By Nelson Schneider - 02/16/14 at 03:15 PM CT
I first encountered the concept of ‘random number generation’ in RPGs shortly after my initial exposure to THE original RPG. Specifically, in the course of running weekly AD&D 2nd Edition games during high school, I came to embrace randomness in every way. I created elaborate tables containing every race and class, then mandated that players at my table create characters not by CHOOSING these fundamental aspects, but by ROLLING for them. My logic, at the time, was to prevent players from constantly creating characters that were cookie cutter carbon copies of each other. I knew that one of my players would ALWAYS be a dwarf fighter, given the chance, and I knew that my own preference would lead me to an eternal cycle of playing as the same wizard with the same complement of spells over and over again.
My introduction to Random Loot came as part of the same game. When I bought a complete four-volume set of the “Encyclopedia Magica” (which I still own), the massive treasure tables at the back of the fourth volume became the sole source of loot provided to players at my table. Kill a dire rat? Roll on the treasure table! Kill a dragon? Roll on the treasure table! It was 2nd Edition, of course, and my friends and I were young and inexperienced, thus ‘balance’ was never really taken into consideration. Such overwhelming randomness led to occasional disappointing potions (of course, potions ALWAYS suck), but all too frequently lead to the disbursement of wands of wonder and an Invulnerable Coat of Arnd.
While we loved it when our characters benefitted from a particularly good roll on the Universal Treasure Table, whichever one of us was the Dungeon Master at the time learned to dread the ‘broken’ and overpowered characters that had spontaneously arisen. It is nearly impossible to create a fun-but-challenging encounter when characters possess overwhelmingly powerful loot. Likewise, it is nearly impossible to create a challenging-but-fun encounter when characters are kept artificially poor and underpowered by unlucky rolls.
By the time I first played “Diablo” (the PS1 port), I had largely grown out of my obsession with randomness. I had started college, and my D&D group had dispersed. My impression of the original ‘random loot’ videogame was that it was a boring, repetitive grind through samey-looking dungeons filled with lots of really terrible loot that was heavy enough to warrant multiple trips back to town to pawn it (No portable holes or bags of holding? Blasphemy!). I never made it past the second level, as I got bored with “Diablo” and, it being the PS1 era, I had so many amazing RPGs to play that grinding away at “Treasure Table: The Game” offered nothing compelling.
Of course, “Diablo” wasn’t really ‘new’ in its addition of randomness to videogames. Computerized RPGs had been the home of random encounters since the beginning (based, of course, on the fact that tabletop D&D had random encounters). “Rogue” and its vile Roguelike spawn had been randomizing everything they could – from encounters to floorplans to loot – since at least 1980. What “Diablo” – and to a much greater extent, “Diablo II” – did was repackage randomness as a pleasure-center-tickling IV drip, jammed directly into the part of the human brain that gets excited about gambling, encouraging players that ‘just one more try’ will result in a big win.
Fast forward to the 7th and 8th Generations, and it seems that everything has gone horribly wrong. All of the great PS1 RPG franchises are dead, and “Diablo” has spread its awful fixation on randomness into almost every genre. Instead of using the increased power of modern hardware to craft unique experiences of a scale impossible to create on older hardware, game developers are using the increased power to offload the most important portion of their jobs to computers.
Taken to its most extreme, the ‘procedural generation’ of digital content has seen a programmer create software that writes videogames from scratch. Is the Games by Angelina project the future of game design?
Why is procedural generation so popular? Back in my treasure table days, I liked it because it was a quick and easy way to generate rewards without putting in any real effort. Of course, this kind of shortcutting turned around to bite me when it was my turn to be the Dungeon Master, as I ended up having to put even more effort into designing everything else. From a developer perspective, randomness is a crutch to prop-up a short game that would otherwise be lacking in content. From a player perspective, randomness creates a lot of replay value. Of course, from either perspective, the procedurally generated material is of sub-par quality, providing ‘more content’ and ‘more gameplay’ in the form of tedious grinding and superfluous padding, with any unbalanced portions handwaved away with admonishments to grind some more. Yet, because the rewards are spread-out in such an uneven trickle, much like B.F. Skinner’s experiments in Operant Conditioning, players will keep performing the same repetitive tasks in the vain hope that the random number generator will produce something good ‘this time.’
The result of overreliance on procedurally generated content is games that waver erratically between ‘too difficult’ and ‘too easy,’ dependent entirely upon the quality of the random rolls in any given session. Instead of deliberately-crafted experiences that have been painstakingly balanced by a human designer for optimal enjoyment, we are being deluged with random experiences that are fun when the numbers are favorable and intolerable when they aren’t. We have already seen a number of potentially flawless games reduced to merely ‘meh’ experiences by stingily-low random numbers, from “Dungeon Defenders” to the ‘Borderlands’ franchise.
Personally, I much prefer to play games that are generated by humans instead of mathematical processes. In a human generated game, if a player is expected to have a weapon of a certain power-level, that weapon will be placed in a specific location and will be there every time, whereas in a procedurally generated game, such a weapon may or may not appear and the location for this appearance/non-appearance may or may not be reachable by a character in need of said weapon. Yes, static locations and hand-crafted environments will always remain the same through multiple playthroughs of the same game, which limits replay value. But if you are really that bothered by having a game’s layout memorized after replaying it 15 times, it might be time to play a DIFFERENT game instead of lamenting the unchanging nature of one from which you have already sucked as much enjoyment as possible.
Sadly, the one area of gameplay design in which procedural generation would be a boon – multi-player deathmatches – is still stuck with a single-player design mentality. While single players and co-op players lament the lack of replay value in human generated games, deathmatch players, be they of the FPS or MOBA variety, seem content to memorize and replay the same maps over and over and over, drastically reducing the chances of skilled newcomers from claiming victory over less-skilled veterans who have committed every bump-mapped polygon in the environment to memory. Is this mismatch of design ideals due to the perceived higher-quality gameplay provided by human opponents instead of AI opponents? It seems to me that providing random arenas when every participant is human would be a natural extension of the superior competition provided by humans over AIs, as no AI could compensate for the quirks in randomly-generated terrain the way a human could (Hell, there are plenty of human generated games where the AI gets ‘stuck’ on geographical features in environments that never change!).
Nobody knows exactly what the future of game design ideology holds. Will we face a future filled with crap designed by Angelina and her descendents? Will a hipster mentality of ‘craft’ games hit the already-nearly-hipster Indie games movement, much like craft beers, to provide a painstakingly hand-made experience to contrast with the formulaic experience provided by competitors? Either way, I want to see gambling theory stripped from videogame design once and for all. There has to be a better way squeeze more enjoyment out of a videogame experience than to pander to a basic animal quirk in human brain chemistry.
Comments
Chris Kavan - wrote on 02/17/14 at 08:39 PM CT
I don't mind the random nature of things but I think there has to be a balance. Games like Diablo and Borderlands would benefit if you were guaranteed a a a rare drop in specific locations or by specific bosses - not necessarily the same thing, but at least you know you will be rewarded for your efforts and don't have to kill the last enemy of the game only to be showered with under-leveled crap that everything else drops. Keep things random, but keep things fair - and don't rely on chance for everything.
