When you're writing a book, or a movie, or really any story narrative, one of the key questions that a lot of people will ask you (when giving advice) is: what is at stake? Sounds like a simple question, right? But, you'd be surprised how many people can get it wrong depending on the situation. Ever watch a movie and feel like a sense of urgency was lacking, or wondering why the sense of urgency was there at all? That is because, somewhere along the way, the movie failed to communicate the stakes to you properly. This can happen in RPGs as well, and when it does happen you'll often end up with either the GM or the players sitting there and scratching their head wondering just what the hell is going on. So, let's look at the stakes, how to set them, and what you can do with that.
First, An Anecdote
So, a few years ago I was running a game. It wasn't a particularly big game, but that doesn't matter. Now, in the game an evil wizard was doing the old "Destroy the World" plot, and was in the last stages of their plan. Simply put, the PCs had a short time frame to break into the castle, make it to the summoning chamber, and stop the wizard. Simple right? I thought so. The only problem was, the PCs didn't go straight for the mage's chamber, they went around killing the other monsters, clearing rooms carefully and thoroughly, and generally looting the room. Eventually, the group even sat down to rest and recover after a fight. What did I do? Well, I smiled, calmly asked them who was taking first watch, and then had them make a listen check. I explained that they heard a rumbling shudder through the castle. The PCs went on guard, took up defensive positions, and that was when the castle crashed down on top of them, wiping the group out instantly. I think explained how, unhindered, the evil mage managed to summon not just one but three greater demons from the infernal realm, and was now marching them across the land killing all who opposed him. Needless to say, the PCs were kind of pissed.
So What Happened?
What happened in that situation? Well, we had a communication break down somewhere along the way, and I didn't take any efforts to remind the players of what was going on. I had thought that I had clearly communicated "If you don't get to the wizard and stop him in three hours, the world ends", but the players had picked up "if you don't stop the wizard, things will get worse" and had completely missed the time limit. Because of this, they thought they had time (I have no idea why players think this...and I have found myself thinking like it myself when a player, so it has something to do with being a player) to clear the castle "properly" when in actual fact they were on a timer.
Now, there was more wrong here than just stakes - and I have grown significantly as a GM since this incident - but the communication of the stakes is still the key issue here. Had the stakes been presented properly, the world wouldn't necessarily have ended, and I wouldn't have had four angry PCs on my hands. Ah well.
The World Is Ending
Now, you don't always have to put the world on the line when communicating stakes. However, you want to remember that the closer the bad result is, and the bigger it is, the more urgency that is expected from everyone. You can break your own stakes by failing to communicate this urgency. Giving the players 1 hour to save the world, then letting the characters sit back and calmly discuss how to do it and amass equipment for 45 minutes is going to break the pacing more than anything else. Especially if you don't then leave them with just 15 minutes to resolve everything else.
Generally speaking, the larger the stakes, the more urgency that is needed. The smaller the stakes, the more relaxed you can be. After all, is there really a need to panic if the stakes are as simple as "if you don't stop that man, he's going to run a red light"? No, not really. However, if you make it "if you don't stop that man, he's going to run a red light and hit that lady and her kid." Well, now we have more urgency and reason to stop him right? Keep going though, "...he's going to run a red light, hit that lady and her kid, and then crash into that train carrying radioactive material and causing a nuclear explosion." Yeah, now we're talking.
Stakes Determine Response
The other thing to be careful about here is that the stakes also determine the level of response you are going to see. Looking at the example above, when the stakes are just running the red light...you're probably not going to do all that much to stop the guy, right? However, once it is clear that letting the guy go is going to endanger hundreds of others, now a lot more in the way of response is justified. Putting a PC into that position, I wouldn't be surprised if a quick acting/thinking PC simply blew the guy and his car up before crossing the red light. Sure, the guy is dead, but you've just saved two people and prevented nuclear fall out from happening.
Basically, you need to be aware of the mentality that "the ends justify the means", and the bigger the stakes are, the greater the ends' ability to justify any means becomes. There isn't much you can do about this either, aside from hit the players with the consequences. My recommendation? Don't fight it, let it happen, and let things fall out naturally from it.
Stakes To Control Actions
So, now - and finally - some practical application. You need your villain to make an escape, but just know that your players are going to chase to the ends of the world. So, what do you do? Up the stakes of pursuit. This is effectively playing with consequences, but it bears mentioning here too. So, you don't want to have those PCs chasing? Give them something else to do. Leave a bomb that needs defusal, a child that needs saving, someone who needs to go to the hospital immediately, or do all of the above in one go.
Just remember to keep the stakes in mind, and make sure to present them to the players clearly. Then, let them make their own choices for how they go through things.