We're returning to D&D 4E after a long absence from it, coming back from Savage Worlds since we've really missed HP-based battles and videogame-like logic.

However, as we talked about our new campaign based on the novels and videogames of .hack//, the question struck me: "So, who are you gonna play this time?"

My players like when I play a PC as well, and I sincerely like to build characters that aid in storytelling. I'm creating some basic NPC allies, for some quests here and there and as part of the narrative, since they like an "anime" storytelling style where there are lots of secondary characters with their own backgrounds.

But they want me to have a full PC traveling with them too, especially since we lack someone in the Leader role and no one ever wants to play it.

I've made characters for 4E before whom I played as, however I've found this has two large issues:

  1. While controlling my character I noticed I pretty much avoided having monsters attack me. When I corrected this, I pretty much started intentionally making them go all out on me. It made the battles weird since I felt I was playing "Punch U Punch me" with my own face...

  2. When that doesn't happen, my PC just makes battles easier. I know my Bosses and Enemies, but I want to hurt my players, not kill them, so I pretty much end up screwing up and accidentally revealing weaknesses in order to well... defeat myself and save my team at the same time.

How can I balance this? I feel like it happens when I give them allies for quests as well, not just when I'm playing a PC.

I've considered letting the players control my PC during combat, but of my three players, two of them do not like controlling more than their own PC (it was a problem in Savage Worlds, where it's common to hand the players control of NPCs during combat), and the third is new to 4E and wants to focus on his Fighter. In case it's relevant, the other two PCs are a Rogue and a Psion.


5 Answers 5


As a fan of the Suikoden franchise, I also like creating basic NPCs that follows the players (the Smith is a common one).

You seem to already avoid the pitfalls of most GMPCs: disliked by PCs, and abnormally strong and awesome. That's great.

Since your problem is mostly related to combat, I would suggest an option I use when NPCs become good enough to be useful in combat (otherwise they stay in the back, or simply give a passive bonus): let your players play them.

Make a character sheet, control the character out of combat, and when combat happens simply hand the sheet to the players. Either have one player control it (the one most familiar with the rules, or whose character cannot play at the moment), or even make it a group thing with all players (you included) deciding what the NPC should do.

For lazy players, consider statting the NPC as a monster, with a couple at-wills and some recharge powers, just like a monster. The (significantly) reduced complexity of the character might make them less reluctant to take on the additional effort of managing the NPC's combat.

However it's statted, handing over control of the character makes it easier for you to avoid most bias, and for the players to better relate to the NPC.

  • \$\begingroup\$ This is good advice in general. Considering this comment left after you wrote it, it might need to be reworked a bit though? \$\endgroup\$ Commented Jun 13, 2014 at 18:33
  • \$\begingroup\$ @SevenSidedDie Ah, indeed... This actually makes the whole answer irrelevant, yes. I will think of a way to reformulate it (any advice about how to do it is appreciated!). \$\endgroup\$ Commented Jun 13, 2014 at 19:18
  • 1
    \$\begingroup\$ I still like this idea, despite that feedback from the asker. Maybe sell its knock-on benefits, such as keeping everyone engaged when it's not "their" turn because they have this other PC to make decisions for as a group, learning to coordinate powers more (easier when you control both), and that sort of thing. Basically, highlight why it's a good solution despite the players' reluctance. Add how to mitigate that reluctance, and it'd be even better. \$\endgroup\$ Commented Jun 13, 2014 at 19:55
  • 3
    \$\begingroup\$ In the past, I and my group have employed this strategy to great effect. To get around the extra effort of running an entire second combat character, we generally stat the NPC as a monster - just a handful of powers, with recharge timers and such, and an at-will or two. It makes their role more clear, still allows them to be very effective, and greatly simplifies running them in combat. The players may not shy away from the extra effort if it's not a full character. \$\endgroup\$
    – Travis
    Commented Jun 13, 2014 at 22:08

You can try a giving the GMPC a "combat mode setting" similar to how it's done in some video game RPGs. For example, your players set the GMPC to play a defensive combat role—meaning the GMPC will defend a given player—or an aggressive combat role—meaning the GMPC will always attack the nearest monster.

Allowing the players to set these roles lets them to choose how the GMPC fights, but not necessarily require them to make its combat decisions. They can change the stance during combat to alter the GMPC's strategy accordingly.


It seems to me like the problem you're actually having is Player Knowledge vs. Character Knowledge, except in this case it's GM Knowledge vs. Character Knowledge.

Using GM Knowledge

You say you "pretty much avoided having monsters attack me". This might mean your character maneuvers in such a way to avoid being attacked, which would be reasonable, but I think you probably meant that when you're controlling the monsters you avoid attacking your character. Then, to fix it, you had the monsters all attack your character. This means that your monsters are taking into account your knowledge that that character is controlled by the GM, not the players.

You say "my PC just makes battles easier". Well that seems fine, I mean you're another PC, you should make the battles easier.

More importantly, you say "I ... end up ... revealing weaknesses in order to ... defeat myself". This implies that your character takes advantage of weaknesses your monsters have that you know about because you're the GM, but which your character wouldn't otherwise know about.

Stop It

This is easier said than done, but here's how you divide character and player knowledge:

When making decisions for your character, only take into account things that character knows.

Simple, right? Well, maybe not. Keep in mind that in the scenario you're describing, you have to do this for not only your PC but the monsters as well.

The trick here is to keep track of what the characters know. They certainly don't know that there are such things as a GM or players, so making a decision about who to attack based on that is right out. When you're controlling your monsters, ignore the fact one of the PCs is yours and make the same decisions you would make if it were a real PC.

When you're controlling your PC, only attack weaknesses he would know about. If this means you have to make knowledge checks in battle, then do so.

Making it Easier

I am about to start a campaign as dungeon master where I am also going to play one of the PCs. The primary motivation for this is that the party only has three characters and they wanted a fourth.

I think it's funny someone mentioned 'Hodor' in the comments up there, because I'm basically going to play Hodor. I houseruled it and allowed myself to create a character with 6 intelligence. This will make it easier for me to keep track of what he knows: nothing. When the party is planning I'll have essentially no input because I'm playing an idiot, so I won't have to worry about already knowing what they're walking into. In combat, I'll just take the most obvious and direct course of action, usually just charging whichever enemy is closest.

My players are completely on board with having an idiot meatshield.

There are other possible solutions, of course. Make the PC speak a different language. Make him a seer, to explain the knowledge he shouldn't have. Make him constantly taunt the enemies so they have a reason to attack just you. Be creative.


We've been playing this campaign for a couple months now. The character is a Half-Orc named Horko. Everybody loves Horko. One of the party members is a former militia lieutenant named Jett, and he is constantly giving Horko orders, often with amusing results.

This can also be a convenient way for me to nudge the party back on track. For example, they were once held prisoner by a priest dominated by a succubus. When they escaped they took the priest prisoner, which I had not anticipated. The priest wouldn't have told them anything, as he was dominated, but he had a pretty flamboyant personality that would've been difficult to keep up longer than I had intended.

When the party left him unattended the priest tried to escape, so Jett yelled at Horko to "get him". Horko rushed him, I rolled attack and damage, and he died. I solved my problem, and everybody had a laugh.

Horko works well in combat too. He is a Fighter Slayer specializing in charging with a greatspear. Every turn I either have him do what Jett (or occasionally one of the other party members) tells him to do, or just charge the nearest eligible target. Two of the party members can grant free melee basic attacks to allies, and they have fun ordering Horko all over the battlefield.


I second Scrollmaster's answer of letting them play the GMPC mechanically if not in roleplay. You could even switch who plays the GMPC each session. This may lead to some players discovering classes they usually avoid, or let them sporadically play as these classes without being stuck with them for an entire camapign, too.

If the players are not up to this or you have a reason for not letting them control such characters, we go back to the issue of how to make these characters and the opposition act towards each other without too much bias.

Let the dice decide!

This is a trick based on the Mythic GM Emulator where you define the probabilities of things happening and roll dice to determine the final outcome.

Here, I suggest that whenever you feel your own knowledge as GM is at risk of introducing unwanted bias into your decisions, let the dice decide.

To choose who an enemy is going to attack, consider their point of view and note any preference they might have. A clever foe aware of the party's individual strengths and weaknesses might have a preferred target. A more common foe might have no idea who's the "best" target, or might not care. If multiple choices are possible, weight them, if needed, and roll a d100.

4 PCs and 1 GMPC, no obvious target choice: 20% chance each.

d100 = 34

You pick the second PC (use the order of the sheets, of players around the table, etc...)

Same team, but one PC should be a favored target of the villain. Yet, you want to remain impartial. Assign 60% to the favored target, divide the rest among the remaining PCs (and your GMPC if he was not the favored target).

d100 = 40 => You pick the favored target

d100 = 77 => Pick the second remaining character

You can use the same concept for tactics. Maybe you know that the best thing to do is to flank PC #3, but you're not sure the gnolls would... Just ask this as a question (see below for the technique).

An added bonus to this method is that you will get surprises as a GM and possibly enjoy playing your GMPC some more as you don't decide everything that's going to happen.

For your second point, how much GM knowledge to reveal through the GMPC, I'd use a similar technique combined with your game system.

First, the technique!

  • Define a simple "yes or no" question you do not want to answer directly.
  • Define the weight of either answer. If it's likely that he knows, pick a number from 51 to 99, with 99 being near-absolute certainty (because the GMPC is a childhood friend of the villain, has adventured with them in the past, etc...) If it's unlikely, go with 1 to 49 (the GMPC doesn't know the villain personally but you defined he did some research on him). If it could really go either way, aka "I have no idea!", pick 50.
  • Roll a d100
  • If you roll below of equal to the chance percentage, it's a Yes to your question. If you roll above, it's a No.
  • If you want some twists, consider results very close to the chance percentage as being "Yes, but" and "No, but" answers and try to come up with the "but" part.

Boom! You get a random but informed (through weighted choices) answer.

Now to the problem at hand, as an example.

"Should my GMPC reveal the villain's weakness?"

Does the GMPC actually know about the weakness? He could but you're not sure? It might be time to use a knowledge skill check of sorts, if not an Intelligence check! I strongly suggest to look at existing mechanical aspects of the GMPC to define what they know and choose to do. If something is a clever tactic, check their Intelligence. If it's a matter of "getting" the enemy, make a Sense Motive check for them.

If it's not a matter of skills and more about GM bias, time to use the previous technique and turn this into a question: "Does the GMPC know about the villain's weakness?" We're looking for a Yes or a No. The GMPC knows quite a bit about the villain, so let's go with 75%

d100 = 95

It's a No, and since we're very close to the upper limit, let's say it's an Absolute No. Somehow, in spite of all the GMPC knows about the villain, he's never found out about any specific weaknesses and is thinking he doesn't have any! Let the GMPC act this out, warning the other PCs and reacting to this lack of knowledge with their personality (being in favor of fleeing, or going all out in a blaze of glory, etc...)

What if he does know the weakness? Should he act on it? Is there a reason for him to, or not to? Is there room for hesitation?

The GMPC does know the weakness of the villain, but this villain was a dear childhood friend before turning... and something it tugging at heartstrings. Will the GMPC let the PCs know and help possibly kill his former friend? Is he going to keep it secret and try to push the PCs away from the confrontation?

"Does the GMPC reveal the weakness?" 50%, because it's such a conundrum. d100 = 45. Yes. Or "Yes but..." considering how close it is to 50%. He does reveal the secret, but maybe not the whole of it? Or perhaps he joins up with the villain right after to somehow make up for this treason?

d100 = 67. No, the GMPC does not reveal the secret. Still, knowing about it, he may use it to his advantage if things start to go sour. Either use this to save the party from a TPK and have a cool roleplaying scene afterwards ("How did you know his weakness? Why didn't you tell us?!") or put this as a new question if you're unsure what to do.

d100 = 1. Yes, the GMPC reveals it, and he also pushes the party towards using it, to finally end this, once and for all.

Oh, and if you don't like the answer, don't let the dice rule, you are the GM. The randomness just showed you what you knew and wanted all along, go with it.

Also, you are still responsible for everyone having a good time and sometimes this means you should have the GMPC give a hint simply to keep things moving. Let the dice help you but don't solely rely on them.

  • \$\begingroup\$ I quite like this solution. It's a natural extension of a process I've always used successfully for the opponents in a combat when I can't decide who they'll focus on or what tactics they'll use next. \$\endgroup\$ Commented Jun 13, 2014 at 21:10

If no one want to play a leader, don't have a leader.

Encounters should be balanced by the XP budget so the lack of a leader should not be devastating. The biggest thing players will miss is healing, but there are plenty of consumables, daily item powers, and the option to multi-class into a leader that this shouldn't be an issue. PCs playing other roles should be able to optimize well enough that they don't need leader buffs, out of turn attacks (a la Warlord) to succeed. Roles are nice and they offer options, but a party doesn't need to have every role covered to succeed.

When initiative is rolled you stop being the Narrator/adjudicator of the world and become the antagonist.

This is not to say that all the monsters the party faces in an encounter should focus fire and coup dying players in an attempt to TPK, but rather that your job as a DM during encounters is to provide a legitimate combat challenge for the players to solve (while trying to enjoy things on your side). You're tracking the initiative order, monster HP, remembering all the status effects on monsters, and trying to remember any immediate interrupt and immediate reaction powers they might have. You're too busy to run a PC for the players.

Make up for the lack of a leader by providing lots of nice consumables in their loot as they go through dungeons and adventures.

Various healing and resistance granting consumables should be given out at regular intervals. There are also a ton of buffing consumables adding damage on hit, boosting AC or to-hit as well as granting a damage type change. All of these should become a staple of the after-the-fight scavenge finds that the party uncovers. Additionally be sure to provide "safe" places for the PCs to take short-rests during sessions so they can spend those surges to heal out of combat.

  • 3
    \$\begingroup\$ This addresses the supposed lack of a Leader very well, but that's only a minor throw-away mention in a question that is primarily about how to use a PC anyway, since the whole group wants it to be done that way. This answer could still be on-point, if it contained a section explaining directly that this preference should be overridden and why. \$\endgroup\$ Commented Jun 13, 2014 at 18:31
  • \$\begingroup\$ @SevenSidedDie I thought section 2 covered that pretty well. I could rearrange the order. \$\endgroup\$ Commented Jun 13, 2014 at 18:42
  • 2
    \$\begingroup\$ Section 2 just says what they already know, though: the question is asking how to pull it off despite all that. Since they already know the obstacles from direct experience, the "no really, don't do that" would have to be much more than just an intro-level description of the DM's job. \$\endgroup\$ Commented Jun 13, 2014 at 18:48
  • 3
    \$\begingroup\$ I'd love to see this answer grow into a "don't do it" one, but I have to agree with SSD that it needs work toward that goal. \$\endgroup\$ Commented Jun 13, 2014 at 21:03

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .