The Two Types of Random in Game Design -

The Two Types of Random in Game Design

Game Maker’s Toolkit
Views: 2282845
Like: 80460
From critical hits to random encounters, and from loot boxes to procedural generation, video games are stuffed to bursting with randomness. In this episode, I look at the way randomness is used in games – and why some forms are more contentious than others.

Support Game Maker’s Toolkit on Patreon –

Have Mark talk at your studio, university, or event –


Uncapped Look-Ahead and the Information Horizon | Keith Burgun

A Study in Transparency: How Board Games Matter | GDC Vault

GameTek Classic 183 – Input Output Randomness | Ludology

Why revealing all is the secret of Slay The Spire’s success | Rock Paper Shotgun

Crate | Spelunky Wiki

Random Generator | Tetris Wiki

Level Feeling | Spelunky Wiki

Plan Disruption | Etan Hoeppner

Fire Emblem True Hit | Serenes Forest

The Psychology of Game Design (Everything You Know Is Wrong) | GDC Vault

How Designers Engineer Luck Into Video Games | Nautilus

Roll for your life: Making randomness transparent in Tharsis | Gamasutra

12: Into the Breach with Justin Ma | The Spelunky Showlike

Find out more

Many faces of Procedural Generation: Determinism | Gamsutra

Why Our Brains Do Not Intuitively Grasp Probabilities | Scientific American

How classic games make smart use of random number generation | Gamasutra

Games shown in this episode (in order of appearance)

Cuphead (2017)
Enter the Gungeon (2016)
Octopath Traveler (2018)
Mario + Rabbids Kingdom Battle (2017)
Griftlands (In Early Access)
Dicey Dungeons (2019)
Hearthstone (2014)
The Binding of Isaac: Rebirth (2014)
Darkest Dungeon (2016)
Dead Cells (2018)
SteamWorld Quest: Hand of Gilgamech (2019)
Into the Breach (2018)
Spelunky (2012)
Armello (2015)
Minecraft (2011)
Chasm (2018)
Downwell (2015)
Middle-earth: Shadow of Mordor (2014)
No Man’s Sky (2016)
Celeste (2018)
Fortnite (2017)
Mario Kart 8 (2014)
Super Smash Bros. for Wii U (2014)
Tekken 7 (2015)
Super Mario Party (2018)
Bloodstained: Ritual of the Night (2019)
Borderlands 3 (2019)
Call of Duty: WWII (2017)
Valkyria Chronicles 4 (2018)
Civilization V (2010)
Wargroove (2019)
Plants vs. Zombies (2009)
XCOM: Enemy Within (2013)
Chess Ultra (2017)
Mark of the Ninja (2012)
StarCraft II (2010)
Slay the Spire (2019)
Apex Legends (2019)
Civilization IV (2005)
XCOM 2 (2016)
Overwatch (2016)
FTL: Faster Than Light (2012)
Card of Darkness (2019)
Diablo III (2012)
Tetris 99 (2019)
Puyo Puyo Tetris (2017)
Phoenix Point (2019)
Fire Emblem: Three Houses (2019)
Tharsis (2016)

Music used in this episode

Cuphead soundtrack – Kristofer Maddigan ()
Tharsis soundtrack – Half Age EP by Weval ()

Other credits

RNGesus original artwork by Dinsdale

Super Mario Party – Luigi wins by doing absolutely nothing | Nintendo Unity

Fire Emblem: Three Houses – New Game Plus Maddening Walkthrough Part 43! | MrSOAP999

Deadpool 2 © 20th Century Fox

Pandemic Card Art © Z-Man Games

Contribute translated subtitles –


  1. Computers can do "True Randomness" when they use external information that is truly random (at least, if we count anything in the universe as "Truly Random." Maybe nothing in The Universe can be "Truly Random," but that's a whole philosophy thing.)
    Anyway, you have computer listen to static noise, or the noise of the universe or something, and generate it's numbers using the RNG off of that… if it has a completely random input and is programmed correctly, it can be a completely random output.

  2. About using dice instead of cold percentages, there is a kind of die called a percentile die, which is actually two 10 sided dice going from 0 to 9, or 00 to 90 (increments of 10 obvs) and then you just treat a result of 00+0 as 100.

    Tellingly, I used to feel differently about percentile probability when used in ttrpg's with dice, than I did when people just tqlked about them, but since I have come to just alleays imagine someone is rolling dice, and it's very useful.

  3. I passed over this video many times figuring it was about true vs pseudo randomness, but glad I finally clicked. A really cool video. Thanks!

  4. Video is about randomness in games.
    Video mentions XCOM by name at least 5 times.

    Me: That seems low…

  5. Seed rand= procedural…

    Normal random, random seed = pretty random…

  6. Fallout 76 takes a bucket load of s*it on this video. It relies on randomness to startup, function, choose weather or not your good enough to combat, your looting outcome. Bethesda is enjoying punishing players with rng just because they are still playing the game. I kinda was hoping you had the fallout franchise in this video too i believe it would have added a little spice to it

  7. 50% had no idea what random number generator means but knew RNG

  8. As someone who played poker in the past, RNG in games barely affects me, because I understand that in the long run winning strategy is winning strategy

  9. I have the board game pandemic, it's pretty fun

  10. But randomness in fps games help create those rare high moments where a single player goes on a rampage and kills 3 people

  11. 11:23 "I've never said the word 'epidemic' so many times". 14 January 2020

  12. in plants vs zombies new players can get surprised by a new enemy and not know what their best bet is for the gargantuar you best bet is to use something like squash and then use the plant that clears the row that i cant spell the name of

  13. The only good luck in videogames is fallout new vegas luck. fuck ive made so many caps playing blackjack

  14. This makes me think of Team Fortress 2. In TF2, there are multiple types of output randomness, including random bullet spread, but the random crits are the one people hate the most. They occasionally give a player triple damage on a shot. Of course, this means another player is randomly getting oneshot by something that otherwise might not have killed them. The biggest problem, in my opinion, and why the TF2 community talks more about random crits than random bullet spread, is that random crit chance actually increases with damage done. Thus, the better players get more random crits than new ones.

  15. This is very true in MTG. Sealed and draft has input randomness and it's a type of random that I like and isn't fustrating because it creates variety in the cards you use and the decks you play against considering how every time you enter a sealed or draft competition the decks will always be randomized so you will never play against the same deck. This is in contrast to playing modern where there are popular decks that many people use and decks that you will most likely encounter which can argubly get boring after awhile. Same thing goes for the deck you are using with sealed and draft every time you enter a new competition you use a different deck where with modern you have to make a new expensive deck if you want to use a different deck otherwise you use the deck you already have and have most likely already used

    However, in the same game there's also randomness with what cards you actually draw. it's the randomness that I hate and is one of the problems I have with the game. I think this is also input randomness but it's the input randomness that I don't like

  16. This video explained why I like randomness in some games and why I don't in others. I caught myself one time saying that I hate luck in games but then realized I still love Mario Kart

  17. Weirdly, I don't like any randomness at all in situations that seem like they're skill based or related to treasure that is directly visible.

    That's what I liked about the Elderscrolls IV and V, the combat is entirely up to your ability and the loot is what your enemy is wearing.

    I tried Elderscrolls Online during the free weekend, and not getting a bow when I killed an archer and having armor randomly generate off of storage racks and abstract floor items was so jarring I quit in an hour.

    I can see the metal gauntlets on the table, why do they turn into cloth mittens??

    I killed a man with a bow and arrows, why can't I pick it up and get the advantage of ranged attacks??

  18. First time playing Dont Starve Together I got a Treeguard in Autumn, Deerclops in Winter, frog rain un Spring, struck by lightning a day after that, and killed by mating Beefalos just three days after. All the time I thought this was ALL super random. Base was massacred, of course.Then I researched and realized it was… The most common thing to happen in game. DST has a deep and dark prize inside my heart: the most cruel game ever created for noobs.

  19. After watching this i may worship rngesus 😂

  20. Glad to hear you credit Geoff Engelstein! Very few do!

  21. My favorite memory of randomness is of borderlands 2. Was just casually playing game i got a pearl weapon from a garbage pile which has like 0.5% or maybe less to drop

  22. "(Luck) can also be a cruel mistress that leads to leads to unfair outcomes and frustraiting failures"
    Me and my 721 hours of darkest dungeon had a visceral reaction seeing it in the background lol

  23. Other than the false expectation of a enclosed/consumed sample space (ie expecting a probability of 20% will guarantee 1 hit in 5) I think with percentages we tend to try to simplify the probability. We think 95% = 100% in such a way a person would avoid a 10% chance with the expectation that it will lose every time, even if it would give a 20x reward (ie 200% outcome on average).

    Yes, saying that losing an 80% chance 5 times in a row is impossible is incorrect but given the chance is 0.2^5 = 0.032% (yes that is a decimal of a percentage) or 1/3125, it is reasonable for someone to be annoyed at such bad luck.

    I think one of the main reason people hate output randomness is its misuse for rewards. By giving something a drop chance of 1/n the developer can semi-reasonably make the player repeat the content n times on average to get the item. For low numbers of n, this can help ensure players don't just ignore content that the developer has likely spent a long time working on (ie the dungeon or boss battle that drops the item). In practice n can easily exceed the 100s if not 1000s as developers try to pad out game-play or just making intentional acquisition of that item unreasonable so that it is a pleasant surprise if they actually do get it. Though never underestimate a player's actual ability to grind. There is a reason why protecting players from themselves is a topic for discussion in game design.

  24. Computers actually can do "True" randomness actually, but you have to have it calculate human inputs at strategic points in calculation, example ask the persons desired name, gender, age, race, a starting item, you can even also on top of all that use a muon detecting like radioactive decay machine etc. and use the input times they made the decision along with the choice data with the usual time, and then with the matrix of all of those to calculate it's final outcome, even if all the people in the world attempted to use it at same time even attempting to answer the same at same times, if the data it's calculating has more outcomes than people by a few magnitudes for each stat and items and levels, then it would be quite rare any would match, and overall the whole thing even if something matched would still be different overall. As choice times in all of those would be different timings just slightly and radioactive decay be different and then you likely have a few people fat fingering some selections by mistake as that many people to remember all answers and do all exact and timings would likely not happen! "Current Time Random we all use + Human choices and timings + The Quantum Radioactive Decay" and bam all that Pre stuff to calculate the game, but if we also do cards and such then the choices they make then further adds to it. – Just give computers more input than just a clock and True Randomness will actually happen. 😉

  25. What about dithering ? Games like Diablo or Doom make the calculations more accurate, by randomizing damage. If an Enemy has 100Life, it doesnt matter, if your weapon does 20 or 21 damage (always five shots), but it makes a difference, if your weapon does 1d20 or 1d21 damage.

  26. Appreciate this video. Output randomness can actually do very well in competitive games at times, because accessing risk management itself is just another skill to add to the arsenal of a player.

  27. BattleBrothersDevs need to watch this to fix their broken anti-player RNG curve. (which favors odds against the players in an insane amount -> 1/20 (95%) is in reality between 1/4 and 1/3. i.e. even with 95% evasion or hitchance, you miss 1/3 hits / get hit 1/3 times.

  28. To mario kart's credit. It's not just luck, the reason battle royales suck is because you're constantly throwing dice to get a chance to get a weapon you're good with, with mario kart you sometimes got that one blue shell before the finish but there's a clear balance between the better items in the back that help you get closer to first place and the weaker items when you're between 3rd and 1st.

    The blueshell in that case serves more of a forced brake to stop you from being too far for the other players to catch up.

    Meanwhile in any battle royale, you're constantly on the same playingfield while not knowing what the other players have, in most other gamemodes like kartracers it's normal to kind of assume what items your opponent gets dependong on the place they are in, and even the you don't get shafted when you lose when you don't get first while most battle royales just say you've lost and you only get a sense of accomplishment whrn you've defeated the rest

  29. This discussion reminds me of the differences between the games Elder Sign and Roll for the Galaxy. Both use d6’s with different symbol faces and you roll your pool of dice on your turn to resolve stuff. In Elder Sign, you have a limited pool of dice that you need to match to objectives on a card that allows for only one and sometimes two bad rolls. You roll, hope to get what you need, and keep going until you pass or fail. Very little you can do to modify the roll results outside of temporarily increasing your die pool or using a character who previously failed the same objective to hold onto a desired roll result for later use. Its a very frustrating game at times. Meanwhile, Roll for the Galaxy as multiple ways of using dice they don’t have the desired face to power other results you do or to completely change a result. A base ability allows you to discard a die (possibly 2) to change another to any facing. Everyone gets this ability, not just some. This makes Roll so much more enjoyable to play. You don’t feel at the complete mercy of random dice.

  30. my dude will show every strategy game besides the series that invented the genre

  31. Ah yes. But don't forget synchronized randomness through the use of a synced clock and an equal seed on 2 devices. That is one of the big m***********s in programming. You want both players to experience the same thing, on screen , in the background, the same skeletons popping at the same moment from the same chest but you can't afford the bandwidth.. This is what you do.
    You make sure you are polling from an equal pool of random numbers when random things need to happen. And as I said.. if your seeds and clock are equal it should all come out the same on both machines.

  32. I just watched a ton Scott video and the voice is throwing me off

  33. great video but i can't help but cringe seeing those xcom 2 complaints from people who don't even understand that xcom's randomness actually favors them. way to misunderstand the core mechanic of the game honestly

  34. D&D is an interesting example for high output randomness that can be extremely rewarding to play

  35. As a writer and occasional public speaker, I loved that closing statement better than most of the video. That said, I did also love the rest of the video.

  36. Luck is just one thing that people do not understand. There is absolutely no true certain probability. Things just happen. If we think future to past, in reality one of the outcomes is a 100% chance, because that's just what is going to happen. Its useful to predict an outcome, but its ignorant to assume the outcome will follow the prediction. The outcome has no bias or need to follow the predciton. The outcome just is.

  37. I think the best way to understand output randomness for me is just dungeons and dragons

  38. "And it can turn players into risk-calculating tacticians" sounds like a good way to say "forced maths"

  39. god, that epidemic section did not age well

  40. If you know how rng is generated in that game, you will win every match in that game (if you are skilled enough).

  41. We think stuff like 99% chances aren’t actually those odds and are a lot lower because we ignore all the times they work, and we only remember that one time you get incredibly unlucky

  42. The worst randomness in gaming is when a CoH2 tank misses the final shot.

Leave a Reply

Your email address will not be published. Required fields are marked *