some of the members don't even open up the app after creating the character except when leveling up. I just use it to keep track of items and spell slots etc.

Love data discussions. This is one of my favorite videos in a long time.

One thing to consider with BG3 class stats is what classes your party members have? First party member you get, is a cleric so that could explain why cleric is the least popular.

I feel your Pain Mr. World Builder, being named Bob is a trip

Makes me wish Larian Studios released a BG3 Players Handbook, including all character options and rules options in the base game (subclasses, feats, magic items, etc) 📘

They are doing some data manipulation on the 2d map stats. This data most likely includes each map load, so if a map has 2 NPCs and 4 players is loaded up 3 times while a DM is testing stuff and 1 time for play that means 24 players and NPCs were played.

8:00 Yeah Foundry is basically like Roll20 but with more stuff and customisation, seperate tabs for character sheets, stuff like that.

Neither I nor any of my 12 players, nor any of my befriended DMs use apps like DND Beyond or Roll 20. We all use pen and paper. So yup, the idea that the majority doesn't use any of these apps is quite accurate to my experiences

Nice analysis, Bob! I think the video game market makes it much easier to lure the casual player into the D&D orbit than more traditional -- albeit virtual TT -- modes of play like D&D Beyond (still 2D maps and verbal role-playing). A ratio of 10:1 (20 million consumers on Steam with BG3 vs. 2 million on D&DB) shows that your average consumer/player prefers reflex-based action games where the "imagining" has already been done for you (visuals are provided and you simply react), rather than the player having to pro-actively describe & verbalize & dramatize, and collaborate on an emerging Story. *Interesting corollary: the ratio of Garycon (mixed RPG bag of events) to Founders & Legends (OSR focus) is also around 10:1. I am estimating around 2000 games for Garycon and 200 for F&L, at least at my last viewing of the Schedule. Both events are held 1 week apart this year on the 50th Anniversary of D&D.

Yea what the graph shows is that most people are not paying for the other races. I know that once I found out that I was going to have to pay to even be able to use homebrew or custom characters I completely stopped using DnD beyond. I started using 5e companion instead.

I just want to say Foundry VTT indeed has a character creator* with an added module. In fact, the built-in module support for creators is one of the best parts of Foundry. It's a one time payment for almost unlimited functionality* some mods are locked behind a Patreon subscription. I think it's worth every dollar, AND they just officially partnered with Wizards of the Coast, similar to Roll 20.

For my table we use the dndbeyond app on our phones for character sheet management but roll physical dice cause we play in person. Could explain some of the disparity between players who rolled dice vs. made characters on the app

About the 6m characters created, I feel like it could be people trying out concepts and also creating a character in dnd beyond but playing irl. Which means that they would create a PC but won't use digital dice. At least my group have both analog and digital character sheets on roll20 but we usually meet irl and use irl dices

6:11 that graph is amazing!

To be fair, I made literally 100 characters on DnD beyond last year lol. So I believe that last stat.

I use dnd beyond character creators, and I made like 25 characters in 2023 for various oneshots, and for premade characters for my players to play. They all made their backups too, and some of their backups got brought in, so I'm not surprised at the average amount of characters created via the dnd beyond app.

with class distribution on bg3 there is probably also an effect of shadowheart being very popular and being a cleric, so there is no need or desire to make their character a cleric as well.

I think a thing to remember comparing the 31 mil BG3 characters made to the 3 dndb characters is that BG3 came out 6 months ago while 5e has been around for a decade. There were people who were involved in multiple long-standing campaigns who made zero new characters last year. Meanwhile I personally made 4 different BG3 characters, but only played 2 of them.

We all know you have a script creating human fighters named bob.

Thank you! 😃

I was absolutely geared up to comment about wishing you had pointed out the terrible scale on the species graph and then when you started to do exactly that I was filled with joy

Baldur's gate 3 did things very well. It`s also interesting to see how BG3 took decisions in direct opposition to what onednd did. Like moving Paladin's oath to level 1 while onednd moved all subclasses to level 3.

Why are we all up so early?

My guess for the disparity in sorcerer players is that it has to do with streaming the video game versus playing with tokens. I know from highlight reels that several big streamers started fooling around with the sorcerer's wild magic surge. Chaos ensued and copycats followed. Also on the topic of player races, I know that some old-school players hate that you are able to play the traditionally more monstrous races like half-orcs, tieflings, and dragonborn, but there is at least one player playing a character of those races in every game that I run or play. They clearly have a lot of appeal. Both datasets you showed seem to represent that as well. It would be interesting to see the data of class/race combinations and how maybe players are playing with or against type.

Re: Foundry There's an add-on that allows you to straight up import your character sheet from DDB, but it's also possible to create your own character in Foundry. You have access to the base classes, but only one of the subclasses (I think?), so you'd have to homebrew anything else. Depending on the mechanic you need, it can be quite intuitive but others take a while to figure out. So while it is technically possible to create characteres in Foundry, creating them on DDB and importing them to Foundry is typically more efficient. I can't be arsed to homebrew an artificer, so my player gets an analogue sheet and her Foundry token will just have the info we need for combat (AC, HP, resistance and the like)

brb omw to make a male human fighter named bob

The company can’t stop lying and deceiving. Even with a simple use graph they have to play all kinds of shenanigans. And the “and NPCs” bit is so characteristically disingenuous. They might as well have just counted “pixels moved.” Thanks for your analysis, Bob.

Statistician here: 100% the adjusted y-axis bothers me.! You are not the only one. Thanks for this deep dive into the data.

I think the reason for cleric being so low down the BG3 scale is because the literal first character you find after leaving the ship is going to be a cleric as shes the only person you will find before having to make any check or go through combat. There are ways to get to other characters without combat or checks but its pretty unlikely you would miss shadow heart.

That charity thing is just a way for the corporation to do tax write-offs. Don't ever do charity through a profit-driven corporation, give directly to charitable organizations.

Everyone that plays BG3 has to play through BG3. Not everyone that plays dnd plays on dnd beyond....

AS someone who plays a good deal of D&D in table top, I think I see some potential explanations of the data just in my own experience. 1: I don't really use character building websites for my characters. I know how the system works now to the point where I just use my books and a form-fillable PDF. 2: I bounce around different sites for online games. Some use D&D beyond, but most don't, and use something like Foundry, Roll20 or one of many other options. This will definitely fragment the data.

there are alot of ways to make a sheet, I made 3 characters this year and made all of them outside DND Beyond. BG3, obviously forces you to make a character in BG3. So yeah, it does outshine DND Beyond, but that also makes alot of sense given the situation.

Dnd beyond doesn’t indicate the popularity of DND as a whole though.

Some of my party use D&D Beyond to create and manage their characters but I think only a couple actually uses the app to do their rolls etc, although we're all in-person play. So we don't really use a lot of the functionality other than just a character sheet manager. We do tend towards paper, pencils, and dice although the app does help with character management so is more accessible than paper for some. But I'd be just as happy without it, TBH. I love the process of creating a character on paper although of course there are some who love it.

Last December I run a quick one shot for 2 friends who never played RPGs before.. I played other systems but I've only played a couple of DeD one shots myself... so after they chose the concept I used DeD beyond to build their characters for them. The free tier help me on that task ..

Will say I shared my DNDbeyond account with my players for making and running their characters.