Experienced Points: Why Make Games Random?

By Shamus
on Feb 16, 2016
Filed under:
Column

My column this week is a reader question where my answer goes immediately off-topic, devolves into a list, runs long, gets sidetracked, and then ends on a cliffhanger.

No need to thank me. Just doing my job.

Enjoyed this post? Please share!



20206Feeling chatty? There are 46 comments.

From the Archives:

  1. Alex says:

    I think the question being asked was why games have a visible and modifiable luck stat for the player. Other stats tend to have a specific impact on visible numbers, but luck modifies random chance and usually requires in depth knowledge of the system to know what it does.

    • Erik says:

      Easy, games have a random factor, which is kinda like gambling (people like gambling).
      Having a high luck stat makes them feel like they increased their odds at “winning”

    • MichaelGC says:

      Right – without a luck stat Josh might not know exactly how far beyond 100% his critical chance has got to.

      • James says:

        Which makes Fallout 4 an odd duck, it has luck but no natural or random criticals at all, its all sneak attack (guaranteed) or earned and used in vats (which makes the hit a guaranteed hit) this makes me more upset then you can imagine.

        • Humanoid says:

          It’s actually one of the few things in Fallout 4 which I can wholeheartedly endorse. Given the level of detail we can simulate with today’s games, there’s no reason left to justify criticals being purely random. In an age without hitboxes, dynamic line-of-sight calculations and relatively smart AI, you no longer need to abstract the extra damage as pure luck-based rolls. What were random criticals in the past now make more sense as extra damage for headshots, sneak attacks, that sort of thing.

          • Geebs says:

            I absolutely agree, particularly in a game which is mostly about guns. Having a difference between a rocket-propelled grenade that the player can clearly see hitting directly between the eyes and a critical RPG between the eyes is intuitively ridiculous.

        • Scourge says:

          There is a mod for that. (Figures).

          Now you can randomly critical in normal combat too.

    • Shamus says:

      Right. That’s how I read it. But like I said right after the quoted question, I felt like I needed to ask another question first.

    • Wide And Nerdy ™ says:

      Putting the Luck stat in is somewhat comparable to Shamus’s example of the two swords that have the same average damage but different ranges of probability.

      You can pump the Luck stat to, say, have a narrower probability range that leans towards the upper end of the former probability range, or you can boost the numbers that are added to that roll. So you either favor higher variability with the potential for higher roll totals or lower variability with a lower ceiling and higher floor but more reliability. Or something like that.

      Depends on the game as to whether it works.

      It can also serve a flavor purpose. We have lots of ideas about Fate, Luck, and Fortune. By giving the player a stat to play with, you signal (in theory) that these themes are an important feature of your world and/or story.

      And it creates different potential characters. You can crank your luck high and be an the hapless schlub who manage to stumble into danger and come out the other side better than he deserves (like maybe an Indiana Jones type, but with more luck and less quick thinking).

      Or he can be the guy for whom everything goes against him but he grits his teeth and makes it through with determination (like Peter Parker or at least how Peter sees himself).

      • Sleeping Dragon says:

        My problem with Luck as a stat is that oftentime in games it’s rather… nebulous. Some titles have the courtesy of showing the player the specifics of how luck affects rolls (say: “each point of luck improves critical chance by 1%”) but frequently it’s generic statements like “improves your odds of critical attack” or, my favourite, “subtly affects everything that happens”. It irks me because if I have to make a choice where to assign points I’d like to know what choice I’m actually making.

    • Cinebeast says:

      I feel like a Luck stat would be a decent in-universe way to control the game’s difficulty. Maybe you’re recommended a Luck of 5/10 for a first playthrough, but experienced players take those points and spend them elsewhere, punching up the difficulty and giving themselves more options, while a player who wants an easier time will do the opposite.

      • Chris says:

        Ooh, I like that idea. I see something like that fitting in better with the classic Fallout titles than the Fallblivion stuff, but it’s a solid idea.

      • Nidokoenig says:

        Final Fantasy Explorers has something like this in the class system. Each class has a set Luck stat that affects item drops, and as a general rule, the deadlier you are, the lower your Luck. For example, even as classes that can use all use Knuckle weapons and abilities, the Ninja and Monk are quite a bit more effective than the Machinist, who has the second highest Luck stat and basically depends on you being good at using Counter, whereas it’s only highly recommended for the other two, as their stats and innate abilities are better.

        I’m not sure how well a stat-based difficulty option holds up under stat customisation, though, since most people who are replaying will adapt against random misfortune and in favour of good drops or interesting random events. Or critstacking, like me.

      • Cuthalion says:

        Oooo, that’s very interesting. I like that and will file the idea away. :P

  2. Kylroy says:

    Like that you linked a picture of a perfectly rolled Diablo 3 item. That game’s success is owed almost entirely to a masterful use of randomness.

  3. Decus says:

    Replayability is a big element, not so much for a Luck stat but for random numbers and randomization in general. Going to guess that’s the “Big One” you were thinking of, given the popularity of things like Randomizers for older games.

    After what I guess has been nearly 20 years of Pokemon Red/Blue I need to use a Randomizer or things like Nuzlocke rules (randomizes your roster somewhat) to keep them interesting. After over 10 years of Fire Emblem (GBA) I need to use things like drafts (randomizes your roster somewhat) or Randomizers to keep it interesting. Both games already have random elements/damage mechanics inside of them, but in the big picture even more random is needed after enough playthroughs to keep me on my toes.

    • Falterfire says:

      I’d argue randomness doesn’t provide much replay value in games that aren’t specifically procedurally generated. Pokemon has replay value due to the number of different pokemon allowing different ways through the game, but if you want to build a specific team and use it to beat the game, you’ll be able to do it in every single playthrough if you want. Outside of a few very extreme examples (Feebas maybe) Pokemon’s randomness prevents predictability but doesn’t itself give you a reason to play the game again.

      Likewise in Fire Emblem, how exactly a level unfolds may change, but the challenges you face will be the same – Every (story) level will be laid out in the same manner and the AI will move in the same ways and you’ll get new units in the same order every time. A few shots hit or missed differently don’t heavily affect replayability.

      Where randomness DOES affect replayability, in my mind, is in cases where you face challenges which differ based on random chance. Pretty much anything where you’re expected to make multiple playthrough (So, roguelikes) where randomness affects not only the outcome of a given move but also your ability to use a specific strategy at all.

      As a specific example, I’d point out FTL – You may be able to use a specific combination of weapons and parts one run that you can’t in another, which means that although some learned skills are transferrable between runs, you can’t just have a plan like “First, always get two Burst Laser IIs. Then get the shield upgrade. Then get the second door upgrade…” and so on without having to adjust it to match what you actually have to work with.

      • Rodyle says:

        Actually, I think the PokeMon randomizers he’s talking about are the ones with actually are able to randomize not just PokeMon spawning locations, but also their elements and moves.

        • Falterfire says:

          Yes, that’s what he means by a Randomizer in the sentence about randomizers or Nuzlocke runs. I was talking more about the bit at the end where he mentions random elements in the games themselves, which I don’t think contribute significantly to replayability.

  4. Neil D says:

    I was thinking about randomness in games recently, though on a somewhat different scale.

    My first computer was the TRS-80, way back around 1979. One of the games I remember playing a lot of was “Swords and Sorcery II”, a game I typed in by hand out of the old 80 Microcomputer magazine. It was predominately text with a few small and simple graphic moments, and I don’t know how many hours I must have logged on it.

    A few years ago in a burst of nostalgia I bought a TRS-80 just like my old one off of eBay, and a bunch of the old magazines as well. I fired it up, and spent some time re-entering that old game, practically giddy to relive the old memories.

    About three minutes into the game I suddenly realized that the whole thing was nothing but a series of random encounters, which would prompt you to make a random choice (Choose path 1) or 2); F)ight or R)un; C)limb out of the hole or Y)ell for help) that would result in a random outcome. Survive enough of those, and you win.

    Even the final fight with the dragon at the end came down to luck. You had to move a stick figure past an randomly occurring line of ‘flame’ from the dragon’s mouth – the course of the line never changed, and you could only move between ‘breaths’. If you were in the line of fire when the dragon started breathing, you were toast.

    Essentially, you were entirely superfluous to the process. You could program the game to make your random decisions for you, and come back later to see if you’d ‘won’.

    Looking back on a lot of the other games at the time, barring pure text adventures and shoot ’em ups, a great many of them amounted to much the same thing. I’m sure a lot of it is just down to the limitations of the hardware, and the fact that the home video game market wasn’t even out of diapers yet, but still, it was kind of an odd realization.

  5. silver Harloe says:

    …because CRPGs are still cribbing from RPGs which have dice. RPGs have dice because Wargames have dice. Wargames have dice because the simulation is imperfect. In reality, a fight can be lost because someone tripped on a loose rock leaving an opening in your formation – in other words, Wargames have dice to *increase* realism. An army with better equipment, better training, high morale, and fewer issues with hunger and illness and fatigue should win, but historically sometimes such armies lost, anyway (or, more often, they had better in some categories and worse in others so it wasn’t really predictable who would win, anyway).
    Just some starter thoughts I threw together.

  6. Alrenous says:

    So many advantages it’s almost a better question to ask why -not- make a thing random.

    Dark Souls (yes, again, sorry) damage is fully deterministic. Further, the basic damage calculation is very basic, it’s damage – defense/2. However, there’s a whole bunch of situational modifiers, which may be why it doesn’t seem boring and static. Even if I’ve got a build one-shotting everything and the exact damage is irrelevant, it’s still good. Perhaps that’s because I know that if I don’t one-shot them, their counter-attack is going to be seriously terrifying.

    Anyway, it clearly doesn’t need the random, and since I’m not entirely sure why, I find it interesting.

    • DivFord says:

      Dark Souls does have randomness though. It’s in the enemy attack selection, and since correctly identifying enemy attacks and reacting to them is such a huge part of the combat system, I’d say that adds a lot of the depth Shamus is talking about.

      In other words, I don’t think you can say that it “doesn’t need the random”, since without any random element, the main challenge of the game would be gone.

      • galacticplumber says:

        Enemy attack choice ISN’T random in Dark Souls. It’s based on proximity and location relative to player. The only legit random thing inherent to Dark Souls is drop rates, and it’s not really core.

        Outside of that you have the multiplayer which is entirely optional.

        • Decus says:

          It is random though–it’s just random based on proximity and location. I can get enemies to show me certain attacks by standing at a certain distance relative to some direction of their body, but not with 100% probability because while that does limit their choices it never limits their choices down to 1 single move.

          • galacticplumber says:

            This has not been my experience. Were you moving or standing still during your test? Yes that variable can matter a lot because enemies can’t stop their animations which would possibly result in attack choices triggered when you’d have move out of the effective range.

  7. ehlijen says:

    Speaking of randomness in games, what’s everyone’s take on XCOM 2’s new approach?

    I’m talking about the random engineering projects, psion training and the mindspin power.

    In case you don’t know:
    There are now 5 random engineering projects (experimental ammo/armour/grenades/heavy weapons/adv. heavy weapons). Every time you complete it, you get a random item of the chosen type for your soldiers. If you like that item and want more, you have to try the project again and hope you roll lucky.
    Disclaimer: the items are always somewhat similar and none are truly important to an effective strategy.

    Psion training works by having the soldier in the psi lab for x days. He gets a choice to learn from a random three or four powers he doesn’t know yet from his talent tree. Given enough time, any psion will learn all the powers, but the game isn’t built to guarantee you have that time. So most likely you’ll have a psion or two who know random bits of their talent tree in the end.

    The mindspin power targets an enemy and then does a random thing to their mind: nothing, disorient (move and aim penalties, some actions can’t be taken), stun or mindcontrol. Sectiods also have that power and are thus fairly harmless 90% of the time unless they happen to roll mindcontrol on their attack.

    My opinion? Not good.
    Don’t get me wrong, I’m happy with randomness. Procedural maps, random missions, % chance to hit or hack etc all good.
    But there’s a right context for randomness and a wrong one. Surely once an engineer has built a thing, he could build that specific thing again without rolling dice?
    And I think I might be more kindly disposed towards psions if most of their powers weren’t available to the other classes as talents or gear already, if they actually had something unique.

    Some things should be random, others shouldn’t be.

    • Content Consumer says:

      Re engineers:
      I suppose it depends on whether or not you’re looking for realism (or at least verisimilitude) in your games. The standard argument of “if you can accept dragons, why can’t you accept swords that weigh n units instead of m units” is usually trotted out at this point, but can be safely ignored. :)
      Yeah, it doesn’t make much sense for someone to get a random result, as if he sets out to make ammo, closes his eyes, and puts his hand into a lucky dip bag to see whether he gets viper or dragon components. An engineer doesn’t set out to make a computer and only after it’s finished find out whether he made a desktop, laptop, iPod, or tablet.
      It all comes down, as most things do, to whether or not you can stand the justification for the mechanic. I don’t like the randomness (if I want tracer ammo, I want tracer ammo, not ten bits of other kinds of ammo before the RNG is kind to me), but I can deal with it. As you said, none of them are truly important to an effective strategy. So I just treat it as the aforementioned lucky dip bag. Tracer rounds go to snipers, other ammo types go to whoever has an empty slot and nothing much to put in it.
      Grenades are a slightly different story – even with the randomness, my grenadiers can almost always make good use of any sort of grenade. Frag or plasma in the 2x slot, and there’s usually some tactical use for any other sort in any given mission.

      Re psion training:
      I agree. To me, the psion class seemed tacked on, as if late into the development process someone said “oh crap, we forgot psionics, it’s an integral part of XCOM” and they rushed to put something out there. Each psionic power has its use, and a fully-trained psion can actually take the place of several different equipment types (Null Lance for Plasma Blaster, Fuse for grenades, Mind Control for Combat Hacking [organic vs robot], etc). So the psi operative is basically a catch-all soldier that can fill most any minor role, with a couple abilities that are still useful in the late game. But it still seems arbitrarily added just because someone felt the need to add psionic soldiers to the game.

    • IFS says:

      I actually like the random projects, and so far seem to have plenty of time remaining to train my psions (staffing an engineer in the facility ups the speed of training immensely). The random projects kind of force you to experiment with different outcomes, and leave some replayability in that you might not find them all in one playthrough. Psion powers are all very powerful (its saying something that the weakest powers is probably detonate an enemy grenade) and I’m not really sure what you mean by ‘available to other classes as gear’. I guess you can throw grenades to replicate that one and you might see flamethrower as a replication of null lance, but so far I don’t see any way to mind control an alien, give another soldier an extra move, or put an enemy in stasis for a turn.

      I don’t like the variance of mindspin though, the level of difference between panic and mind control is too steep for my tastes though there is counterplay both in the existence of mind shields and in the ability to just shoot the sectoid so that’s not all bad. Plus I much prefer panic as the result of alien mind powers as opposed to ‘oh no I was shot at, better shoot my friend’.

      • ehlijen says:

        I’m not opposed to the engineer saying ‘hey I could tinker with this stuff and maybe come up with something cool’, resulting in a random thing. My verisimilitude problem is when that engineer then can’t remember how he did that and I don’t get to make more of the thing I’ve already got. Once you’ve gotten a random thing, you should be able to make more of that specific thing. Unless the engineers really are doing all this while high and don’t take notes.

        As for psions doing what other classes and gear do:
        Mind control: hacking. Works on different kinds of enemies, but is in principle the same ability, only question is if you’d rather want a muton or a sectopod on your side.
        Null lance: plasma blaster (aka one of the least useful heavy weapons)
        Vortex: any explosive, capacitator discharge
        That power that does a little guaranteed damage: Combat protocol
        Immunity to environment: hazmat vest
        Immunity to mind messing: psi shield

        While psions are very useable, I don’t think they are unique enough to really be an interesting full class. Between needing a psi lab moderately early to get a power variety by the endgame, getting a free psyker in the final mission anyway and the class doing so very little you won’t have other soldiers for already, I just don’t see it worth the engineering time to build and staff the lab, or to upgrade the psiamps on top of the other gear.

        • IFS says:

          I’d say you’re underselling a lot of the psionic powers, the vortex can be upgraded to chain into insanity, mind control has a much wider choice of targets than hacking (and is also permanent unlike hacking), and the immune to mind effects is an aura which means you don’t need to bring mind shields on other soldiers. Stasis is also an incredibly useful ability you didn’t bring up a neutral counterpart to (don’t want to deal with that sectopod? Just put it off a turn. Need to get closer to skulljack a codex? lock it in place, etc). I’ll concede the plasma lance though, having just unlocked that last night it is very much a neutral null lance.

          Personally I view the psion as Xcom’s version of a wizard, they’re incredibly versatile and powerful. Yes some of their abilities have neutral counterparts but those counterparts are often worse or limited use and a fully leveled psion makes for incredible support in any fight. Plus they make for an easy option to level recruits you don’t want to throw into the field late in the game which is nice.

          Now I am only playing on veteran (and haven’t finished the game yet), so its possible my opinion might change for higher difficulties but I definitely feel like the psi operative is an interesting enough class to deserve to exist and require its own gear upgrades.

          As for engineering I definitely agree that you should be able to rebuild items you already discovered. It seems like a bit of an oversight that it’s not an option to do so.

          Edit: oh hey, so that’s what the avatar I get with my current email address, I’d almost forgotten I was using my old one on this site…

    • crossbrainedfool says:

      As far as I know – mindspin/insanity has a success chance (like straight MC does) which is influenced by your Psi stat and your opponents will stat and if it succeeds it always does something.

      As for the Experimental gear – not allowing you exactly what you want is the point. It’s another way to shake up your toolbox, and make you rethink things. Sure, tracers would be great, but you got Venom instead. Okay, how do you use that? It’s another way to put you into a jam for you to think your way out of.

      It’s worth noting that all the experimental gear is optional utility stuff – weird grenades (you already start with Flash, smoke, and frag), ammo types, 1xmission heavy weapons, and so on.

      I think it’s intentional that Bluescreen/EMP isn’t gated like this, which allows you to kit out to kill the late game high armor enemies.

    • Falterfire says:

      I like XCOM 2’s randomness, but only because I don’t think any of it is actually make-or-break to a campaign. You can win whether you get Acid Grenades or Inferno Grenades or Gas Grenades, and unless you get incredibly unlucky on a Psi-Trooper you’ll probably end up with a soldier who is useful in the field after you get three or four abilities.

      I would feel differently if I thought the game was significantly harder if you didn’t have Dragon Rounds or if having a Hazmat Vest was necessary to beat certain enemies.

    • Daemian Lucifer says:

      Procedural maps, random missions, % chance to hit or hack etc all good.

      Ive come to hate the % chance to hit or hack,very much.Because of its stupid binary nature.So I can kill this alien(and only THIS ONE ALIEN,not the one directly behind it) outright,or I can miss it and it will kill me on the next turn.That ridiculous.Especially in how every guy has EXACTLY one action,no more,no less*.I much,much prefer the old way,where you pick the trajectory for the bullets,and then they travel somewhere around that trajectory,and even if they miss your intended target,they still can hit another target.The old way was much fairer and better balanced.

      But randomness in building stuff and training?Eh,thats ok.You can always have a do over if you dont like it.You dont lose if you get a less desirable result.

      *Ok,there are powers later that influence this,but thats not important.

      • Humanoid says:

        I don’t mind random chance to hit, but it should be able to be mitigated through CLOSE RANGE far better than it is. Point blank shots with rookies were 100% in the previous game, and having it be capped out at 85% (with normal rifles) now seems a pointless regression.

        Any argument that guns at that range should be unreliable is undermined by grenades being 100% reliable at any range, the game isn’t interesting in that level of simulationism.

      • ehlijen says:

        It’s certainly a more abstract game than the original UFO was, which has it’s ups and downs.

        Downs are significantly less emergent gameplay (XCOM:Apocalypse was amazing for that) and much simpler gameplay.

        Ups are greater approachability (seriously, the original was infamous for its impenetrability to new players) and faster gameplay (which allows for more missions to be played given the same amount of player patience/endurance).

        There are a lot of things I miss: being able to give any soldier any weapon, different attack types (aim, snap, spray), base building that doesn’t force satellite/radio spam…

        But I still enjoy the new games a lot (even if the engineering crew apparently does do peyote now and then while building grenades).

        • Ranneko says:

          Individual soldiers are much more interesting in XCOM compared to X-COM or Xenonauts. Having specific classes and abilities makes them feel much less interchangeable and impaction in a way that remembering Squaddie X has a high range does not.

          • Humanoid says:

            That and because they’re no longer expendable commodities. :P

            Sure, you have to deal with losing soldiers in the newer games too, but there’s a massive difference between losing soldiers occasionally when you stuff up, versus losing them all the time as part and parcel of playing well. Sure, in order to accomplish/balance that they had to massively dial down the squad sizes, but I’m fully on-board with that change.

            And on the subject of this article, I also love the Absolutely Critical option in EU. Criticals as a reward for good play, as opposed to reward for …being lucky. Indeed if I went back and played the game again I’d outright remove base critical chance. And on another RNG-related note, I’d mod all air combat hit chances to be 100%, but reduce the damage per shot proportionally. All of this should be easily changeable with Long War installed, since they’d just be entries on the one text file.

    • Humanoid says:

      Yeah, I liked the idea of Blind Research in Alpha Centauri, but really, once you’ve learned it, you’ve learned it. I’d be happier if “Experimental X” was a research that unlocked a random doodad of class X, which then you could build infinitely. XCOM’s research tree is shallow enough as it is, so adding new ones is not a problem instead of having it as an engineering item.

      Besides, Experimental Warfare was already a research item in Enemy Unknown.

    • Galad says:

      Wait until you see the final battle in XCOM 2. I am NOT happy about it, and the reason is a slew of RNG that does not appear anywhere else in the game, not to this extent anyway.

      • ehlijen says:

        I played the final mission. I didn’t notice much more randomness in the final mission than otherwise. What do you mean?

        Then again, because my first experimental ammo and grenade attempts both produced poison specimens, it never clicked with me that it was a random dip draw until I read about it afterwards. I didn’t want anymore poison stuff, so I never tried those projects again.

  8. Rodyle says:

    Are you going to talk about history in your next article? I mean, the randomness in videogames pretty much comes from the tradition of using dice rolls in TRPGS.

  9. Richard says:

    Oddly enough, one of my favorite JRPGs ever, Paper Mario on Gamecube, was fun because it actually removed many elements of randomness typically found in JRPGs. Your attacks would always do the same damage based on your weapon, level and the enemy’s defense, with criticals happening only if you press the attack button at the right time (or let go of the left stick, or whatever, depending on the attack). Defense was the same, with you able to reduce damage taken with a perfectly timed A-button press.

    Since the numbers were generally kept small and non-random, you could use a bit of quick and simple math to plan out an entire battle from the beginning, determining the fastest and/or safest way of defeating the enemies. It was entirely possible to find yourself in a situation where you could defeat the enemy with 3 perfectly timed hits, but missing one of your chances for a critical would cost you the fight or at least cost you a healing item that would be better saved for the upcoming boss battle (your inventory space is incredibly limited, only allowing you to carry about 20 items or so at a time, and items don’t stack.)

    Buffs and debuffs worked similarly, having a set number of turns that you or the enemy is affected. “I need to do 10HP of damage, but only have 5HP left. A perfectly timed defense will allow me to only lose 4HP instead of 6HP, and a perfectly timed attack will do 8HP of damage instead of 4HP. I could use this turn to attack, defend against the enemy’s attack, and then finish the battle with a final blow and 1HP remaining, but if I miss the defense or both criticals I’m dead. I could also use one of my very limited items to put the enemy to sleep for 3 turns, giving me enough time to safely dispatch the enemy even if I miss all my criticals.”

    Add on top of that different types of enemy defenses and enemy positioning that affect how you can attack (can’t jump on spiky or fiery enemies, your hammer can’t reach flying enemies or enemies further back in the row) and it ends up being a wonderfully deep battle system with very little, if any, random elements.

  10. silver Harloe says:

    and then there was Tower of The Sorcerer – combat as a space-traversal puzzle because of how predictable the results are.

  11. Randomness in games for me is all about what *kind* of a game it is. If combat is all about timing your attacks and blocks and dodges properly, having an additional roll to see if you hit is just *obnoxious*–but random damage isn’t so bad. But it’s not totally necessary either–see the boss fights from Prince of Persia: The Two Thrones. Those were 100% timing, no randomness at all. They were still pretty cool.

    Random gear was fun for me in Diablo 1 and 2 but I hated it in Diablo 3 because you did actually need to assemble sets and so forth for Diablo 3, so after a while you weren’t even LOOKING at random stuff, just waiting for the one set piece you needed to drop and hoping the stats on it were good. That kind of stuff makes me want to gouge my own eyes out. It’s the difference between “at any moment, I could see something cool!” and “How many more times do I have to run this crap before X drops?!”

    Random dungeons/locations tend to be pretty boring, although they can be MADE interesting by adding a second or even a third layer of randomness on top of them, like random bonus spawns or random champions or random super-bosses. Random dungeon, bleh. Random dungeon where you suddenly have 35 champs all in one room . . . okay NOW it’s interesting. The thing is that there has to be some kind of interaction between the gameplay and the environment. Rooms that are in a slightly different configuration, who cares. Rooms where the configuration can lead to craziness, that’s different.

  12. kdansky says:

    Talking about randomness as if all randomness were equal is doing game design a disservice. If you randomly set up all chess pieces, and then play a round of chess, you still play chess. No matter how much you shuffle, it’s still a strategy game.

    If you however use the normal starting locations, but every piece you take gets to roll a die and has a 1/6 chance of defeating the attacking piece instead, then you have a vastly different game. If you make this a 1/2 chance (a coin-flip), you have basically removed all strategy from the game, because having your piece taken is just as good a move as taking a piece. It would put an incredibly low skill ceiling on the game.

    That is why I think we should start to distinguish these two cases of randomness: The kind that comes before you make a decision and lets you play around it (drawing a card), and the kind that interferes with your decision after you’ve taken it (randomly missing a shot).

    Modern (AAA) gaming has a ton of the latter, and that’s a big detriment: Games seem deep, but that’s just our mind playing a trick on us because we cannot understand why we won. Not because there are deep rules, but because we literally cannot learn anything about it, as it was pure luck to begin with.

Leave a Reply

Comments are moderated and may not be posted immediately. Required fields are marked *

*
*

Thanks for joining the discussion. Be nice, don't post angry, and enjoy yourself. This is supposed to be fun.

You can enclose spoilers in <strike> tags like so:
<strike>Darth Vader is Luke's father!</strike>

You can make things italics like this:
Can you imagine having Darth Vader as your <i>father</i>?

You can make things bold like this:
I'm <b>very</b> glad Darth Vader isn't my father.

You can make links like this:
I'm reading about <a href="http://en.wikipedia.org/wiki/Darth_Vader">Darth Vader</a> on Wikipedia!

You can quote someone like this:
Darth Vader said <blockquote>Luke, I am your father.</blockquote>