
The goal is fun, right?
The other week I made the following prediction:
Providing that an eventual 5th Edition continues in largely the same direction as 4th Edition, the next edition of D&D will eliminate the concept of rolling to hit. Powers will always be effective to some degree; only their degree of effectiveness will retain a random element.And I said if I got comments, I'd provide the rationale. Well, I got (as of today) 23 comments, so I guess that explanation is owed.
1) Minimising player downtimeIt's a truism to say you can only enjoy playing the game if you are, in fact, playing the game. When a player has no meaningful input into the proceedings, they're not a player, they're a spectator.
D&D historically has had four key situations in which players were not able to meaningfully contribute. Firstly, when players were unconscious or dead. 4th Edition addresses that issue by making accidental death much less likely, and by giving unconscious players saving throws to avoid slipping closer to "death's door" (with an associated 1 in 20 chance of regaining consciousness).
Secondly, when players are engaged in a challenge which tests the skill of only the most proficient member of the party.
Diplomacy is a classic example, where the best speaker is often the only speaker. 4E hasn't done a lot about that, although the skill challenge system appears to at least recognise the problem.
The third situation of downtime is during combat when it's not the player's turn. The "attack of opportunity" system gives players an out-of-turn action under certain circumstances, thus requiring them to pay attention to the board-state. Also, the increased emphasis on team positioning and buffing party members means that players need to stay alert to call for bonuses and request backup.
The fourth and final situation is the most relevant for our purposes, and that is when, during the player's turn, they take a null action. That is to say, an action which creates no change to the state of play. The most common example is rolling to hit and missing. Play goes on, with the player having contributed nothing.
Missing is simply not fun. Having waited a full round of initiative and then achieving nothing is functionally identical to skipping your turn. If you expend an encounter power or daily power and miss, you're actually worse off than when you started. It's hard to argue that powers with an "on-miss" effect, or powers with the "reliable" keyword (not expended on a miss), aren't palpably more satisfying than options with higher rates of risk regardless of the proportionately higher return.
It's a big issue for new players. Coming to D&D as a newcomer and watching roughly half your attacks go wide can make you feel impotent and a liability to the team. It's immensely frustrating and more than a little dull. Newcomers more than anyone need to see their initial experiments yield positive results, but it's hard to optimise at low levels and you're locked out of the most effective power options. Your first experiences with the game can be the sessions where you're most likely to watch entire encounters go past without having achieved anything.
Improving the overall experience, eliminating frustration, and making the game more accessible to newcomers means eliminating the possibility of missing. If you're going to continue to go down the path that 4E has started it's an inescapable conclusion.
2) Tactical thinking4th Edition emphasises tactical thinking. That is to say, it asks you to choose between known options with predictable outcomes but complex interactions. The real skills it calls upon are not managing risk, but rather efficient targeting, optimised positioning, and teamwork.
Risk runs counter to tactical thinking. It can turn bad moves into accidental successes and solid plans into disaster. Yes - that is a realistic outcome. But it doesn't automatically make for good gaming.
Tactical gameplay is learning gameplay. It's about experimenting with new ideas and assessing their effectiveness. It's about adapting to known external changes and evolving your technique to deal with new threats.
Risk inhibits learning. It provides a discontinuity between action and results. A good idea can be made to seem bad, and sloppy thinking can be hidden behind improbable victory. Risk makes it harder for new players to see where they're going wrong and it's frustrating to experienced players who are denied the results of their tactics thanks to occasional probability skews.
To create a more open, understandable game table, and make the game more accessible for new players, risk needs to be minimised.
3) RedundancyRolling to hit is redundant. When players make an attack action, they are making two separate rolls to determine its effectiveness - to hit, and damage.
There's no need for it. Why not have a single roll? Either just a "to-hit" roll, with the margin of success determining the damage, or just a damage roll, with every attack assumed to be successful and only the extent of success in doubt.
Every third thread on the official forums is, "How can I make combat go faster?" When we're looking at the next edition, there's a clear improvement staring us in the face: eliminate the attack roll redundancy.
4) ComplexityThe roll to hit is by far the most complex roll in D&D. The to-hit equation is 1d20, plus half your level (rounding down), plus trait modifier, plus weapon/implement modifier, plus feats, plus buffs, plus race/class bonuses, plus conditional modifiers including charging, cover and combat advantage.
All that maths goes to a single question: did I hit? It can be eliminated by uniformly answering: yes.
Maths is not, normally, fun. And in any case, it's not what D&D's here to do. The rulebooks don't bill it as a game of "bold warriors, mighty wizards, and mental arithmetic". There's been a consistent trend since early editions to do away with this kind of number crunching, which covers the elimination of THAC0, the re-engineering of dice rolls to make "high" always equal "good", and a major overhaul of the Armor Class rules.
The multi-variable maths attached to every attack is an albatross around D&D's neck and it's hampering the game's acccessibility. In 5th Edition, it has to go.
5) Over-specificityRolling to hit is unnecessarily specific.
Look at it this way: let's say you hit on a 5. So that's 75% of the time. You hit on three out of every four attacks.
Why are we micromanaging? Why not just say you hit
all the time, and do
33% 25% less damage when you hit? It's the same end effect. And if you want to (despite the above) retain risk and randomness, you just vary the damage, so that the new damage per hit is an average, subject to dice roll.
How it might work.
So if we eliminate the to-hit roll, what does the game look like? What's armour good for? What's weapon specialisation about?
The simplest answer is to put it all into the damage roll. Armour represents damage mitigation, soaking a certain amount of incoming HP loss (which is more realistic anyway). Weapon competency increases your damage, or decreases the effects of armour, or results in debuffs or other non-damage penalties to the target.
Another answer is to rip out the to-hit subsystems entirely. Eliminate them. Feats can concentrate on buffing particular powers or classes of powers; proficiencies unlock new power categories.
The final result, I have to confess, will not look a lot like D&D as it's been understood between its origins and 3rd Edition. It will, in fact, be a hugely different game. But that's clearly not something that's significantly bothered the 4th Edition designers, and in the end result a better game is a better game.

So - that's my rationale. Now go nuts. Defend the roll to hit.
UPDATE: This post was featured in the July 2009 RPG Blog Carnival at 6d6 Fireball, which had the theme of "Dungeons & Dragons". Big thanks to 6d6 Fireball for hosting and for including this article. Now go check out all the other excellent entries!