Dungeons & Dragons has an underlying tension between the power of the monsters and the power of the characters. There will be combat in a D&D campaign, and what happens then is of paramount importance to how the game plays. If every combat ends with the monsters being destroyed in one round, then it’s not a very satisfying game. Likewise, if the characters end up losing every battle in one round, then, once again, the game will quickly fail.

No edition of D&D has had things that bad. One of the elements that mitigates that is that players and monsters have a large range of possible power levels. Although a group of trolls might tear apart a first level party, a fifth level party will find the job a lot easier, whilst a tenth level party will tear the trolls apart. Incidentally, this also provides one of the bonuses for good game play: player characters that survive gain the ability to confront and defeat greater and greater threats, evidence that they are getting better.

The original form of D&D, the megadungeon, places monsters in the dungeon based on how difficult they are to defeat. The easiest monsters go on the upper levels of the dungeon, the most difficult monsters go on the lower levels of the dungeon. The players are able to find their own difficulty level based on their experience. You go down until you reach a level that provides a challenge (and thus good rewards), but isn’t too difficult to survive on. The balance between monsters and characters is one that finds its own equilibrium, assuming the monsters are ranked correctly. Due to the simplicity of original D&D, this isn’t too hard to do: Hit Dice do most of the work for you.

Unfortunately, that self-balancing of encounters goes out the window once you start writing adventures for specific character levels. An adventure like Steading of the Hill Giant Chief doesn’t allow the players to go away and do another adventure instead: they must survive or perish! So, the adventure designers need to judge how difficult individual encounters are and how difficult the adventure is as a whole.

The earlier editions of D&D are fairly inexact about this. There are a few guidelines based on hit dice totals, but, for the most part, the judgement depends on the skills of the DM. As the monsters are fairly simple, this isn’t too hard to do.

3^rd edition, however, is a much more complicated system. In AD&D, you can directly compare a 5^th level fighter and a 5 HD monster. In 3^rd edition, the two can be greatly divergent from each other. Indeed, two 5^th level fighters can work in completely different ways, and that’s before you even consider the plethora of additional classes and options that came out for the game! The solution, as envisaged by the D&D designers, was to invent a Challenge Rating for each monster, which would rate the monster in comparison to an “average” party of four characters.

The Challenge Rating system was a bold experiment. It had a lot of really good points, and although it failed, it didn’t really fail that badly. Challenge Ratings, even when they were inaccurate, weren’t that inaccurate. Unfortunately, the system had a fundamental flaw which stopped it from being as accurate as it could be. Interestingly, there are two solutions to the problems raised by 3^rd edition. 4^th edition took one route, and Next is taking another.

The problem arises from the scaling of Armour Class and Attack Bonus as levels increase in 3E. D&D has always had fighters improving their attack bonus as they’ve gained levels. However, 3E had them increasing it faster than it ever had gone before, or would again. In general, the fighter gained a +2 to attack for every level they gained. AD&D has the rate as +1 to attack for every level, or just a little above that, as does 4^th edition. When you add to that the fact that the Armour Class of monsters is also increasing at about the same rate – although not codified that way – a difference of even three levels between characters makes a huge difference in how effective they are in combat.

The effect of this scaling is that a group of lower level monsters just can’t hit the players, and are hit back entirely too easily. Challenge Ratings work quite well when there’s between 1-4 monsters, but once the numbers get up, the mathematics gets problematic.

The 4^th edition solution to this was to design monsters so that the number appearing also was significant. Thus, we got minions, standard monsters, elites and solos. Each of those categories can take on a particular number of player characters. Thus, instead of a monster being rated as a monster that challenges 5^th level characters, it’s a monster that challenges two 5^th level characters. It’s a major shift in paradigm, and it solves – quite effectively – the problems that 3^rd edition had with its own Challenge Rating system.

The failure of 4^th edition in this area mostly came down to minions. Minions are a great design idea – four minions can take on one character – but the mechanics proved problematic. One hit point? Not bad. Defences that challenge the PCs of an appropriate level? Fantastic! Player powers that kill minions without needing to hit them? Oh dear – things fall apart in a heap.

D&D Next is promoting the idea of bounded accuracy, which – in a lot of ways – is like the D&D we used to have. In original D&D, there is no monster in the rulebook with an Armour Class better than platemail and a shield. None. The effect of this is that players can always hit monsters, even when they have new characters. Likewise, monsters can always hit characters – the very best AC a player character can attain in original D&D is -3 (interestingly, a magic shield only works 1/3 of the time…) AD&D made the potential ACs of monsters much, much better – and thus more troublesome – but the vast majority remain in the hittable region. D&D Next, in theory, will pay more attention to how it works than how AD&D ended up.

With Armour Class a lot more static, we’re back to the idea that hit points (and Hit Dice/Level) are important in determining the threat of monsters. They’re not the only thing, of course. The attacks the monsters and characters can make also have a lot to do with the balance of it all. However, if the players are always able to hit monsters (regardless of level), you don’t get the problem of adding monster numbers not always raising the threat level.

One interesting factor about the playtest is that it considers that once you double the number of players in the group, the difficulty of the encounter goes up a step. This is quite fair; numbers make a huge difference to how difficult an encounter is. Minions, in theory, might have run into this problem, but their ease of destruction negated a large part of their potential threat. Solo monsters ran into major problems when controlling magics (hold, daze, stun) were factored in. So, a horde of orcs, even for higher level Next characters, is likely to be an interesting challenge.

I still haven’t played enough with D&D Next to properly assess how the implementation of bounded accuracy will work in the longer term. Several of the playtest packets were quite problematic in terms of the monster stats – they were often too fragile or too tough – but I’m quite willing to suspend judgement until I see the full game.

One thing is certain: in this age of published adventures, allowing designers to make informed decisions about the relative power between monsters and character parties is incredibly important to the health of the game.