Published by Hasbro's Wizards of the Coast, Dungeons & Dragons Fifth Edition is the latest (as of 2020) revision of the world's first published role-playing game.
There are many great things about the 5th edition of D&D compared to previous editions, and there are some other considerations that should be taken into account when evaluating RPG systems & settings used with at-risk and high-risk populations.
One of the best improvements of D&D5e compared to 3rd and 4th edition, is the extensive enhancements to developing more multi-dimensional player characters with tools to help flesh out a richer background than the defaults for previous edition.
Many of these enhancements have little or no real mechanical advantage, and really only help with enhancing the role-playing experience.
We do recommend 5th edition over 3rd or 4th edition D&D if you are looking for a recent version of D&D that brings more of the focus back to ROLE-play instead of ROLL-play. However, not as much as the earliest editions: OD&D & BECMI D&D.
Short-comings/considerations/concerns for at-risk and high-risk populations
There are several significant problems with many modern RPGs in recent decades, and specific to D&D 5e (as well as 3rd & 4th edition to some degrees).:
1. Lack of behavior guidance tools (we recommend using AiME with D&D5e instead, or better yet TOR or DWAITS if looking for non-D&D settings).
2. Extreme moral relativism, lack of in-game-system rules consequences for PCs doing evil things in the game world. Older editions and other RPGs include clear game system consequences for maladaptive / evil acts. Instead the GM/DM must use in-story reinforcements/punishments, such as city guards, etc., to try to guide the players to play their PCs as heroes (unless playing an intentionally evil campaign, which can be done well with mature groups but requires extensive additional training and exceptional players to keep it running and not devolving into chaos).
3. More video-game like design for resource consumption (short rests, long rests) unless implement optional rules to increase "grittiness" to be more in line with previous editions. This tends to mean that participants will be less effective at longitudinal planning than other editions and RPGs where such resources need to be better managed over time.
4. Combat is not very deadly, which leads to increase in more "murder enthusiast" type behavior using combat as solutions due to feelings of low risk of consequences to excessive combat. The abstract hit points approach of D&D has always suffered from this effect compared to other RPGs, but only really become more dramatically an issue in the last few editions as now it is very difficult (relative to older editions) for a PC to die, unless adding optional rules for grittier consequences. It is a fine line between effective longitudinal thinking, risk assessment skills development, and those who are excessively risk averse.
5. One of the greatest problems is the reliance on either the Mentor Model or mass-learning high barrier to entry for brand new D&D 5e players and game masters (and indeed most versions of D&D and about 99% of RPGs except for BECMI D&D).
6. Point inflation and power bloat starting as already very powerful characters also has a long list of consequences in reducing a lot of benefits found in earlier editions and other RPGs, that is significantly reduced in the latest edition. This also makes it very difficult to have a very long-term campaign where the PCs become so powerful they have lost most challenge. About 10% to 25% of the developmental benefit from earlier editions and other RPGs is missing due to starting with such powerful characters at the beginning.
7. "Dumbing down" of the math, is shown to have reduced the prior mathematical benefits found in earlier editions. These earlier editions, when presented correctly were just as accessible to as many people as the new editions (actually more accessible for other reasons). While research studies in the past showed that D&D helped improve match skills, that is now all but missing in the latest editions except for the most rudimentary of younger youth developmentally.
8. "Dumbing down" of the writing. The continued decline in vocabulary, sentence concept length, reading comprehension, found in many educational institutions and most RPGs is also showing up in the research. Whereas in the past most RPGs were shown to quickly and dramatically raise the reading level, vocabulary, cognitive capacity, reading comprehension, and other related aspects for all ages of participants from about 8-9 years old on up, research is showing that these benefits are now far less and slower in newer editions. While older editions "raised up" everyone and were actually more inclusive, newer editions start at a lower reading level and don't raise up the reader anywhere near as much as earlier editions, and have counter-intuitively became less accessible because of a slew of other issues (font, art, styling, contrast, etc.).
More details are pending, we are still transferring decades of content, thousands of content items, from the old servers (www2.rpgresearch.com and w3.rpgresearch.com ) to the newest one (www.rpgresearch.com).