EPT"While sex heads a great number of lists, we all have things we like to do in between."
In the past few days, two posters brought up the subject of competitive random play. One was little more than a rant thread, and so has naturally gotten much more attention than the second, which attempted to discuss the flaws of random in a competitive setting from a reasoned standpoint.
Pick one: either the rant or the argument. Either one essentially comes down to this: facing a random - and assuming for the purpose of the argument that the player does not reveal his race - throws a further element of uncertainty into the metagame, and (this is important) that element damages the game as a competitive activity.
"Gerard would have chosen that moment to attack ... Benedict... would have had one [eye] in each pocket by now..."
The standard metagame depends on being able to choose an ideal path in the face of your opponent. Whatever the matchup, the traditional analysis has broken openings into three categories:
Rush play, also known as "cheese", has two main varieties: all-in or pressure. An all-in intends to, in the (out of context) words of Day[9], "Just go fucking kill him!" Anything short of a total win is normally a total loss - infrastructure and defensive potential are sacrificed to the attack. A pressure rush is designed to slow the opponent down in order to guarantee a win later. Particularly poor defense may enable an easy win, but the rusher still has plans for a longer game - though he will be at a disadvantage if the pressure is rebuffed too easily.
Standard play, sometimes called "safe", is generally considered to be the normal state of the game. Two players playing nicely standard will generally give us a decent macro game. Different builds come in aggressive or passive varieties, but in any case early pressure is limited or nonexistent.
Economic play, or "greedy" builds, are designed to secure a big infrastructure lead early. Normally these indicate a player wants to play a macro game and steamroll the opponent with weight of metal whenever his advantage builds up to the breaking point. Sometimes, though, an early economic build is designed to turn out an unstoppable timing attack.
The normal consideration is that rush > economic > standard > rush, though particularly vicious all-ins can disrupt even safe builds, or safe standard builds can hold off economic timings long enough to equalize - or better players can simply disregard these RPS effects and beat worse ones.
In a normal game, where the opponent has picked his or her race, a catalog of probable builds and responses pops up in the good player's mind. Scouting then reveals actual plans, and responses can be tailored effectively to whatever the other guy does (short of unscouted cheese - and even that usually is known to be coming in some form).
"But they are genuine hero types. Me, I just stood there..."
The random player himself faces this planning, with exactly the same choices, and an additional advantage for the beginning of the game. His opponent is forced to choose a generally effective build, which may not counter the actual race of the random player as well as a specific build would. To take the most blatant Brood War example, a Protoss facing a random player has to seriously weigh whether and how to expand. Not expanding against a Zerg could lose the game; expanding early against unscouted Protoss pressure might also lose the game.
There are several options: play against randoms could generally be a sign for pressure builds. (Compare Brood War ZvZ, which generally sees only two combat units used, and even bringing in queens is extremely rare.) The goal would be to force random players to play passive standard builds, gaining no advantage from their time hiding.
Or, the player with a chosen race could opt to play safely himself, trusting in a solid defense and superior knowledge of the game to rebuff any cheese and regain any advantages lost in the opening.
Because, after all, playing random has weaknesses the player with a chosen race would not have. A random player has to practice nine matchups, not three; to face a specific opponent, he has to practice three matchups in three different races while his opponent practices the same three matchups, but always controlling the same units.
"'I get the idea. How have you been operating?'"
What kind of player would be good at random play? The player has to be adaptable. Hydralisks one game and dragoons (or stalkers) the next are enough to make anyone lose balance. They are probably going to play best with low-tier, fairly similar units: spells and effects tend to diverge more in characterization the further up the tech tree you climb. And they are probably either planners or whimsical: that is to say, already inclined to look for the unexpected.
I am going to outline three well-known Brood War players and why they might have succeeded reasonably well playing random.
Kwanro was the first player who came to mind of modern-era players. He coupled devastating ling micro with a certain limited approach to strategy, but he also got at least some credit - despite his brief time with the team - for a marked improvement in ZerO and Neo.G_Soulkey's ZvZ play. But the main reason I think he would have been at home playing random is that he always seemed to fall apart once he hit Hive. I could be exactly wrong, and this is a sign he's too limited to play any way but the way he did; but I suspect it shows aggressiveness well-suited to taking advantage of the information imbalance provided by random play.
fantasy is an example of a thinker. His preparation is among the best and his coach is unrivaled. His prepared builds are all but unstoppable (though he tilts hard when they do come apart). If you handed fantasy random play, his tournament runs would be truly amazing - though he would likely have even greater struggles in team league play (or in weekend events like MLG) than he already does.
Stork is my Protoss pick to try playing random, because he already does weird stuff and already wins an absurd amount of games while self-admittedly not really practicing. Given more ways to do goofy things, and with his pristine micro, random-Stork would be pretty scary. (It's worth noting that Protoss in general seemed to have fewer candidates for the adaptive type of play randoming would demand, so I was forced to go with one of the best.)"Individually, they were no match for me. Even a couple of them together had not made that great a showing."
But the fact is that playing random is hard. Most random players are not particularly successful; the most notable in Starcraft II was
GuMiho - and he's been playing Terran now for months despite fairly good results with random play. There has not been a successful Brood War random player at any time in the "modern" era (ushered in by oov's macro concepts) of play in Korea, and I don't know of one before then. Even Garimto, I believe, was really a race-picker. Some 2v2 players would random on some of the worse maps, but that is a different question.So far, my defense has been fairly simple: Complaining about random players is hard, because they are doing worse than you. True, some players will random and cheese up the ladder - how is this different then the people who 4-gate or 6-pool their way there? They're a nuisance, but not the real threat. The ones you want to worry about are the ones who are actually trying to play well, not just stroking their egos.
But take this a step further. Not only should you not worry about facing randoms - there should be no stigma about it. Competitively, it's clearly the hardest choice - most give up. Maybe on the ladder it gets you an advantage - use it. The point is to win.
All quotations from Random, in Roger Zelazny's Sign of the Unicorn
