Vague Patch Notes: Why do MMOs have randomness?

Order from disorder.

If there’s one concept that is nearly universally reviled among MMO players, it’s randomness. Yes, pretty much everywhere. You could make a little fairy that flies around randomly and dispenses helpful buffs at random, and people would hate it because what if the fairy doesn’t show up when you need a buff, or you get a lousy buff? No one likes things being random; the term is as close as you can get to being an accepted dirty word.

So why do MMOs still have so much randomness in them?

Well, because games have randomness in them. But that’s only part of the story; the fact is that randomness does two different things in MMOs (and games in general, to be fair), and it’s fun to pick apart why we still have so much random stuff when no one actually likes anything being random. And yes, if you’d like to point out that computer have a hard time generating truly random numbers, we’ve covered that before. Let’s move on.

So where should we start? Tic-Tac-Toe, or Noughts and Crosses, or whatever you call the game with a nine-square grid played by two players. You know, the game without any actual randomness.

If you and your opponent are playing optimally to win (which is pretty likely, since otherwise you’re either just losing for giggles or playing someone too young to know better), this game is always going to end the exact same way. It’s even possible to map out every possible move quite simply. The only reason that the game persists is because of people playing sub-optimally; put two optimized computer opponents against one another and they will play to a draw every time.

Now, let’s say that you’re designing a computer version to play against a human player. If both the computer and the human play optimally, the outcome is going to be the same every time. So you add some randomness. For example, every other move the computer makes is optimized, but the computer’s first move is always random. That alone gives the game longevity because even if you play optimally from that point on, the computer has only one chance in eight of making the optimal move right away – and you might screw up, because it’s an unfamiliar situation!


That’s one of the big goals of randomness right there. Randomness prompts variance. Why are critical hits in combat? Because sometimes that means you can deal a lot of damage sooner than expected… or can take a lot of damage unexpectedly. It makes optimal play a lot harder to hit, even when a lot of other elements of play are completely under your control.

You might immediately be pointing out that “variance” doesn’t matter a whole lot if, say, you’re killing wolves for Wolf Fur and half of them don’t drop fur. And you’d be right! Variance is a big deal in tabletop games, but randomness has another tool in video games that’s not as present in tabletop experiences even if it does have similar roots: investment.

Rest assured, this existed in tabletop games too, as surely as there’s a whole section in my vintage Advanced Dungeons & Dragons rulebooks discussing the rules for exactly what magic items can be found in a treasure hoard of a certain type. The result is the time-honored process of being encouraged to go out and slay more stuff for a chance at the magical weapons or armor or whatever you really wanted. So your run today makes you want to play more tomorrow and maybe buy a new sourcebook or two so that you can go run in the Cave of Remarkably Easy But Treasure-Laden Monsters once you whine enough to the person in charge of the game.

Once you pass the age of 14 or so, though, you realize that the game will let you just… give characters whatever. Those Boots of Striding aren’t real. You can just tell the GM you want them and what do you know, the next time you get a magic item, you have them. Or the GM just never has them show up because they’d wreck some encounter or another.

Video games, on the other hand, do not have someone running the game who can be reasoned with that directly. Instead, you have code. And herein randomness can help capture your investment in one of two ways: You keep running content to have that 1% shot at something you want (a time investment), or you keep buying boxes of random stuff in the hopes of a 1% shot (a money investment).

Really, what matters there is more about business model than anything else. If the game has a subscription model, a time investment is way more valuable, since you develop more of a habit around the game. If it’s free-to-play, a money investment is better. I’m listing both because they’re both fundamentally taking something you have in limited quantities and asking you to devote it here.

Is that a bad thing? No. Also, yes. It’s a complex thing.

All these moments will be lost, in time...

We have generally accepted that money invested for power breaks the game, due to the fact that… well, it does. Money invested for cosmetics? Not so much, especially if there are other cosmetics available that don’t require money. But time investment is one of those things that we still generally accept as part of games, simply because we all know that most games require a time investment to stop being awful at them.

We were a little less accepting of this back when, say, games like Darius were new and going concerns, requiring both time to know how to play the game properly and more money each time you tried again. But at this point, we’ll buy games that tell you things like, “You have infinite lives, and you will need every one of them because this game will kill you over and over again.” Time investment means skill.

However… randomness kind of breaks that sense of things. You farm for eight hours to get the materials to build a new jacket, then you build the jacket, then you do more damage and take less damage. So you feel like you’re better now, but… what really happened is just that you sunk time in, and eventually that time rewarded you in exactly the way you knew it would. It’s not a matter of more skill, but it sort of fakes its way into that.

None of this is to say that randomness is inherently bad or good. Variance might feel like it’s the “good” reason for randomness to exist, but variance can either be too far (some fights you can’t use any of your abilities, others you just sneeze and kill your target) or too limited (if your only variance in combat is critical hits, people will not be invested in the hundredth identikit combat encounter). They’re both just reasons.

Instead, it’s a useful way to think about the random stuff that you’re experiencing. To pay attention to the game and ask if you would still play the game without these random deviations in either form. And, perhaps more importantly, to determine if the problem is that the random chance is off in some direction, or if maybe this would be more fun if you weren’t randomly hoping for your Boots of Striding to show up when the game could just give you the damn boots.

Sometimes you know exactly what’s going on with the MMO genre, and sometimes all you have are Vague Patch Notes informing you that something, somewhere, has probably been changed. Senior Reporter Eliot Lefebvre enjoys analyzing these sorts of notes and also vague elements of the genre as a whole. The potency of this analysis may be adjusted under certain circumstances.

No posts to display

newest oldest most liked
Subscribe to:
Jon Wax

I think maybe there’s a schism here. Things aren’t really random. Or I should say randomness is just a probability of a scripted event. Variable is probably a better word. Weather is variable but not random. If you know the probability of an event, you’re just chipping away at triggering the event. A variable environment wouldn’t allow for such predictability. Hopefully quantum computing can create truely dynamic environments


“why we still have so much random stuff when no one actually likes anything being random.” Says who? That’s a pretty huge assumption.

“You can just tell the GM you want them and what do you know, the next time you get a magic item, you have them.”
Sorry to Scotsman this but: No decent DM on earth would give a player what they want just because they want it.

The role of dice (lol) in games isn’t variability, in any game more complex than Candyland. Dice are used to black-box the host of variables that aren’t worth simulating.

When you fight a wolf, why is there only a 75% chance you get fur? Because lacking a combat modeling system where you get to control precisely where and how hard you strike that wolf while fighting ostensibly for your life, there is no way to non-randomly model the likelihood that your various stabbings, whackings, cuttings, etc haven’t damaged enough of the skin that whatever is your purported reason for GATHERING said skins is thwarted or not in this instance.
So the designer wings it and says “you have a 75% chance to get a skin” and moves on to much more important things.

Now….black-boxing it like that ALSO takes away player control. If you’re a supremely skilled weaponmaster, presumably you could more carefully control HOW you kill that wolf to preserve the skin better…if the dev felt that was “worth” simulating, they could build in a variable rate of recovery compared to your weapon skill level, invulnerability to the wolf’s attacks, or magic used to name a few.

If you want to see a game where EVERYTHING IS PRETTY MUCH MODELED…play Dwarf Fortress. And enjoy their definition of “winning” because the more complex and detailed a model is…well, then the better the players have to ACTUALLY be.

Randomness in this wolf-model protects the PLAYER from having to understand the nuances of killing a wolf without damaging its fur. Is that something we really need to learn to play an MMO? Would that be fun?

As George Box (IIRC) said – all models are wrong, some are useful. Meaning that any ‘model’ of reality is going to shortcut aspects, and that if you ultimate accurately represent every variable your model IS AS COMPLEX AS REALITY so you’ve gained nothing by modeling it.
Randomness specifically allows us to have models of complex things that are simple enough to be fun to play.


Sorry to Scotsman this but: No decent DM on earth would give a player what they want just because they want it.

Outright give, no. But plenty of good DMs will allow the player to earn the item they want or need, bypassing the loot roll, through specifically crafted gameplay that involves tracking down the item.

So the designer wings it and says “you have a 75% chance to get a skin” and moves on to much more important things.

That is the lazy way out. And a particularly bad one because it removes player agency, making the whole thing less interesting or engaging.

Monster Hunter actually solves this issue in a quite interesting way; the attacks you use and where they hit influence not just what you get when you win, but also how well the monster fights back. For example, repeatedly hitting a monster’s head will cause more stuns and break any horns, preventing the monster from using horn attacks and making the fight easier – but it will also prevent you from recovering an unbroken horn you might want.

If you don’t want to simulate to that extent, just allow the players to accept a penalty to their attack rolls in order to avoid damaging whichever monster part they want to recover; most tabletop RPGs I’ve played have rules for that.


Well sure, I agree – if you’re collecting spider gizzards, I can certainly see a “attack carefully: double your chance of getting the thing you need, but you suffer X penalty to attack% and -Y to damage”. That makes perfect sense to me…

(BTW I’ve – seriously – played a majority of the tabletop FRPG rule sets issued commercially since 1978, at least through about 2005 – I can’t think of a single game that had an explicit rule for that sort of thing. But that’s the joy of an actual, human GM vs a computer rpg generally: they can make such a ruling on the fly that rewards particularly creative, lateral HUMAN-style thinking.)

..IF that’s a level of detail the dev wants to go into.

In me experience in development, need for features is endless (cf Star Citizen) while resources aren’t (but apparently NOT including Star Citizen…).
Does it make sense to have a detailed extra UI item for a “monster body part recovery” mechanic and rules, and NOT have an ESO-like detailed UI/mechanic for lockpicking? Would it make sense to have those and NOT have a more-detailed and engaging EQ2-style crafting system? As a dev, you have to draw the line somewhere.
Let’s not forget that this is the internet and it’s 2019…for every person CHARMED by your attention to detail, there’s likely to be some whiny bitch on reddit complaining that you ‘wasted time’ with this ‘stupid detail’ while expounding in a 41 page manifesto why you should have spent more time on boob physics.


Oh, and FWIW…in my RPGs if a player wants a magic item that does X, they have basically two choices:
– research if such a device exists, and if so, then where they might find it and then go about getting it through whatever means are necessary, or
– if such a device doesn’t exist or gaining it would be prohibitive, develop the skills and resources necessary to create it on their own if possible.

Basically, there is null chance* in any game I’ve ever run that someone saying “gee I wish I had a magic bow” would in any way meaningfully result in me “putting that in some treasure pile for them to find”.
*ok maybe not null – in this case, “random” treasure generation is something that would work in the favor of the player. I’d NEVER place ‘desired’ loot in a random hoard, but likewise if such loot is fortuitously randomly generated that is precisely what a player wanted, I would never NOT put it in. In that case, the luck of the gods was in their favor.

I concede, that could certainly be a generational thing. I’m a first-gen RPGer. I stoutly *don’t* believe in RPG rules, games, or settings that ‘cater’ to the whims of players. Ick. That’s like reliably getting precisely what you asked for, for your birthday present. How incredibly dull that would be?

Bruno Brito

Ick. That’s like reliably getting precisely what you asked for, for your birthday present. How incredibly dull that would be?

If you’re playing a Vampire, let’s say a Brujah, and you have three Potence dots, and you engage into combat with a minor ghoul, it’s pretty much staple that no combat rolls are needed and you just destroy the ghoul with a thought.

The point is simple: By allowing the vampire to do whatever he wants and having the power to achieve that, it lulls the vampire into the feeling that is the game: Being a powerful creature that sooner or later will have to deal with the consequences of what they do.

Good storytellers are directed to, instead of impeding the player, they make them pay the consequences of their actions.


Except we weren’t talking about the results of actions, but the rewards therefrom: should they always be what benefits a player or fulfills their specific needs?
I’d say no, that’s dull.

Bruno Brito

but the rewards therefrom: should they always be what benefits a player or fulfills their specific needs?

Rewards should ALWAYS benefit a player. If you’re giving players patches of grass for a work well done, then your players will feel that their hard work is for naught.

Now, i don’t think all rewards should be useful towards every single situation, but they should be useful.

Rewards are rewards. They’re supposed to be good.


Meh, I’m not sure if we’re just poorly communicating or if you’re being deliberately obtuse.

What I was discussing is specifically what I talked about: if a player says “I really want x” no way am I going to simply, on the basis of that request, make sure that x is in their next loot pile. No way.

Does that mean the alternative is to give them utterly worthless trash? No, of course not. At worst, “loot” – ie whatever they monster felt worth keeping – is likely saleable for some value. Or they may find things otherwise useful and precious – but not specifically what they asked for.

Personally, my monsters have treasure that makes logical sense for them to have accumulated in their existence; what they have CERTAINLY has nothing to do with player desires, except perhaps coincidentally.

Edit: as far as I can tell, my players want a realistic world to function in, not a dm that is Santa fulfilling their wishes every time they express one.

Bruno Brito

Poorly communicating. I’m not someone to shittalk DMs.

Nor do i think that players should have everything they want without working for it.

I just think that the randomness that led us into Classic WoW territory of Horde getting Paladin gear is bad. Rewards should be useful.


That’s like reliably getting precisely what you asked for, for your birthday present. How incredibly dull that would be?

Actually, from the moment I first learned how money works I kinda stopped getting enthused about birthday presents that I hadn’t personally chosen. Unless it was something I would have purchased myself if given cold hard cash equivalent to the present’s value, such a present had a fairly large chance of not actually pleasing me, but instead making me frustrated at the sheer waste of money.

As I said elsewhere, I hate any kind of randomness in rewards. That extends to the real world; giving me a present I had no role in chosing tends to be a waste of money unless whoever is chosing the present knows me well enough to reliably pick precisely what I would have asked for.

Arnold Hendrick

As a game designer (and frequently producer) in computer games since 1983, and in paper games going back a couple decades before that, I have a slightly different opinion of randomness.

Randomness is good in a game when there are a lot of small random events whose average likelihood is predictable. For example, hitting a MMORPG mob with your weapon (or spell) has a certain probability. But in a well-designed game, in each battle you’ll need multiple hits to kill them. By comparing the general rate of incoming damage to you, versus your damage to them, you can decide whether to keep fighting or run. If it’s close, you realize that chance could determine the victor.

However, the point is that the player can see it coming, and take action, based on how much risk they like. This is why many small random events are generally good. Of course, the game UI might be so poor that the player can’t understand what’s going on. But that’s a different design problem.

Randomness is a bad in a game where one random event has a very large psychological effect on the player. Would you enjoy a card game where, after shuffling the deck, any red card means you win, while any black card means you lose? Where’s the gameplay in that?

A more sophisticated example is from the old board wargame “Afrika Korps” (Avalon Hill, 1963). After considerable play and replay, the best moves for each side boiled down to the Allies putting all their strong troops into Tobruk to defend it, to which the best Axis move was to attack the place. A single dice roll using the “2-1” column on the CRT (combat results table) determined the result. Each side had a 50-50 chance of winning the battle. Winning that battle pretty much determined who would win the game. Players soon decided to stop playing the game. One friend of mine said, “Hey, let’s just roll the dice to see who wins? Why bother spending an hour setting up and moving around the counters just to reach that one battle?”

I believe this is why MMORPG players hate randomness, especially when it comes to Epic/Legendary loot. Why must they spend a couple hours doing a dungeon to see if the right loot drops? If they could, they’d want to see what loot they would get before spending those hours in the dungeon.

This design rule applies to subscription and F2P games. Just selling the loot for cash raises the spectre of P2W, which is the kiss of death with North American and Western European audiences. Regardless of how the company makes money from a game, it is critical to keep people playing and enjoying the experience. Successful game companies know that customer LTV (Lifetime Value) is more important than any one item sale, or any single subscription renewal.

Sorry to be long-winded, but design and business tradeoffs, when overly simplified, cease being useful.

– Arnold

Robert Mann

Randomness is there for both those reasons, but there is yet one more. TTC. That’s ‘Time Toward Coding’. In short, without some randomness involved, and the more randomness the more benefit is reaped here, coding cannot keep close to consumption.

This combined with investment is where games often go off the deep end. They have limited the scope of activities, and the parts of the world that really matter usually (generally new content, and generally with combat that means the most recent dungeons and raids) as their ‘ideal’. As there is a lot of limitations to the random factors for those bits of combat content, there is a longer TTC. This means that investment for that highly limited content area is a big problem. Thus two things must exist: Barriers, and randomness for investment in loot.

That’s the traditional model for MMO combat centered around ‘endgame’. It is also fairly reasonable to be reviled, as the entire goal is “Pay us to stay here doing the same thing you have done for months, because we made this low chance so you don’t play other games on us!”

It is, in all honesty, part of the reason why I have lost faith with combat as the main focus of MMOs. Combat is a good thing for excitement, when not overused (hint, it’s overused). The issue is that combat requires a lot of that TTC, which means the randomness and delays for more content are inevitable. It also means that the more complex and thoroughly planned that content is, the greater all those factors hit. So where combat can and should be part of our games, I believe the key to MMOs that are fun, and lacking in the insane grind toward rare drops, is making a variety of goals to be met (indeed, sometimes combat should not be about the drops, but other progress).

Matt Comstock

I wouldn’t say universally reviled, as it depends on the circumstances in which the randomness presents itself.

If it takes a huge amount of time and effort to get to the point of the random element, lets call it a loot roll, and the odds of getting what you want are extremely low, I would agree that is universally reviled — as failure requires that you sink another huge amount of time and effort to get back to the random element with an extremely low chance of success. Which means there is an extremely high chance that you will continue to sink time and effort without into the game without reward (i.e. wasting our time).

MMO Players, want, need, to be rewarded for the time and effort they put in to the game. Particularly players who don’t have all day to play games ;)

My egregious example is the BDO weapon/armor upgrade system — if you’ve played the game, you know what I mean.

The solution is to have a meaningful safeguard, in that after “X” attempts you get what you came for. Safeguards such as getting tokens from each encounter, used to purchase the coveted gear. And, I think most MMOs have this to some degree or another.


I was under the impression that randomness, particularly in combat, was there to help simulate “real” combat and not just for varience.

In a real life sword fight, it’s not really random but sometimes you’ll hit, sometimes you’ll miss, sometimes you’ll block or parry or evade. What happens is based on the skills of the individuals.

In a game, we lack the input devices to make that sort of realism possible. A sword fight is simply too fast for us to be able to individually aim each strike, or to halt a strike mid swing and switch to a parry. There is no button for “suck in your belly” so that you can avoid a mid-aimed swing, and even if there were, the lag involved in gaming (whether thats input lag, networking lag, graphical lag, processor lag or whatever) would make the skill virtually useless to most of us.

So, they give us skills, stats and randomness instead.

We may lack the input devices for parrying effectively, but I can build up my parry stat, balance that against the opponents aim stat, and use randomness to determine whether I parry or not.

This is part of the reason why I like “old school” mmo combat the best, with loads of skills and skill bars. It gets the closest to simulating combat that I’ve seen so far. I may not be controlling the action directly, but I’m giving commands to my character who is then simulating that combat. Compared to action combat with just a few skills and a lot of spamming, traditional mmo combat seems much more engaging to me, much closer to realism. (and yes, i know none of it is actually realistic, but action combat always seems about as far away as possible from real combat as you can get).

At some point in the future, we’ll get full virtual reality, as in exoskeleton suits that act as input devices and feedback devices. Then we can get rid of the randomness in combat because we can finally have the precision and speed of input that is really needed.

Robert Mann

To some point originally. MMOs are not, however, aiming toward realistic combat in most cases anyway. So where that applies to the PnP games that would inspire people, I hesitate to agree that is why MMOs use it. In fact, if that was the goal it would be completely opposed to so many of the other aspects of combat in games that use randomness in combat that it would make even governments flinch at the discongruity.

Bhagpuss Bhagpuss

“If there’s one concept that is nearly universally reviled among MMO players, it’s randomness.”

You’re kidding, right? I love randomness. It’s a primary attraction of playing these kinds of games. I’ve posted about it many times and had plenty of agreement (and also the opposite). It’s far from “universally reviled”.

Rheem Octuris

I’m with Mr Bhagpuss here. The source of the hate is when you KNOW you want item X to drop and you keep getting item Y. This is the problem with prebuilt sets, rather than a series of slight upgrades over time. With prebuilt sets, you know there are only 3 upgrades for this slot, and they are tiered and until you get that last tier you are going to feel cheated.

The problem isn’t randomness in the acquisition of items, its the lack of randomness in the items themselves that make the random component unfun. The fact that only that one item could be an upgrade versus ANY item being potentially an upgrade is the real problem, and it usually comes down to laziness by designers copy and pasting the same item over and over again but with different tiers.

Robert Mann

The problem here comes into play with best stats. Sure, randomized items can be interesting in some cases. However, if the best combination of stats is always (for example) crit and cooldown reduction… then unless something is super OP the same scenario will play out, just with 90% of the drops that could be good for the person being some other combination that is likely not an upgrade.

I believe that the drops are a big source of the frustration, but I believe that the solution isn’t just changing that (but rather giving more motivations than loot to the genre).

Morgan Filbert

I agree with Bhagpuss as well. I really enjoy randomized mechanics in games.

Robert Mann

Depends on the randomness involved. There’s good randomness indeed. In fact, many games are quite enjoyable due to randomness. Randomness can mean more content, new outcomes, and new adventures. To get there, MMOs would have to stop being married to the idea that specific dungeons and encounters are the constant that must be kept.

On the other hand, there are factors in MMOs where the randomness is very bad. Not only does it make the loot chase an even greater focal point (and generate a lot of drama in turn) but it emphasizes the time between big patches. Instead of randomness to make more content, it is randomness to keep people from moving on. It is this reason why so many MMO player rail against randomness, and they are only looking at this negative side.


I really hate randomness in two situations: when it’s used to determine rewards, and when it can make you lose regardless of how well you play. In most other situations randomness can be quite enjoyable.


Got to agree and even expand on what was said I dont even want Boots of Striding have always the same stats but rather have some random range in their stats. There is a reason diablo 2 did that with uniques for example.


Yup randomness is what makes a mmo come alive. If there are no downs or chanc of failue (variance), then you will experience no ups either (excitement); You can argue whether mmos with much randomness are deliberately (over)playing on this psychological effect, but if it weren’t there at all the game would be boring because it would be completely predictable.

Personally, I find it extremely rewarding to get something rare that I had to put effort into getting, that appears slightly out of reach, that I know the next player after me needs to be lucky and put in effort to get – That game experience gives “feeling of accomplishment” and no FLAT (no randomness) design can ever compare to it.

Randomness is not universally hated by players. Some may think they don’t like randomness, but they will experience the exact same thing if it was removed; boredom. Especially for long term games like mmos, randomness is very important. Obviously shorter games do not need as much randomness, and if you think you dislike randomness, maybe it is because you are actually more into shorter games and shorter game sessions, and don’t want a high degree of variance between those short sessions.

RNGesus is the saviour of games, and mmos in particular.