I truly wish I could blame the y-axis on these graphs on AI, but I am pretty sure they were made by a person.  

These graphs are cursed.

"Why doesn't WotC engage with us more?!"

*Me gesturing vaguely to all the bullying posts happening to a community outreach staffer, a real life person, after the staffer made a fun article for the community*

Some things will remain a mystery forever...

Quote from Erik_Soong >>

"Why doesn't WotC engage with us more?!"

*Me gesturing vaguely to all the bullying posts happening to a community outreach staffer, a real life person, after the staffer made a fun article for the community*

Some things will remain a mystery forever...

I completely agree. I mean, come on, its a frickin bar graph. give the staff a break. Personally, I enjoyed the article, I thought it was fun to look at the statistics and see the good things that happened this year. WOTC made mistakes, they messed up. but we need to give them a chance to redeem themselves. they know we're watching them. they know what we expect. that doesn't mean we have to remind them of past failures ever chance we get. so I will just say one thing:

Good Job, DNDbeyond team, I love this website/app and don't know what I would do without, so thanks. thanks for the awesome space where everyone can bond together and play our game.

My class is always wet blanket, and as such i preferred a pie chart to a graph to be honest. I feel it really delivers easily digestible information into bite size pieces, but I’m no expert.  I’m just a soggy blanket.

To anyone not understanding why the graphs are a “big deal” to some of us, here are a couple of points to consider:

1) There is a significant segment of the population that has difficulty reading graphs properly. Assuming no ill intent on the part of the graph makers, the poor design contributes to that and can make it even more difficult, exacerbating the problem. 

2) A variable scale doesn’t happen by accident. You’d have to go out of your way to create a scale that starts by 20s, then by 10s, then by 100s. Someone made the conscience decision to “flatten the curve”. Pure speculation on my part, but I see two possibilities: A) They want to obscure the fact that some classes are failing, “See?! Artificers and Monks are *almost* as popular as the others!” and/or B) Downplaying the most common classes to encourage more variety by “playing down” rather than do something more to improve the classes. 

3) It’s dishonest. It further erodes what little trust players and DMs have left in Hasbro’s stewardship of D&D.

I loved the product before WotC/Hasbro bought it and I really want it to succeed. So while I greatly appreciate the staff and all the work they put into improving D&D Beyond, I don’t trust whoever made the decision to manipulate these graphs; especially if these were just supposed to be for fun.

Absolutely. I use graphs for a living. Thank you for this. Get ready to be slammed.

Would love to know that numbers for the most popular sub classes and feats.

My only gripe is that the species/class graphs were not evenly laid out. That's it.

Looks like Druid has clawed its way up to the number 10 spot, with an indeterminate number more than monk. Shouldn’t the scale be in 50 thousands just to have an accurate and consistent graph? That’s what my math teachers said at least.

Quote from lygerseye >>

To anyone not understanding why the graphs are a “big deal” to some of us, here are a couple of points to consider:

1) There is a significant segment of the population that has difficulty reading graphs properly. Assuming no ill intent on the part of the graph makers, the poor design contributes to that and can make it even more difficult, exacerbating the problem. 

2) A variable scale doesn’t happen by accident. You’d have to go out of your way to create a scale that starts by 20s, then by 10s, then by 100s. Someone made the conscience decision to “flatten the curve”. Pure speculation on my part, but I see two possibilities: A) They want to obscure the fact that some classes are failing, “See?! Artificers and Monks are *almost* as popular as the others!” and/or B) Downplaying the most common classes to encourage more variety by “playing down” rather than do something more to improve the classes. 

3) It’s dishonest. It further erodes what little trust players and DMs have left in Hasbro’s stewardship of D&D.

I loved the product before WotC/Hasbro bought it and I really want it to succeed. So while I greatly appreciate the staff and all the work they put into improving D&D Beyond, I don’t trust whoever made the decision to manipulate these graphs; especially if these were just supposed to be for fun.

 I did not know that 'big deal' necessarily meant that collective bullying was permitted.

1. Utter nonsense. This population you speak of does not need to know that users built 325,454 fighters or 47,666 custom lineages. What would you have done with this data that is now rendered impossible because of this egregious offense?
2. Again, the data, the goal of the presentation, and who one is presenting to are all important. What is far more likely, drawing on my own professional experience presenting regular monthly and quarterly reports to executive stakeholders to one of the largest healthcare organizations in the entire bloody world, is that because there is a large disparity in the data, that rigid to-scale graphs are going to make the data at the very least ugly as sin but also runs the risk of being nigh unreadable. Imagine all the harassment to workers unrelated to anything to do with executive decisions we would see here if the graphs were to scale and insanely challenging to read as a result. Further, do you seriously think that underrepresented classes by users are going to experience a meaningful change if, for some reason, staff were to present the data stupidly, as you suggest? I think not.
3. It is dishonest? Erodes trust? How very dramatic. You would have to turn yourself into a pretzel to link this data to the perceived nefarious actions of literally anything else done by the company, but I would sure love to see you try.

If you have to punch down on workers to communicate your grievances to a company, your grievances are not worth airing to begin with. A person made these data for our collective enjoyment, not because you are some actual important executive stakeholder at Hasbro, who would be making sweeping changes to the company/game based on the information presented here. This smacks of entitlement. Because these graphs are made for fun, they do not need to be held under a microscope unless someone is trying to find a reason to be angry. You will do nothing significant with the knowledge that artificers failed to clear 80,000 and you would do even less with the knowledge that artificers reached only 77,777 in number. For the purposes of who this data is meant to engage, more than 60,000 but less than 80,000 is all that matters. Where you direct your anger matters. A lot.

You guys have never seen a Logarithmic scale before??

Yes it's not the best scale for a bar graph, but when you have values ranging from 700k to 40k in the same table a Log scale is needed to avoid a bar graph the size of a A4 bit of paper.

This kind of data reporting is perfectly normal, so attacking WOTC staff for presenting "trend" information in this manner does nothing more then highlight how toxic some parts of the community has become.

Mmm.. A logarithmic scale is a nonlinear scale often used when analyzing a large range of quantities. Instead of increasing in equal increments, each interval is increased by a factor of the base of the logarithm. Typically, a base ten and base e scale are used.

I am not going to assume any malice, I am not even sure where it would come from here; however, I will chime in and say these graphs are pretty bad and should have gone through a second set of revisions. The class one particularly so, where the Y increments by  20k,20k,20k,20k,10k,10k,100k,100k

Quote from JustinDarkness >>

Mmm.. A logarithmic scale is a nonlinear scale often used when analyzing a large range of quantities. Instead of increasing in equal increments, each interval is increased by a factor of the base of the logarithm. Typically, a base ten and base e scale are used.

 
I don't know if the differences in data are large enough that it makes sense to consider a logarithmic scale, all data would be between three rungs if they did base 10, and any other base is likely to confuse a general population audience, but it would have made more sense to do that than what they decided to do. Honestly, putting in a break in the Y axis to capture Human and Elf on the species option might be the way to go if they didn't want to just increment by 100k.

Is anyone else calling BS on some of this data? In particular the line about 180 million PCs or NPCs played with in the Maps VTT. Even granting them four whole months (from September 1 through to December 31), they are suggesting that approximately 1.5 MILLION people used Maps PER day.

Are you kidding me? It makes it sound like your trying to boast having more active players than ACTUAL VTTs like Owlbear Rodoe or even Roll 20 on any given day in that period.

Quote from Erik_Soong >>

"Why doesn't WotC engage with us more?!"

*Me gesturing vaguely to all the bullying posts happening to a community outreach staffer, a real life person, after the staffer made a fun article for the community*

Some things will remain a mystery forever...

 i don't think people are bullying a staff member when they're pointing out the graph flattening. this is just general push-back and awareness from people who have developed a flinch with regard to interacting with manipulated facts. an alert chime. we flinch, yes, and rightfully so! put the gom jabbar away because this is a very human survival mechanism in an election year following a year of watching ai climb out of the uncanny valley. the flattened graphs are easy on the eye but they are deliberately curated data. in other places predators have used this camouflage for more than entertainment value and so a murmur goes up. if someone picks up a pitchfork and torch, then we can talk about pitchforks and torches.

We need to rebel against the human fighters and Bobs.

Its not a logarithmic scale.
On the Species graph the top few lines are of equal proportion going from 700k to 500k and then 500k to 200k and then 200 to 100k.  On a log scale the top portion can't be smaller than the one below it.  Its just really bad.

A log scale might be a reasonable way to show this data, but this isn't a log scale.  Personally, if I was advising this graph crafter I would not advise a log scale.  Humans generally don't think/reason well with bar graphs and log scales.

I appreciate the information they have given us, and its cool to think about.

It’s not properly written when it starts by going up by 20k then then half way through it goes up by 100k at the end when a proper graph should be same all throughout.

I made some quick graphs in google sheets using my estimates for the data here. The numbers aren't very exact, but the curve should be similar. I think. tell me if I got something wrong and i'll fix it.