Who's Multiclassing With Who? More D&D Beyond Stats!

mach1.9pants · Feb 17, 2019

Nathan Shehorn said:
This is the greatest article, I have read. And the first comment I have ever left. Bravo. My analytical OCD is so happy.

You were so happy, you left two comments!

FrogReaver · Feb 17, 2019

EzekielRaiden said:
I could do without some of the...inaccurate data presentation. Most of the graphs are fine, but the very first one is almost certainly badly constructed: the way the bars are stuck together, it looks like nearly 50% of all Warlocks are multiclassed and over two thirds of all Fighters are. That's...not true. It's not 33.36% and then a further 39.74% of them all, it's 33.36% and then a further (39.74-33.36) =6.38 percentage points. The graph as shown heavily distorts an otherwise very clear data visualization.

Considering the first graph doesn't even have a title I don't know how you would have a clue what it is showing without reading the paragraph explaining what that graph is showing. Once you've done that it's perfectly clear.

Morrus · Feb 17, 2019

FrogReaver said:
Considering the first graph doesn't even have a title I don't know how you would have a clue what it is showing without reading the paragraph explaining what that graph is showing. Once you've done that it's perfectly clear.

That’s probably just my rubbishy screenshotting abilities.

FrogReaver · Feb 17, 2019

guachi said:
Yeah, the first chart is a mess.

Is like whoever made them went nuts in Excel without really knowing the best way to present data.

I like how the first graph's data is presented. It also takes up much less vertical space on a screen that way than trying to do vertically side by side horizontal bars. I'm sure that was a consideration for doing it this way as well. More importantly, the brief decryption below the graph clearly explains what is being shown. It also acknowledges a few important things about the presentation that others might think is strange.

Mistwell · Feb 17, 2019

Link to source, skip to around minute 22

First graph is titled "Most Popular Class for Multiclass Characters (all levels vs level 20)

Hussar · Feb 17, 2019

Always nice to see my biases confirmed.

FrogReaver · Feb 17, 2019

Morrus said:
...The fact that you can take more than just two classes means that these won't add up to exactly 100%...

Morrus, I know you are probably just the messenger, but this bit of the first post is inaccurate. Even if they restricted their data to "Active Characters with exactly 2 classes" they would still have a class graph showing over 100%. In fact in that situation it would show exactly 200%. So the reason about having more than 2 classes is incorrect, it's because they are showing a population that accounts for each character once but showing a breakdown that has each character accounted for more than once.

An interesting observation: Their chart for level 2+ characters adds up very closely to 200% without Druid being shown. With him added in it's likely just a little over 200%. What this tells us is that very few characters in that population of level 2+ characters have more than 2 classes.

Prakriti · Feb 17, 2019

with *whom. Yeah, I went there. :heh:

Also, even after years of playing 5E exclusively, I will never get used to seeing "Fighter/Rogues." They will always be Fighter/Thieves to me.

TwoSix · Feb 17, 2019

Since these charts agree with my biases, the methodology must be accurate. Science!

AmerginLiath · Feb 17, 2019

Prakriti said:
with *whom. Yeah, I went there.

Also, even after years of playing 5E exclusively, I will never get used to seeing "Fighter/Rogues." They will always be Fighter/Thieves to me.

You just beat me to it (making the grammar save, that is — although I agree on the AD&D terms too)!