The Delta of Randomness - Can You Balance for RNG? - Extra Credits
HTML-код
- Опубликовано: 10 ноя 2015
- Randomness has three major functions in game design: 1) it creates exciting moments, 2) it gives the weaker player a chance to win occasionally, 3) it forces players to adapt to changing circumstances. When adding random elements, game designers must look at the delta of randomness, or the difference between what baseline impact that element has without a random effect and the realized impact it has with its random effect.
Subscribe for new episodes every Wednesday! bit.ly/SubToEC (---More below)
_______
Get your Extra Credits gear at the store! bit.ly/ExtraStore
Play games with us on Extra Play! bit.ly/WatchEXP
Watch more episodes from this season of Extra Credits! bit.ly/2wUpNyb
Contribute community subtitles to Extra Credits: / timedtext_cs_p. .
Talk to us on Twitter (@ExtraCreditz): bit.ly/ECTweet
Follow us on Facebook: bit.ly/ECFBPage
Get our list of recommended games on Steam: bit.ly/ECCurator
_________
Would you like James to speak at your school or organization? For info, contact us at: contact@extra-credits.net
_________
♪ Intro Music: "Penguin Cap" by CarboHydroM
bit.ly/1eIHTDS
♪ Outro Music: "Gambling Nights" by Abreu Project
bit.ly/1kM8OYT 
Игры
Oh! Just to let everyone know, the reason I usually talk about things in the context of arena when talking HS is because I have no idea how the meta will change between when I write these and when we release them (which often is a month or more), whereas the arena meta changes much more slowly (so if I say something truly foolish at least I know it'll still be 100% my fault ; )
-JP
P.S. Oh and I was always more of a limited player...
+Extra Credits Hey Next time you do a hearthstone related video, could you maybe have +Kripparrian or +Trump give their opinion too? They are both have a huge amount of insight and knowledge about the mechanics and what not in hearthstone.
+Extra Credits I would love your input on something. I have been working on designing a strategy combat card game (played with a standard playing deck) and I am curious what you think the delta of randomness on a certain game mechanic is.
When you have your fighter (say, a queen {2atk/15hp}) attack an enemy fighter (again, we will use queen) the defender has the choice to counter attack. If they do the defender flips a coin. If the defense wins the attacker takes double their own initial damage (hits shields first) the the defender loses they take double the attacker's initial damage (ignores shields and does damage straight to defender-not countering and taking the hit normally would allow shields to block hits) The counter does, of course, account for any buffs the attacker has but defender's buffs are irrelevant)
+Extra Credits When you balance for RNG in gaming, do you balance it around basic expected (average) value, or something else? E.g. Would you want to assess and compensate for a player's risk preference, or does that have too little bearing in a typical game?
+HellsingRuler I really think Kripp is significantly saltier than the norm, so the level of RNG that is unfun to him is not the same as that which would bother the average player. I mean, sure, great player to talk about mechanics, balance, and strategy. But game philosophy? Eh. And I'm saying this as a fan.
Tudor Naconecinii How about trump from a mechanic standpoint? I mean he has his arena site. And also I may just be annoyed that Extra Credit has said that Blizzard cant nerf cards post release just as before they did it with Warsong Commander.
And let me just say this, Extra Credit are amazing and super good at academic stuff, but when it comes to discussing a specific game.... I just don't get the feeling they are on the same level, like far below that level.
I'd honestly like to see this discussion within the context of pen-and-paper RPGs like Shadowrun, CoC, and D&D- especially the latter example since a d20 roll can be a very feast-or-famine element of RNG.
+VyseElric Ya know, I'm almost certain CoC isn't the game I'm thinking of.
On-topic, though, Tabletops are a little trickier to discuss because they have a human arbiter, who might make different decisions on what constitutes a roll or not. For example, I have a dungeon master who would make me roll to try to scoop up and carry another medium-sized creature with my 18 str Barbarian. Meanwhile, I personally would say that the barbarian can just do that, but if he wants to try holding up the collapsing ceiling trap, THAT would take a roll.
+VyseElric Ok, I'll start. I have played only one partial game of dnd because I can't find a DM, and the campaign fell apart due to scheduling. It was 5th edition and I realized that we didn't have any spellcaster except the cleric, so I made a warlock. Knowing who was playing the cleric I should have made a healer, but hindsight is 20/20. I was so afraid of a bad roll of the dice killing my warlock I made him a mountain dwarf with 20 constitution, medium armor, and lots of temporary hit-point production. He was so beefy he could off-tank, though that may have had something to do with the rest of the party, save my sister's fighter, being glass cannons. Including the cleric, who decided just before the campaign to go tempest and make his highest stat strength. It turns out I was completely justified in doing so, as I don't think I ever made a save vs an attack in any boss fight, and as the spellcaster got targeted for termination. We had a reskinned Chimera (in a sewer, so there was nowhere to maneuver) make a full damage breath attack on the party (before we knew what it was), and only I didn't make the save. Despite having the third highest save modifier. It took out 3/4 of my health, so I then promptly decided to perform a tactical retreat into cover. We barely survived the fight, since the cleric decided melee was a good idea. The second fight I saw it coming from a mile away (Conveniently geometric area with terrain blocking ranged attacks and preventing/heavily restricting movment around it, with a circle in the center that glows. It screamed boss fight.) and, deciding to not just take cover but to try and prevent the boss fight before it had even began. Knowing our DM, I should have known better. I decided to bomb the !@#$ out of the summoning circle from afar with hellfire (really just a fireball spell) and follow up with scorched earth (merely burning hands). The DM, I think he wanted a boss fight, had this trigger the summoning of minor Cthulu villain expy 404. Fortunately, I thought, I had taken cover in a massive tree and was out of melee. I then failed my paralysis save vs constitution... with 20 constitution. While still in the tree, and I had climbed high enough to try and get out of range of, well anything, that the fall damage would kill me. Fortunately I could make a constitution save to break free of paralysis, constitution being both my highest stat and the maximum you can legally have in 5th edition, baring enchanted items. Of course, I failed every save for the entire fight... and the party member with the minimum legal constitution made his immediately. The only decent rolls were the "not fall to my death" rolls that were completely random, and they barely passed.
Mercifully, that was the last time we did that campaign as scheduling issues prevented a game for a minimum of 2 months and everyone lost interest. I have a possible game lined up, and once again I'm planning for the worst.
And this time I'm playing the healer. I'm thinking a mountain dwarf bard with 20 constitution, medium armor, and nothing but healing spells and only attacking with cantrips as a last resort. Healing spells can't miss... right?
ilovethelegend True, but there are certain axioms of the game that are best left unignored, such as needing to roll to hit, save, and so on (in the case of DnD; you don't want to fight in Call of Cthulhu). DM Fiats can only be considered for so long- eventually there has to be law within the chaos of the narrative.
And speaking just within those axioms, I think a discussion on the game's design with respect to RNG would be just as compelling as any other game type.
Play a life cleric if you're gonna go healer. You'll have a lot of the best healing spells on tap and they won't count against the number of spells you can prepare per day, you get access to heavy armor, and your channel divinity becomes a powerful heal as well.
Anyway, I would suggest getting a new set of dice if you've had such a huge string of bad luck; it's possible that your D20's balance is off. You can check by floating it in a cup of salt water and seeing if it likes to settle on any particular face.
+VyseElric *Flashbacks to the too many to count 1s* So many crit-fails.
So how about Dr. Boom, basically a 9/9 for 7 (already far above the curve) with a random effect that could deal 2 upon death (which is ok) to 8 upon death (which is broken) How did this card ever see the light of day?
+CommonChicken
Yeah, Dr. Balanced FTW. The only legendary i crafted out of spite.
+Dr. Flash ...you do know that other neutral 7 drops don't see play because Dr Balanced exists, right? Why would you play anything else when you could play him? And at 7 mana, there are plenty of tempo cards - see War Golem and Captured Jormungar. Hell, even at 8 mana you have Force Tank MAX, and 9 you have North Sea Kraken.
But OK, two other neutral 7 drops that see play - Baron Geddon and Stormwind Champion. Both significantly below Dr Balanced on the power curve. The other neutral 7 drops are so far below Dr Balanced that they will never see play outside of specialised decks. So yeah, Dr Balanced is well above the average, or the power curve, for 7 mana neutral crads.
(P.S. - I have Dr Balanced, opened him in one of my free GvG packs. I'm not raging jealously against a card I don't have. The card is just objectively OP.)
+Prof 1 The issue with the neutral 7 mana slot is that before Boom came to picture neutral 7 mana cards were mostly shit,Dr.Boom was the first good 7 mana neutral that could easily go into many decks.Baron Geddon only ever saw proper play in control warrior and all the other good 7 drops were class cards.
War Golem,Captured Jormungar and Force Tank MAX are not really tempo cards since they do absolutely nothing on the turn they come into play and North Sea Kraken while it is a tempo card it is very late and has pretty bad stats.
+CommonChicken Boom is one of the cards I'm watching. How they deal with Boom will tell me a lot about their design philosophy and just how much they value not nerfing cards...
-JP
+Extra Credits Given what happens when they DO nerf cards (yes, still sore about Warsong Commander thank you very much) let's hope they value not nerfing cards very highly...
Closing out with a Casino Night Zone remix.
I see what you did there. B)
Wow, I didn't even realize! XD
+PCachu You caught me. :D
My favourite sonic level
Could you please, for those of us who don't play hearthstone, give a quick rundown of what a card does before jumping straight into analysis in future episodes where you use it as an example? I had trouble keeping up, and this is a concept that applies to a lot of games.
+Eon2641 I mean specific cards. I have to pause the video on the split-second they show each card to read what it does or else be confused for the next section of the video.
+Jake Pillsbury To put it simply, starting with your first turn, you add 1 mana cap until you reach 10 mana. The amount you spend comes back each turn. So, if you last long enough, you can spend 10 mana a turn. Your hero (player) has 30 health. Minions are your way of attacking. If a minion attacks another minion, they both take damage accordingly. You don't have to kill the minions to attack the hero, but it's advised to kill them so they don't kill you.
+Eon2641 I was more referring to the ones where he was talking about what specifically they do right. He'd say something like, "The delta of randomness is here pretty low. If you ignore its random effect..." And I'd just go, "What delta? What random effect? What happens if you lose a roll?"
+Jake Pillsbury 'Delta' is not a game term, but a mathematical one. When they say delta they mean "change in" or "difference between". So the "delta of randomness" of a card is the difference between how good it is when the randomness goes your way vs how bad it is when the randomness doesn't go your way.
Stepping away from HS, imagine a weapon in a game that does 1 to 100 damage vs one that does 41 to 60 damage per hit. They both have the same average damage, but the first has a higher delta of randomness.
I'm neither a mathematician or a game designer, so my apologies if my explanation is lacking.
+LameytheClown I know what Delta means. It's just that they flashed a card and started talking about how its properties affect gameplay without saying what its properties are.
What role does randomness play in game design, and how can it be balanced?
+Extra Credits Hey, good video.
One thing I'd love to see you talk about in your next video is the new "discover" mechanic, to be introduced tomorrow, in the upcoming adventure. How good of an RNG design cards with the "discover" tag are.
Because I really don't know :)
+Extra Credits Why would you lay off that topic for a while? It's incredibly deep and interesting... Most games, especially board- and cardgames are good/bad almost entirely because they knew/didn't know how to implement randomness.
The perception of randomness is another really interesting topic. Mark Rosewater talked about how the Wizards otC StarWars cardgame was perceived badly, because players had to roll with many dice at once for random factors and perceived it as too luck-based, when in fact the opposite is the case.
+Extra Credits You pretty much only talked about the first attraction of randomness which is to lead to exciting moments, but I think the third one is even more important in many games.
Take roguelikes for example. Their whole draw is that you have to learn the _rules_ of the game, not just a specific level layout to win.
Bad roguelikes are often just unfair because your success is dependent on the roll of the dice with no way for you to deal with the randomness. Good roguelikes are careful to give you enough information to be able to avoid bad fortune and take advantage of good fortune.
I think Nethack is a shining example of the latter type. I forgot the player's name, but somebody actually managed to win a run with every class _in a row_, showing that, if you're good enough at the game, you can win every time, regardless of what the game throws at you.
+Extra Credits
Thinking about the differences between WC3 and Starcraft2, and the reasons for I liked the first one much more.
Small randomness vs None.
It keeps the game from being nothing but a numbers game
Now THIS gives me a lot to chew on. I'm going to have to watch this one more than once to take in everything that just flashed past my face. The game development I work on uses a lot of random numbers, and I think a good bit about how to use randomness to make my games more interesting without taking away from the skill that goes into it, so any discussion on randomness is more than welcome to me!
You want random... there is a deck called Randuin Wrynn which pretty much harness the full potential of random(that is... it pretty much has almost every single card with the word random on it)... now that is really fun to play to troll other people...
+caramida9 I love that deck and even though I don't have all of the random legendaries like sylv that is still my favorite deck to play. It feels so good to beat secret paladin with it, especially in casual mode where decks like Randuin belong and not secret paladin
I managed to beat a face hunter with it... just mad bomb all his 1 health minions and take his explosive trap with keyzan mistyc...
+caramida9 I prefer Kirby Air Ride for RNG fun. Watch a playerthat was crushing you grab all the weight upgrades, only to find out that the minigame is highjump XD
i think a card game was a really bad example to use for randomness creating exciting moments, a better example (off the top of my head) is in metal gear online, sometimes a sandstorm will occur, this mixes gameplay up as it forces ALL players to change their playstyle, instead of 1 thing possibly swinging the game in the losers favor which really should NEVER be a thing for any sort of competitive game, its not good. period. (unless there is like a random fun mode, but that distinction has to be there)
i made a board game for a university project and it had a bidding round at the start were players bid money on items that were drawn from a deck of cards that they could use for different actions or whatever, i think THAT was a good example of randomness in a game, as it meant that players could opt in to bid on an item if they want, it allowed different power dynamics to happen during the game, none of the items could force a win and it allowed for interesting different things to happen during the game, plus players could trade items as well. i think randomness is better used when it creates new scenarios and forces all players to adapt to the situation, or loot based randomness as well
and now we have yogg saron who makes every game a coin flip. blizzard should watch your videos..
couldn't help but watch these videos again when thinking about the night in Kharazan expansion. Yogg Saron and Barnes have really taken this game into the wrong direction
blasterjosh I don't mind barnes that much since, unless you build a combo deck for it, the 1/1 isn't that bad.
Soviet Toaster
it is pretty frustating though when you see that combo actually get pulled off
your guys' art direction is getting SO MUCH BETTER.
I really appreciate all the HS related topics. It's extremely interesting to learn about the ways Ben Brode's Dev-Team is creating cards. :)
I understand Hearthstone is a good game to use for these examples but I really do wish you would use other games to at least attempt differing scenarios.
Not everyone quickly picks up card game mechanics. I do because I've played MtG for years, but not everyone.
+Ronyo I play hours of hearth everyday, and I also wouldve like to see where this RNG is involved. Only other competitve game I play is CS so not much RNG there. I doubt another game has the RNG problem Hearthstone has ...
+00TWB00 many games have RNG problems. In Dota2 for example, there is a hero(character whatever you wanna call it) that stuns enemies on every hit for a 25% chance. You can potentially stunlock and kill every enemy with no interaction, but that happens very rarely.
Comeptitive shooter usually only have inaccurate weapons as random element (sometimes random crits, too) and are among the lowest RNG games. Cardgames have the highest random factor, even the drawing of cards itself is pretty impactfull randomness.
Strategy games tend to have something I would call normalised RNG: Many samples of RNG with small impact.(eg. every unit has a 50% chance to deal 10% bonus dmg wich will average 105% dmg of said unit. Can turn the tide in theory, but usually makes no difference.)
+Slam United "Comeptitive shooter usually only have inaccurate weapons as random element"
In CS:GO, the AK47 and Negev are good examples of this.
Yeah getting a little bit tired of the Hearthstone references myself...
"The use of randomness in skill based games"
immediately adds TF2
Once again, you guys make the best videos. I've binged through season 1-5 and I'm still trying to get caught up. I've *liked everyone of them, good work guys. (also really liked the art in this one)
I really appreciate the artist in this series. Great work!
Actually cards like Flame Juggler, Knife Juggler, or Flamewaker are a very poor use of random effects in the game. Hearthstone is different from card games like Magic, because the attacking player has initiative, they decide how to trade their minions, unlike Magic where the defending player can set how creatures interact. This by design makes Hearthstone more snowbally, since the board state will get increasingly better for the player with advantage as more minions are played. When a minion has an ability that has a random chance to affect the board, especially in the early game it can actually change the outcome drastically. For instance if I play Flame Juggler on turn 2 and it kills the other player's 2 drop, I am in a very commanding position on board. The board state in Hearthstone is much more important thing compared to other card games (this is a very minion focused game). The delta is actually very high on whether flame juggler hits the minion you want or not, it is not simply a matter of 1 damage, that 1 damage in the early game is whether a minion dies or not. An example of good randomness in the game is something like Undercity Huckster from the most recent expansion Whispers of the Old Gods. That card is great because it is slightly below the power curve and provides a card that still needs to be played from the hand, it also doesn't effect the early game board with this effect. I disagree with your evaluation on Spellslinger a little, yes I think its range of cards it can pull is too much causing for a higher delta of randomness, but it usually isn't the actual reason for the outcome of the game, usually that was decided much earlier than that by a card that affected the early game board such as Flamewaker or Flame Juggler. It sometimes isn't as apparent that a 2 or 3 mana play decided a game that goes to turn 8 or so, but that level of snowballing of the board really does have a profound impact.
I'd like to bring up the recent Tales of Zestiria game in addition to randomness. Namely that, in the new game, weapons and armor come with certain inherent abilities; but only the first ability it set. As a result, in order to get the better abilities the player not only must play on the harder modes but, even then, there is far too likely a piece of gear will drop that simply doesn't have the right skill or has it in the wrong slot rendering the whole thing pointless.
I've been waiting all week for Extra History. I hope you guys release an episode today or I die
Fantastic video! Many people are either way off in the no randomness camp, and just as many want it everywhere. I think this video did a good job showing how randomness can add to a game (and sometimes even raise the skill cap)
This was a chance to talk about something other than Hearthstone...Pokemon WiFi battles for example. Attacks have Accuracy, Power and a chance of a secondary effect. Scald does 80 damage, has 100% accuracy and has a 30% chance to burn. Whereas, something like Will-O-Wisp guarantees the burn, but only has an 85% accuracy. A move comparable to Scald; Tri Attack does the same damage, same accuracy (less PP)- it only has a 20% chance of a secondary effect, but the opponent cannot make a switch in of a fire type expecting the burn since it may equally be a freeze or a paralysis. Predictable, and reliable status infliction therefore has some sort of balance in terms of switch potential compared to an attack with a secondary effect that may trigger.
+Mr Sebasculin You can talk about pretty much anything, such as axes in rpgs (More damage but possibility of less damage)
While yes this was a chance to not talk about hearthstone pokemon may not be the best example as rng never makes both players excited at the same time
+Mr Sebasculin The interesting thing about PvP in Pokemon is that you have the power to swing RNG in your favor (to a certain extent). Having status effects on the opposing side, Pokemon with the Skill Link ability or any ability that raises Critical Hit ratio, and so on.
7chaoemerald Like any RPG? .-.
+tjtj012 Unless we're talking Metronome or Gift.
It'd be nice to hear how randomness effects design in some contexts other than hearthstone. Binding of Isaac for example or any other PvE design scenario. Talking about anything in depth in these short videos is a challenge I'm sure.
I could watch videos about game design fundamentals like this for hours.
The art style in your vids is AMAZING!
Digital Extremes should watch some of this.
+Marcelo Cabral Lima still looking for that Bo Prime Handle XD
+Marcelo Cabral Lima I really feel you there.
I hope starchart 3.0 will bring good Rng.
True. Though Bungie could use it more.
***** Somone call the police, because shots have been fired!
I surprised you didn't bring up Darkest Dungeon. That game pretty much screams RNG
That's exactly the episode I wanted. My ultimate objective would be a DnD-like video game with random d20 roll, and the random element is what's giving me the most trouble in that big years-long "pre-production" phase I'm currently in (as in I'm refining the game the best I can while I'm still studying so that I'd be immediately ready to start the project once I get out of game design school). Videos like these help me know exactly what I should do to make the random element engaging rather than frustrating, which is what I strive for.
+Ryoga Hibiki How did you find this board? You were trying for Furinkan high and got lost I bet...
lmao! when that picture came up at 4:37 with the card saying "fight me IRL" I lost it.
never laughed so hard during a youtube video. spoke to me as a HS player
Now I'm curious here. You all over at Extra credits seem like very positive and knowledgable people when it comes to history, game design and gaming culture. My only query so far on your game design segment is: Are there any mechanics that you can list that could be considered wholey negative and bad in a game experience? I'm talking blatent mechanics, and not story, art or character design, and this isn't something that can be corrected, only removed, so no saying " bad 'X'" or " a messed up version of 'Y' ", it has to be bad on its own.
i'd love to have this explained in a context i understand. because hearthstone and TCG's are not my cuppa
we cant explain it in a context suitable for you if you don't specify which games you play...
First time I came across your videos, cool video, neat explanations, chill dude. Keep it up! :)
Scott had some amazing artwork in this vid, nice job Scott!
Speaking for every Warframe player: Screw you RNGesus.
Randomness is not inherently bad for the player, because they control wether or not to play the card.
Let's say there's a spell: 50% chance of winning the game, 50% of losing it. That would be absolutely amazing and auto-include, even though the average result is null. That's because you control wether or not to play that random effect. If you're losing, it's great! If you're winning, you dont play it.
The non-random counter part to this card is: Do nothing. It's obviously terrible. Another shitty version of this card is: when you draw it, 50% of winning, 50% of losing. This becomes terrible because you lose control of when to play it.
RNG can be desirable when you're behind, and isn't inherently bad. Though you'd be right to say it's bad when the game is even.
+Steam1901 In a vacuum, you'd be right, but there are other rules and things to consider. I dunno about hearthstone, but in MTG, you're only allowed to have 60 cards in your deck, and 7 cards in your hand. So, pulling that spell in your opening hand, then having it sitting around, taking up a slot and waiting for a turn where you're way behind and the flip would be worth it, would be really bad because that slot could be taken up by something that could win you the game more consistently through careful planning and good plays.
+ilovethelegend Well it would for sure, by definition, ensure a 50% win-loss ratio, which is average. Then, every game you win while being encumbered by this card in your hand would count towards an above 50% win rate, which is above average.
Thus for most people (more than 50%) it's a good inclusion, even though a pro player might choose to not include it when playing against people of a lesser skill level.
+ilovethelegend HS uses 30 card decks, so the card Steam suggests is even worse. But as Steam says, ensuring that at minimum you have a 50% chance of winning is great for new player. However, a pro-player would want a more synergized deck as they would not need that minimum chance. Steam also forgets that you do have to draw the card, and in HS most games do not make it to the bottom of the deck and there is no tutor cards.
No, it wouldn't. What if you flip wrong two games in a row, or have games where you don't even pull it? A card like that would be nothing but dead weight.
But that's not the crux of what we were saying anyway. The point is, sure, you have control over when you can play a card with a random effect, but if it's only worth playing in a very specific scenario, then its lower delta is even worse than "it does nothing".
When you dont draw it you dont draw it. It's not a dead weight if you dont draw it... It's worth playing in any scenario where your chances of winning are bellow 50%, which is not very specific at all.
It's delta is litterally the biggest delta that can be, as the up side is the best effect in the game and the downside is the worst effect in the game.
This card would see SOOOO much play. At the strict minimum, it would see play in 50% of the decks. If your win ratio is bellow 50%, which is is for 50% of the player, you should, mathematically, put that card in your deck. This is actually a fact.
I already love this series of episodes.
Props to who ever animated this, i love the faces
Good points, unfortunately these days, "Fairness and Skill" is not what is on developers and companies' minds. They design their RNG systems to psychologically torment you until you put down a pretty penny, or you spend hours grinding.
Destiny and Halo 5 are the perfect examples, 2 otherwise good games, but ones that are creating a very ugly trend in AAA games.
This mostly accounts for multiplayer gameplay, but what about singleplayer rng balancing?
+Cheetah Scar for strictly single player games, you generally want everything good to have above average odds of happening. as said in this video, the feeling of getting RNG to give you something super awesome is an exciting moment, but only for the player that got the thing. in a single player game, the opponent is a computer, and thus, wont get angry and rage over a lucky roll. so you should be far nicer with RNG to the player, and at the same time, make the game a bit more difficult to balance that.
i mean, im no designer, so i could be very wrong, but i know i for one would NOT play Monster Hunter without friends, unless they change the RNG completely.
+Cheetah Scar For a single-player game, you pretty much need to stack the odds in the player's favour in any individual encounter, simply because the player faces hundreds of encounters while each enemy only faces one. Let's take a simple game - 10 levels, and in each level, you have to fight 10 enemies. The combat system is set up so that in each encounter, you have a 1% chance of being killed, and a 99% chance of getting past (in which case you recover fully before the next fight). If you run the numbers, your chances of winning all 100 encounters in a single run are 0.99^100 or 36.6%. In other words, even when you only lose an individual fight 1% of the time, when you're facing 100 fights, you'll lose almost two thirds of the time.
A good way to think about it is to assume that the RNG will turn on the player and savage them for a while - having a streak of worse-than-average luck lasting the square root of the length of the game would be typical (assuming I remember the relevant results correctly - it's been a while since I last looked at this sort of maths, and my Google-fu isn't finding good enough references quickly enough to be useful...).
Of course, there are games where it's acceptable for RNG to wreck a player completely, so it's not the end of the world if bad things happen - just be aware that Murphy's Law is a fact of life if someone plays for long enough
+Cheetah Scar
tldr: You can grind for a success state with random events in single player games so the chance of success should be higher.
One thing to note is that each individual outcome of randomness in a multiplayer game tends to be significant whereas this is not the case in single player games.
Multiplayer example: You play spellslinger in hearthstone and now each player has the extra spell and the game moves forward without any way of going back.
Single Player example: You attempt to pickpocket someone in Skyrim. If the random roll succeeds you save your game and move on to the next person having acquired the stolen item. If the random roll fails, you reload to the save a few seconds earlier and try again.
Randomness in single player games (especially rpgs) often degrades to being just a time investment. This happens for stealing systems, percent chance drops for items, etc. It often can end up like this for combat encounters too which can lead to players being frustrated if the chance of failure is high. Thus (as justdonttask3 mentioned) you tend to want the chance for success to be higher than you would have in a multiplayer game.
Personal note: The way I see it, the only real way to make random systems interesting in single player games without it turning into "a grind" is to remove the players ability to save and reload states before the random event (something like a hardcore mode, although most games with hardcore modes tend to not have randomness as it sucks to lose a character you just put a lot of time into to a bad dice roll).
It's interesting to note how some players will impose this restriction on themselves just to try and give significance to randomness systems. An example of this is nuzlocke runs for pokemon which imposes 2 extra rules:
Rule #1. You can only capture the first Pokemon you encounter in each new area you explore.
Rule #2. If one of your Pokemon faints, you must release it. It is dead.
+justdontask3 I suspect it depends on the player though. Using a multiplayer game to make the point, but I have about a 20% win ratio in Wargame Red Dragon and I love it *because* it is challenging. So long as skill can potentially compensate for the RNG, you don't necessarily need to stack RNG in the player's favor, just give them a skill-based way to overcome it, adding to the sense of danger.
Again I am using a multiplayer game, but look at Armello. The chances of getting killed in a battle are far higher than the chances of surviving right off the bat with equal stats. 3 out of 6 die faces cause damage, only the shield protects you. However, a player can compensate by taking items that grant the "evade" ability (instead of hits, you gain defense for all dice, such that it is impossible to cause damage but your chances of survival are high, a second tradeoff being that you must give up your hexagonal space in either case). Alternatively, the "Sun Shield" item allows two die faces that tradeoff being counted as hits and misses depending on the turn phase (sun and moon faces) to instead treat the one that would have done nothing (e.g. a sun symbol during a night phase) to grant defense instead of nothing. There is armor too, but it gives a flat bonus. If the tree symbol (for normal creatures) shows up, you get another die per tree stacking ad infinituum such that once a player has a lot of attack, the advantage will be slightly exponential, rather than purely linear.
However, the effects on the early game of Armello is to make battles exciting. The weaker player is almost guaranteed to die in a battle, but they might be able to take their attacker down with them due to luck.
In a game where RNG is against the player, the player simply needs to stock up to counter the anticipated misfortunes. This would actually be pretty enticing for a survival-based game. It is just important to note that there need to be few "critical failures" among them. The effects can be biased towards disadvantageous, but not disastrous. A bias towards disastrous only makes the game pointless, not challenging.
I disagree look at rouglikes but the key to negative outcomes of random effects in single player is that they should be mitigate able. Usually by skill and attention is best but occasionally needing to have an positive random tool saved is also acceptable.
This episode was sweet. You guys managed to articulate how I feel about randomness in HS short and sweet, in a 7 minute video. It's not an inherently bad thing, it just needs to be carefully tempered by design.
I keep coming back to EC for amazing articles like this. Bravo!
Note to EC. Some of us don't play Hearthstone. Throwing random card names out there does nothing for us.
+Robotfan987 I've never played Hearthstone, but they explain how the cards work in the video. It's perfectly fine.
Good thing that they're showing the cards and explaining what it does
+Robotfan987 Not everyone plays every game. You can look up all of the cards on the internet if you want to know what they do and why they're good/bad.
+Robotfan987 Try watching the video when you start it. It'll help.
+Robotfan987 if you are watching this for entertainment, its not important that you understand everything. and if you are watching this for the education factor... you should already know what hearthstone is.
"Random critical hits are fair and balanced." -no one ever
Even though I barely play it, I like that they use hearthstone for so many examples. Makes it really easy to follow along.
I love your animations or progressive drawings the personification you put on the objects make the videos somewhat funny. Keep it up and yes I kniw this video is possibly several years old but still I love your content.
Why isnt this a design club episode?
Where is the next design club episode?
Hearthstone should be a design club episode
Talk about Mario Maker?
You mean the game where Nintendo went "fuck it, we can't be bothered to design levels, so let's throw out something that's on the same level as a flash game and sell it for $60". I actually have played flash games with better level editors.
+Willie Barcott "Can't be bothered to design levels" - In the Mario titles between just 2010-2015 they created like 400-500 levels for the series. Good level design is kind of Mario's thing, you can't really talk like they've been slacking.
Granted it's a shame the NSMB games are lazy in terms of aesthetics but the actual designs and level hazards and gimmicks have remained varied and consistently fun.
JezMM Im not saying they were lazy before mario maker, and im not complaining about NSMB. THat was quite the straw-man. I'm questioning their deicion to give up on that level design and release a flash-game quality game for $60
Edit: clarification
Willie Barcott Fair enough, I enjoy Mario Maker but can definitely see where it's lacking compared to some fan-made level creators. Apologies for misinterpreting your arguement, pretty much everyone I've seen so far who has accused Mario Maker of lazy have been the same people who accuse the Mario series as being lazy/uninventive since Galaxy (while simultaneously wanting them to "just make Galaxy 3" lol).
wow this episode was solid. great job.
Your HearthStone videos are the best! Keep it up :3
Extra Credits isn't as enjoyable to listen to now since half of the videos are pretty much just talking about mechanics of Hearthstone.
So are you sticking with hearthstone because you're being sponsored or is there some goal here?
+tjtj012 Problably because Hearstone it easier to show the delta of randomness. Like Dan said, you can see, very easly, the RNG happening.
+mestre12 Still it's three weeks in a row of hearthstone and next week is also going to have it.
+mestre12 Never mind I just looked up the videos and it hasn't been in a row but they are noticibly close together
+tjtj012 Basically I think they've been playing hearthstone recently so that's what the topic is. Plus it is, for better or worse, one of the most popular digital card game around that provides some very clear examples to demonstrate the concepts they're talking about.
+Killer_Tapir They could still at least try to use other games for the topics
I'm pretty sure James has been on a *huge* Hearthstone binge for the last month or two, between the few EC episodes, and the Extra Play series.
You guys should talk about randomness in board game design :). I've made a few myself and it seems like that same dichotomy exists, but even more so with dice-based combat.
If you enjoyed today's episode, come watch James play Hearthstone on #ExtraPlay! He's drafted a mage deck: bit.ly/1NLVMpU
+Extra Credits I think this is a lot more applicable to Warhammer 40k than harthstone... as anyone who plays Orks can tell you.
+Extra Credits Please stop advertising for Blizzard. Card Hunter is another game that has random features that are controlable and it isnt a AAA game to worship. Your sounding like a COD fan where every conversation is relivant to the game.
+oathblade because this is a game that so many people know so it's easier to get the point across to the viewers? And maybe they just enjoy playing Hearthstone? If you're trying to use a sports analogy with people, do you think everyone will be more familiar with the rules of golf or Futbol?
Have you ever played payday it has randomness to change guard spawns and where different loot spawns as well as other things like camera spawns and even where the main objective is this means you can play the same mission 100 times and it still feels like a new interesting mission.
+Extra Credits I'd love to hear what you have to say about RNG determined rewards (does Hearthstone have that? "Congrats for the game, here's a pack of cards"?). And especially weighted RNG ... (looking at you, +PlayWarframe)
please stop talking about hearthstone...there are enough videos about it, and most uf us dont play it
i dont, that is the point
+Filipe Brizolara the video isn't about hearthstone, the things they're talking about can be applied to any game, Hearthstone is just the example, and it provides some very good examples that are easy enough to understand for non players. It also has a number of mistakes that show what happens when you get the balance wrong.
Even though, that doesn't change the fact that this game help fairly easy to explain this topic compared with other more complex games like Magic The gathering, Yugiho or Pokemon trading card. The importance in the episode is the topic, not the game used as an example.
+deadlydemon666 And using those games would mean that they reach a smaller audience. Hearthstone is very popular and for most people, easy to get the grips on - thus making it an excellent tool for examples
My problem is that they have made no effort to take the topic out of the realm of card games let alone Hearthstone in particular this is bad form and not up to the usual quality of an extra credits episode not devoted to analyzing a specific game. And many single game analysis videos compare to other games the problem is that they are limiting a really cool topic to such an narrow band and ignoring how to use it in any other type of game.
I love these vids, learning so much about game design.
... so do you already have a plan for what you are going to cover after this series of randomness is done, or are you going to pick... at random?.... I'll go back to under my rock now.
thanx for this one guys. sheds some light
I like the discussions on randomness, so I'd welcome more!
The art in this episode is hilarious and awesome
oh hey, this ep right when I'm starting a roguelite project :D
As always, I love the animations!
Oh my god I love this animator, the facial expressions are so wonderful.
Who did the art on this one? It was fantastic! Almost like animation in so many places! My sincerest compliments!
Well that is an incredibly one sided way to look at randomness. There is the other side of that coin:
- It creates exciting moments/It also creates frustrating moments
- It allows the worse player to win/It also allows the better players to lose
- It forces players to adapt/ It hinders players from planning
Whatever randomness gives, it also takes away from somewhere else....
I love your hearthstone discussions.
Interesting episode as usual. :) See you next week.
Thank you for the video!
Could you sometime in the future go into specifics on percentage-based accuracy RNG games? Like XCOM or Blood Bowl
5:58 I would use the example of counter strike's bullet spread, where you can effect the amount of spread by how fast you are moving, or in pokemon where the critical hit and accuracy ratio are a stat that can be changed in play
Pretty neat discussion as usual. :D
I don't know why but the board flip image is soooo adorable where the girl looks like she is holding the card in her mouth like in an anime.....
And that image of the stronger player with a silver trophy is now my steam avatar. Thank you so very much.
This actually reminds me of a game I played called Wizards101. I mained a storm wizards (basically I hit the hardest, but I died the easiest and I was more likely to fail at casting. The low health is what really evened the school out because mobs could one hit us as well back in the day). There was one card that only storm wizards could use in the first arc of the game, before they came out with the second and third arc. This card would deal 1000 base damage, which was a lot in the first arc! No other card came close to it, and it would always one hit a mob if you had boosts on. The thing is I think this was part of the "randomness" you talked about, because this card also had the highest chance to fail, fizzle, with a 10% cast rate. Most storm cards had a 70%, and while they were all strong, none of them could top that one card. But because it only had a 10% chance of success, you rarely saw it in play. It was a last ditch effort to win the boss fight when you and your teammates were out of attacks and heals and reshuffles.
Of course, as the second arc came out they changed the card so it was more usuable. With the second arc came stronger cards for the higher levels. If they wanted to keep that card in play for a bit longer it had to change, since it was a card that you could only get late into the first arc. Yet they still kept the randomness with it, by chancing that you might get 10, 100, or 1000 damage. They also raised the success rate to 70%
Meanwhile, sniper from X-COM: "...wait. Did command say they want me to shoot that berserker, who is currently tearing ass through our squad..., or that potted cactus? Man, that cactus sure looks menacing, I better take it out first."
MOAR HS VIDEOS PLS! :D
Love you guys
I like the new outro. It's nice.
I love those very specific videos
Good call with the Casino Night Zone theme from Sonic 2
I really like the randomness discussion. I'm working on a paper and pencil RPG system (Looking for info on design is how I found the channel) and one of the things I first set out to do was to lower the randomness curve. In D&D it's highly unlikely but possible for a house cat to over power a storm giant in a strength contest. I've gone through several ideas but I've finally settled on a proficiency/potential system. Each stat is represented by two parameters. Potential is how much raw power you have and determines your dice size, proficiency is a measure of intellect and intuition with that stat so it brings up your minimum roll. So if you had 3/1d10 strength you would roll a 1d10 and add 3 but you're capped at the maximum result of 10. Proficiency is capped at half your Potential so there will always be some chance involved.
+Malcom Chase True. It's still silly results. In practice the system is pretty simple to use.
+Malcom Chase In D&D absolutely but that's far from the only thing I'm doing differently. Just happened to be relevant to the video.
So I play World of Tanks and Armored Warfare. WoT has a 25% up and down range for it's random elements of penetration power and damage but Armored Warfare only has a 15%. The difference here is AW is completely random while world of tanks has stable ranges between it's random chance most often you will do within 10% of your penetration average and damage average per shot. While AW you can do up to 15% above or below but also the damage is multiplied based upon critical hits (WoT doesn't have this at all). Which set of Randomness do you believe is a bit better in this case?
I missed the comedy skits you used to slip into episodes.
Also, still waiting for Designing for Touch Screens Part 2.
5:41 I was half-expecting some visual puns there. X )
I would like to say that I do like your points on randomness in a game I feel you full short on talking about frequency of randomness this has one of the biggest effects on the game play!
when design a random element that has a high frequency that the ability/mechanic is activated or used lessness the random chances witch increases the power of the element.
Arcane Missiles was a good example of randomness done right, and it was one of the first cards released.
There's also what League of Legends did. They called it adaptive or scaling randomness, but what it does is when a player with %crit chance doesn't score a crit their crit chance goes up a little and when they do land a crit it goes down. This makes the chance of getting a long strings of crits or not crits much less likely because getting several crits in a row makes your crit chance very small and vise versa.
I'm looking forward to the "Extra Hearthstone" re-branding soon! /s
Netrunner (the card game) is probably my favorite example of randomness in games. It is full of moments where your opponent needs one more point to win and they access your hand which you know has points in it. The best thing about it is, is that you have to keep a complete poker face so they dont know you have points in your hand while you are sweating bullets on the inside.
I know you guys dont tend to do tabletop games but I recommend at least checking out netrunner because the design is fascinating and completely different to how more traditional card games like magic and hearthstone function.
for those of you who never played it CasinoStone is really fun if you dont mind the occasional PJSalt and BabyRage
Hearthstone is such a good example for explaining RNG like this. Can't wait for the next video!
I already have designed a card game , but before I allow it to become a public offering I'd like to have a definite system to determine any cards strength so how do I make a power system exactly?
that card with sunglasses is extremely adorable :3
>really in depth
>7 minute video
yeaaaaaaaah
All I was think of during this video was: Time Wizard so tempting but could destroy everything you set out to do. Few cards were going to help you comeback from a Time Wizard or catch up if Time Wizard did't go your way.
Lots of Hearthstone stuff lately if you count the gameplay channel and this one. I don't see anything wrong with that but I will admit to a little confusion because I have not played Hearthstone but James's gameplay videos helped a lot
I really hope the next episode discusses the methods of and inherent flaws in how randomness can be implemented. Ie. the differences between straight-up RNG call (any result in range can come up, even if it creates a sequence or trend), an arbitration (new results are chosen relative to previous results to prevent sequences or trends), an exclusionary call (previous results cannot pop up at all in repeat rolls), and so forth. It's a subject that, on a technical level, many developers just plain do not understand. It doesn't help that western culture mistakenly equates randomness with arbitration, either.
This episode reminds me of this one MTG pro tour match, where someone was playing Jeskai control and the other was playing splinter twin combo. The splinter twin guy had a deceiver exarch out, and the jeskai control just needed one more attack to give him a win.
All he needed was a splinter twin, and...He top decked it! Then his opponent cast wear//tear.Sad, but a good example of randomness.
This video is just in time with the launch of the next adventure wing that introduces the keyword Discovery, which seems to address a possible way to shrink said delta of randomness. Perhaps the followup video can touch on this new mechanic.
Kindof like the Ultra-autocannons in Mechwarrior online, you can fire it more often, but you run the chance of jamming if you do.
Number 85: Crazy Box ... outskilled, outplayed XD
I'd love to see an episode about using RNG as part of an in-game rewards system, and how to balance that to make it feel rewarding instead of feeling like an arbitrary grind to lengthen play time. Because lord knows that's how most games with RNG-based reward systems feel, never get anything worth a damn out of them.
Mahjong is a good example of a traditional game with randomness. The player can improve their hands tile by tile and choices about whether I should make it a 20% chance of winning 12000 (risking losing 4000 or sometihng) or 70% chance of winning 1500 constantly creates these moments constantly. While it is possible to win without even drawing a tile, these are so rare that getting rewarded for just the lucky initial draws feels fair.
So, I can see how too high a delta would be a problem, but I would assume (not having any real knowledge on the subject) that it is also possible for the delta to be too small to be meaningful, or just situations where it would be an improvement if it was higher, even though it is not currently zero. So, I assume that there is (for a given game and target demographic) an optimal value for the delta of a particular action. (holding all other aspects of the game equal that is. I suppose there might be complicated interactions if the delta of multiple things is changed?)
I really enjoyed this episode, and look forwards to the follow up one!
This feels like something that would actually be applicable for an individual making games as a hobby, which I like.
I assume that this does not only apply to TCGs or other similar genres, and that there would be at least an analogue of the power curve for those things as well?
Again, I really liked this episode. Thanks!