There was an optimization question recently in my D&D game regarding defenses and hit. My investigation into the matter took me perhaps a bit further down the rabbit hole than was strictly necessary or than I really intended to go, but I did glean some interesting information that only people like me really care about and a discussion I wish to weigh in on. Enjoy.
In 4e, the expected hit chance for a PC is 60%, which is lower than in previous editions. In addition, player scaling is noticeably inferior to that of monsters, resulting in an even worse chance to hit for most characters by the final levels. There are many proposed ways to fix this (in fact WotC has been somewhat addressing the issue in the PHB2 and 3), but the one want to talk about is quite possibly the most drastic: use 3d6 for attack rolls instead of 1d20.
The idea behind this is that it’s no fun to miss 3 rounds in a row, and multiple smaller dice summed give a bell curve average instead of a linear. This means you’re going to see a lot of 9’s-12’s, and very few high or low numbers. To the credit of this idea, that does mean that you’ll hit consistently in an encounter where your Attack Bonus +10 is enough to score a hit (based on average dice rolls, your most common rolls are going to be 10 and 11).
After that, it breaks down. The problem, as I see it, is twofold. First, the system assumes you are going to fight monsters with defenses at or around your Attack Bonus +10. Unless you and your GM have some sort of understanding this simply isn’t going to be true. Soldiers, high level encounters, solo mobs all have defenses significantly above expected values. Yes, you’ll be rolling lots of 10’s and 11’s but if those are misses you’ve not gained anything. And with a maximum roll of 18, you’ve cut off two rolls that will hit high defenses. Furthermore, your bell curve means that once you reach the point of needing Attack Bonus +15 (uncommon but not exactly rare, especially at high levels) the 3d6 system actually lessens the chance for you to hit. The only way this system helps is if all you fight is a whole bunch of “average” monsters.
Second, and far more grievously, this assumes that the only metric by which to measure the game is by the consistency of its combat. In fact, many articles went on to lament that they wished 4e was “more like WoW” (here I assumed you meant more like video games). I love D&D. I love WoW. They are not the same game and if you think they should be you are doing it wrong. Saying that “more consistent combat” will make for a more enjoyable game experience is akin to claiming that the suitability of a car to your needs can be directly measured by the horsepower the engine produces. It’s so stupid I feel insulted for having to read it. There are so many facets to both games beyond the RNG that to suggest that it can stand alone, or even as the most significant factor, in determining a game’s fun is asinine at best.