One of the hardest parts about D&D 3.x is the wide variety of strengths any given party can have. We are given hundreds of monsters and told, with shaky rational, that these monsters should be able to fight a party of adventurers of that level. The DM's job begins there, where he must figure out how good his players are at character building and battlefield tactics, and then challenge them appropriately.

DMs fail for a variety of reasons. Some are poor storytellers. Others lack the objectivity to remember that they are not playing against the PCs. These appear to be traits within the personalities of those particular people and are much harder to correct than the single most common issue that plagues new DMs - the lack of experience with the multitude of possibilities that D&D offers.

These are the sorts of DMs that just can't quite seem to defeat the players even when it seems like they're actively trying. They're the DMs that don't know enough to realize when something shouldn't be allowed, and can't seem to correctly identify why a particular character does so well in combat.

While I was pondering this article, I spoke with a few DMs that I haven't personally played with before, and by combining their experiences I found out that this problem occurs even when the group consists of just one experienced player and an entire field of newbies. I came to the realization (in a crazy, monochrome flashback sequence) that in games where I was the sole experienced player, I was practically directing the flow of battle all by myself. Each player would look up to me for advice when their turn came around, and I would marshal them to victory each time - much to the chagrin of the DM, who was designing encounters for a mostly inexperienced party.

It's something I've dubbed the "Munchkin Osmosis Effect", wherein the power level of the party is more or less equivalent to the most experienced player in the group. This holds true especially if the player is present when the others are creating their characters ("No, don't take toughness. It's worthless"). You could have a group comprised of six raw recruits and one seasoned veteran, and that roper is getting totally smoked and there's nothing you can do about it.

The initial reaction seems to be to bump the challenge level of the encounter up, and you certainly will need to do this. But that's not what this article is about; we may revisit it soon, but for now let's discuss some ways to limit the Munchkin Osmosis influence and let new players discover D&D in their own ways.

1. DO Stay in Character - During a battle scene, keep things in character. It's much harder for the veteran to influence a fight if he's discussing his tactics openly with the other characters. For best results, be loose with this rule but make its intention clear to the players - they are here to make their own decisions, roleplay, and enjoy the game.
   
2. DO Limit Advice Giving - This will seem absurdly hard at some points, but one of the risks of letting the veteran control the flow of battle is that you are depriving newbies of one of the most essential learning exercises in tabletop roleplaying - trial and error. Veterans have already tried and erred - they know why you should never read the runes on the wall. But newbies don't. Let them have their day in the sun.
3. DO Craft Seperate Challenges - When crafting an encounter, consider somehow seperating the veteran from the main party, and then throwing appropriate challenges at them both. What will happen, essentially, is that the new players will have a chance to see how they hold up to those big bad stat blocks behind your DM screen without the guiding hand of their guardian angel, and the veteran player will feel like an absolute, total badass for soloing an encounter while the rest of the party muddled about.

On top of this advice, there are a few things that you should pretty much never do. They have the right intentions, but typically leave you with unhappy players

1. DON'T Make the Veteran Play Crappy Characters - Asking a veteran player to make sub-optimal choices simply for the sake of being in line with the other players is like asking someone to play through Halo using only the needler - and you're on CO-OP mode. Besides the obvious discomfort the veteran will have with performing sub-optimally, it's a blurry directive at best - how good is too good? What spells are ok? What spells aren't? As long as a character isn't game-breaking when compared to the game itself, he should be able to be played.
2. DON'T Ramp Up the Combat Difficulty Superficially - A lot of times challenging an experienced party is all about knowing what they can do, and working out a counter against them. All too often have I watched a DM get totally crushed, and knee-jerk his way back to victory during the second session, leaving the party broken and bloodied. Worse, sometimes the party doesn't actually care if you send another 10 fighters at them. Some parties will always make quick work of fighters. When improving combat difficulty, work smarter, not harder. Understand what the veteran does well and plot against that more than anything else.
3. DON'T Expect the New Players to Adapt - Although many do not realize it, veteran players slowly begin to erode a new player's self confidence and independence in the context of D&D. They will lose their will to try new things, lose their likeliness to develop their own fears and tastes, becoming optimization machines because they've been taught to do it right or not to do it at all.

It's important to stress that I'm not an opponent of high-level gaming (I mean in the optimization sense, not actual character level). Most of my experience in character optimization and build theory comes from my own endeavors in character creation. I have created some monsters, and typically run very high powered characters myself. I do, however, understand that it takes a lot of time and experience to arrive at that point in a gamer's career, and if there's one thing that D&D has taught us, it's that it isn't about where you're going. It's about the journey there.