I'm not really starting a series, just echoing the title of J.D. Wiker's post Sacrificing Sacred Cows: Random Encounters. I really like random encounters and the events they precipitate, and I have to take issue with the points against that he makes.
"Random" means the GM literally doesn't have any control over what happens from encounter to encounter. The GM can, of course, discard results he doesn't like, but then it's no longer random--and you might as well be just choosing them.
GMs have always turned to random generators to introduce ideas and elements that they might not come up with on their own. That doesn't mean that GMs must be slaves to the dice, as there is quite a lot of room between "no control" and "complete control". I would even recommend that a GM using any random table consider discarding the results they get if they don't immediately inspire an appealing encounter scene. Picking and choosing the results you like means that you get the best of both full control and complete randomness: you get a quick injection of ideas and you retain complete control. Anyone who has no control over their game because they're using random elements is going to be in trouble as soon as dice hit the table.
Random encounters make for random Encounter Levels. The GM has to set up the encounters in a pretty narrow band of ELs to match the party's average level, so varying the encounter components makes for more lines on the encounter chart (10 goblins, or 4 goblins and a goblin Ftr 4, or 4 goblins and 4 wolves, and so on).
This is a persistent fallacy of 3e D&D play. I can't find a source right now, but I was reading recently someone's point that ELs were never meant to constrain what encounters you offer to the players. The EL is used to assess the strength of the encounter, making the GM's job easier. Religiously keeping ELs at the "right" spot for a group encourages your players to shut off their brains and charge into any fight they are presented with. Sometimes, they really should run away. Reconnaissance might be in order. They might even want to consider parlay. Not everything should be met with immediate force.
The creatures randomly generated have to come from somewhere, but often don't have a logical point of origin; they simply wander around the encounter area, never interacting with the other creatures that live there. (Some random encounter systems do this better than others, admittedly.)
There are two problems with this criticism. First, this isn't a flaw of random encounters per se. A random encounter table needs to be tailored to the area it's for, and using any ol' table will give nonsensical results. Part of the point of the table is to make a place feel alive and dynamic, something more than just a static string of scripted events.
Second, rolling the encounter and just plopping the creatures into the game without inventing a reason (even if only in the GM's head) is certainly not worth doing, but that's an argument against a lack of rationale, not the randomness. It's the GM's job to figure out what the random results mean and weave them into the game. Coming up with an explanation for the result might inspire something wonderful. I've run entire sessions that were completely randomly generated. Some systems like Burning Wheel make this interplay between random results and creative interpretations a core part of play.
Along similar lines, you can have the same encounter multiple times, which tends to stretch credibility after a while.
This objection I just don't understand. How about: Don't do that, then? If I had to guess, I think this objection comes from the belief that once a die is rolled, you're morally bound to the result as a GM. If you think about dice as random idea generators (as I do), this just isn't a problem. Two encounters with orcs in a row? Maybe the second is looking for the scouts that didn't return. Maybe the GM should just roll again for a different result that makes more sense.
Random encounters drastically change the XP:resources rate of an adventure. If the PCs have an easy time with the random encounters, they rack up extra XP and throw off the challenge of later fixed encounters; if they have a hard time, they expend too many resources, and, again, throw off the challenge of later, fixed encounters.
I don't have a problem with this. If the later challenges are so far off that the PCs will level from random encounters before then, the GM is engaging in a great deal of plot railroading. Scrap those encounters and recycle those notes for a different purpose. A campaign that is responsive to player actions just won't have this problem.
Running the PCs down in resources enough that the later encounters are a problem can also be seen as a result of railroading. If the PCs simply cannot pause to rest, roll fewer random encounters. Remember that they're only meant to be supplemental. This is also less of a problem with groups that haven't been trained to tackle every encounter without any forethought. Heroes that don't pace themselves by picking and choosing their fights, and who charge into combat with the Big Bad when their resources are fatally depleted are making poor strategic choices that will soon lead to their death. I say let 'em learn that lesson if they need to. Smart players know when to sound the retreat.
Unless you either have a lot of miniatures ready to go, or don't really care, then you may find yourself discarding random encounter results because you don't have a way of representing them on the table.
I can understand this being a problem with groups that really enjoy the miniature sub-game and hobby of D&D. There is something exceedingly pleasant about seeing a battle played out with well-painted, properly-representational miniatures on a professional-looking game mat or scratch-built terrain. That's part of why I play miniature wargames. I don't look to my roleplaying games to satisfy my model lust, though.
For the majority of roleplayers, pennies, spare dice, paper chits, or just any old miniature works great for this. A game of imagination works surprisingly well with whatever comes to hand to track the position of PCs and NPCs. Other gamers choose to forego the gaming mat entirely and just don't have this problem.
In conclusion, it's really a function of your group's play mode whether random encounters are a problem or a boon. There are many players that enjoy a well-scripted plot that they can pace a character through, and I would have to condemn theatre as well if I said that way of roleplaying had no value. I do think that random encounters, as a tool in the GM's kit, have a lot to offer:
- A well-crafted table sets the tone of an area of play. The table for a bustling merchant city and a howling wilderness are going to be very different. Sandbox D&D games rely on encounter tables to theme different areas.
- Surprising results can spur the imagination in unpredictable directions. The necessity of making sense of a table result can be the mother of inventive play.
- Random encounters flesh out an area and give the players a sense of a living, breathing ecosystem, society, or operation.
- Random encounters keep players in the game instead of the meta-game. They can't just trudge forward with the blind confidence that the GM will only provide level-appropriate encounters that are tailor-made to be defeated with the "right" amount of effort and expended resources.
- Random encounters make it more difficult for the GM to railroad the plot or micro-manage PC advancement.
Whether those features are good or bad is up to each group, but I hope I've at least shown that there are redeeming features if one is willing to look. Wiker's preferred play style might not work well with random encounters, but a lot of games benefit from them.