To begin with, some general advice on the nature of social skills challenges:
First, Skill rolls are not, in terms of mechanics, a substitute or parallel system to fight. You, as a GM, do not have to allow players to exchange skill rolls for combat, simply because they do not have to decide that skill rolls have a significant chance of success. It's like a player who insists on doing a strength check to lift a two-ton rock. Sure, you can let me do the check for the sake of the show, but without any extenuating circumstance, there's no reason for it to succeed even if it shoots. Twenty Some social roles, like some force controls, are simply non-starters.
They should everyone Be not beginners? It is probably too severe and borders the railroad tracks. But it is also problematic to allow NPCs infinitely malleable, persuasive and deceptive.
Second, Even if you decide that a certain situation is susceptible to this treatment, you do not have to resolve it in a single test. The 4e notion of skill challenge is a very useful model, even if the mathematics needs to be adjusted a bit for the situation. There, the central idea is that multiple runs are needed, preferably used against multiple skills in concert, to achieve great results.
Third, In connection with this, you are not required to allow the player to control the pace of this type of extended skill challenge. As GM, you have rhythm control, and if you decide that it would take several long time units (either in the game for several days, or in RPG time of several sessions) to achieve this goal, that is your goal. prerogative. I would try not to be arbitrary In this respect, but it is in your hands.
Quarter and finally, you do not have to treat this as a pass / binary failure where the players are successful and have a new band of minions, or fail and have no consequences. It should be possible, for example, for players to commit errors that are so serious that they are even worse than before (in some way) or that there is partial success, or that something strange and unforeseen happens (in relative terms) that nobody planned on.
To continue, some concrete suggestions for this meeting:
Disclaimer: I have never made my players try to get a band like this, but on many occasions I have tried to influence a large number of people, and I have used the skill challenge motive, and I have this advice to give:
FirstEven if it takes a few minutes to calm down, take some notes about how many successes and failures are necessary before deciding the outcome, and what kind of skills are useful. I would not treat that as stone engraving (especially if the players have a creative idea) but I have found that it helps me to stay objective and prevents me from falling into a too adversary role.
Second, If you are going through the path of extended skill challenge, it is critical to give his players comments on whether they are good or not. My first failures in this sense happened because I knew that the players needed (for example) four successes, and I had achieved three, but I could not communicate a sense of success or progress, so they gave up.
Third, Sometimes it helps to personalize this kind of thing by turning it into a two-sided conflict: the PCs, for example, in front of the Redbrand leader and the main lieutenants, where the prize is the course of action or the loyalty of the rest of the group. This increases the drama (the opponent has a face) but also offers much more scope for reactions, role play and understanding of why the group is not susceptible to persuasion. (Say it, because Lieutenant Thuggy McThuggerson reminds everyone what happened the last time someone tried to tell Captain Brute what he should do: that would be the list of intimidation or persuasion of existing leaders).
Quarter, On the idea of non-binary results, consider carefully what a total catastrophe could be, or how it could be a partial success. Are the thugs so fearful at the regional level that they could put a reward on the heads of the PCs? Could PCs take off a quarter or a third of them and start a gang war? Could local authorities exploit that? Complexity is the friend of the GM.