D&D 5E Nobody Is Playing High Level Characters

According to stats from D&D Beyond, above 5th level characters start to drop off sharply, and above 10th level, the figures are very low. The exception is level 20, which looks like it's probably people creating experimental 20th-level builds.

Some of them say 0%; this isn't strictly accurate, but levels 16-19 are used by an insignificant number of players. Interestingly, there are more 3rd-5th level characters than there are 1st-2nd level.

D&D Beyond has said before that under 10% of games make it past 10th level, but these figures show the break point as being bit lower than that. DDB used over 30 million characters to compile these stats.

 

[BadEye]

This is a valid point, and leads me to a thought: if DDB only supports the base rules as written and doesn't do variants or houserules then speculative extension kind of suggests that it'll end up defaulting to a user base of more casual players and-or players playing under more casual DMs.

Saying DDB "only supports the base rules as written" is not an accurate statement. The biggest gap at the moment is the gigantic UA content drop for variant class features that was recently released. We will be shoring up that gap in January.

Furthermore, this entire line of thinking makes assumptions that folks that want to play highly-customized games are a larger percentage of the population of total D&D players than any data suggests. Unfortunately that is not data that I am able to share, but I will mention it here. And by the way, yes, it absolutely could be the case due to a chicken-and-egg situation. That doesn't make it any less true as things stand now.

 

[Tiggerunner]

Not specifically picking out your post, but a 30 million character sample size is pretty significant, considering there are roughly 40 million D&D players globally at this point. It would be far more representative, for example, than a survey with less than a few thousand responses.

Mathematically, a random sample of a few thousand is better than a non-random, bias sample of a few million. A larger -N of a biased sample just leads to a more pronounced bias. Additionally, WoTC's choice to count characters instead of players is, well, interesting. It keeps any of us from comparing apples to apples without assumptions about the average number of characters per player.

 

[Don Durito]

Not specifically picking out your post, but a 30 million character sample size is pretty significant, considering there are roughly 40 million D&D players globally at this point. It would be far more representative, for example, than a survey with less than a few thousand responses.

The very established at this point in the life cycle of the entire Dungeons & Dragons franchise fact is that most groups do not engage in high-level play regularly. You will see that in our data on D&D Beyond as well as any other research conducted formally.

As a fellow fan, it's hard to think this fact doesn't impact decisions about the game.

See my most recent post. I'm not questioning the overall trends. But I think there are questions that can be asked about the degree of confidence in the specific numbers.

What does it mean where it says in the graph that these are 'campaigns'? Is there some way of tracking that these are characters that are actually being played in active campaigns, rather than just experimental builds?

 

[Umbran]

It was buried deep within the survey write-up, if memory serves.

No. It was right at the beginning of the survey write up.

From the copy google found for me ( RPGnet News )

"Section 1: The Segmentation Study
...
A two phase approach was used to determine information about trading card
games (TCGs), role playing games (RPGs) and miniatures wargames (MWG) in the
general US population between the ages of 12 and 35. For the rest of this
document, this group is referred to as “the marketplace” or “the market”, or
“the consumers”.

This age bracket was arbitrarily chosen on the basis of internal analysis
regarding the probable target customers for the company’s products. We know
for certain that there are lots of gamers older than 35, especially for
games like Dungeons & Dragons; however, we wanted to keep the study to a
manageable size and profile. Perhaps in a few years a more detailed study
will be done of the entire population."

Yes, this excludes anyone who was born before 1964. Today, those gamers are over 56 years old.

 

[BadEye]

Mathematically, a random sample of a few thousand is better than a non-random, bias sample of a few million. A larger -N of a biased sample just leads to a more pronounced bias. Additionally, WoTC's choice to count characters instead of players is, well, interesting. It keeps any of us from comparing apples to apples without assumptions about the average number of characters per player.

That is all data that we have readily available but have not shared. The intent of sharing high-level summary data in our development updates is just that - to provide a 30,000 foot view. We don't have the time or resources, or even a real reason that would benefit the community more than us putting our focus into other areas, to deep dive into all of this.

I'm glad that our summary data provokes this kind of interesting discussion, but it definitely is not provided to appease data scientists. (Although our internal data scientists do plenty of their own research that would be far less exciting/accessible for the community.)

 

[BadEye]

See my most recent post. I'm not questioning the overall trends. But I think there are questions that can be asked about the degree of confidence in the specific numbers.

What does it mean where it says in the graph that these are 'campaigns'? Is there some way of tracking that these are characters that are actually being played in active campaigns, rather than just experimental builds?

We have previously shared that there is no foolproof way to guarantee that any of these characters are being actively played, but take reasonable steps to filter the data the best that we can. For instance, we look at characters that have leveled up on different days and have adjusted hit points in the past and taken short or long rests, etc.

And we have far more than 30 million characters. The 30 million number is filtered down to what we reasonably call "active."

 

[Umbran]

What does it mean where it says in the graph that these are 'campaigns'? Is there some way of tracking that these are characters that are actually being played in active campaigns, rather than just experimental builds?

D&D Beyond has a concept of a "campaign" - it is an object you can create, and you can invite people to assign characters to your campaign. If they are assigned, they become visible to those in the campaign, and so on.

As to whether they are being actively played... well, I don't know why you'd create a campaign in D&D beyond, and have various people put their characters in it, except for purposes of play. They should have have information that could be used to identify campaigns that are being used. Various kinds of activity - character hit points getting tracked, long and short rests being used, spells being cast, leveling up, and so on - could all be used as indicators. I don't know if they did use that information for this, however.

Edit: Whoops! Didn't see Badeye there. Glad to see my guesses weren't too shabby.

 

[tetrasodium]

Saying DDB "only supports the base rules as written" is not an accurate statement. The biggest gap at the moment is the gigantic UA content drop for variant class features that was recently released. We will be shoring up that gap in January.

Furthermore, this entire line of thinking makes assumptions that folks that want to play highly-customized games are a larger percentage of the population of total D&D players than any data suggests. Unfortunately that is not data that I am able to share, but I will mention it here. And by the way, yes, it absolutely could be the case due to a chicken-and-egg situation. That doesn't make it any less true as things stand now.

No it's much worse than that, ddb has never added support for WotC's own variant rules...

• "Variant: Skills with different abilities" is detailed on PHB175. "let me setup a thing & struggle with it so I can tell you the results of my barbarian's strength(intimidate) roll"
• "Hero Points", dmg264. oh boy "I don't have a box for that on my sheet and it's a lot of work to track through other means [so we are just going to use but never actually expend them until you the gm takes them away]"
• ""New ability scores: Honor and Sanity", dmg264-265 "Hold on.. um... I just gotta find where I recorded it here"
• Near as I can tell, tracking of downtime(phb87 dmg127) on the sheet suffers the same problem as hero points

God forbid I want to use the variant UA stuff or supplemental rules extensions like anything on homebrewery, this & other forums, darker dungeons, or anything I create myself. Ddb launched in 2017 & here we are counting the days till 2020 with it still not supporting stuff published in the phb/dmg

 

[Umbran]

God forbid I want to use the variant UA stuff or supplemental rules extensions like anything on homebrewery, this & other forums, darker dungeons, or anything I create myself. Ddb launched in 2017 & here we are counting the days till 2020 with it still not supporting stuff published in the phb/dmg

Mod Note:

I don't think turning this thread into a laundry list of faults in D&D Beyond is going to be constructive. The folks at WotC have official channels for feedback. Let us not inundate someone in the know with such, as it is apt to disincentivize them from giving us information on the topic at hand.

Thanks, all.

 

[BadEye]

No it's much worse than that, ddb has never added support for WotC's own variant rules...

• "Variant: Skills with different abilities" is detailed on PHB175. "let me setup a thing & struggle with it so I can tell you the results of my barbarian's strength(intimidate) roll"
• "Hero Points", dmg264. oh boy "I don't have a box for that on my sheet and it's a lot of work to track through other means [so we are just going to use but never actually expend them until you the gm takes them away]"
• ""New ability scores: Honor and Sanity", dmg264-265 "Hold on.. um... I just gotta find where I recorded it here"
• Near as I can tell, tracking of downtime(phb87 dmg127) on the sheet suffers the same problem as hero points

God forbid I want to use the variant UA stuff or supplemental rules extensions like anything on homebrewery, this & other forums, darker dungeons, or anything I create myself. Ddb launched in 2017 & here we are counting the days till 2020 with it still not supporting stuff published in the phb/dmg

You're actually missing my real point - the amount of the player base that uses those variants that are available in the DMG is incredibly small.

Since we have hundreds of things on our roadmap, those variant rules used by such a small percentage of players do not take priority over other features that would be more widely used and impactful.

Of course I would love for us to support everything that players need (and we won't stop working until we get there), but the reasons behind our prioritization still stand.