How much RNG is Too Much, Inspired by Priestly


The original thread for this was locked due to getting derailed by criticism of one poster’s posting style, but I think it’s actually a pretty important question: how can you determine how much RNG is “too much”?

Continuing the discussion from Counterplay, is this RNG-style design a permanent fixture?:

I swung back and forth over and over as I read the original thread, but I had an insight as I re-read this post for the fifth time: while this kind of analysis is a good starting point, it misses the mark. That’s because, as you get higher and higher in skill level, of course the games are going to be determined less and less by differences in skill or significant mistakes…which means of course the games are going to come down to RNG elements. The same would be true if there were zero RNG cards, however: the RNG would come down to who drew the better card for the situation on-board.

The net effect of RNG might be slightly larger, but realistically, at high levels (which probably means ‘Diamond or higher’, looking at the replays) the likelihood that any one game is going to be determined by draws even in the ideal zero-RNG world is extremely high, since mistakes are rare. (Or, even more confounding, the games might be determined by deckbuilding skill, which makes determining RNG factors by watching replays all but impossible.)

In general, RNG effects favor the less skilled player, which is precisely why they aren’t bad for the game as a whole…and exactly why people who are decent or better at the game complain about them the most vociferously. :wink:


I remember watching the first Battle Royale (a classic in my book) movie for the first time and thinking ‘Why do the teachers give the students such vastly different weapons in terms of power level?’ Most kids get like a crossbow, or a knife, maybe a utility item like a GPS, but then other kids get freaking firearms! All based on the luck of the draw. The most likely winners of each Battle Royale was decided by complete random chance before the match even started, negating many of the skills individual students might have had to win. You might be good at sneaking, but your odds against someone with an Uzi are still pretty crappy.

We like these circumstances in movies because we like rooting for the underdog, especially when they’re faced with unfair disadvantage. But in an actual game of skill involving ourselves it’s no thrill to lose or win a match based on these random events, if you’re a skill player that is. If you’re in it for the cool moments and upsets you might enjoy them more. The point is: if you make a game targeting skill players, then include elements that work against skill play you’re going to rustle some jimmies. Finding the sweet spot for a broader demographic that still respects skill play is the real trick imo.

Where that sweet spot is continues to be a nuanced issue, as it will probably alway be.


That would be a pretty valid concept, but for two issues:

  1. The very nature of a CCG (in which RNG literally decides what ‘weapons’ you have available to build your deck with) makes skill-based play meaningless until you approach a full collection, and

  2. Once you do approach a full collection, you have the option of using the ‘weapon’ of your choice, meaning you get to elect whether or not you want to roll with a few high-swing RNG “comeback” cads in your deck, or instead roll with consistent power cards on the assumption that you will generally be winning and need to secure the victory.

In the first case, “skill players” should be aware of the fact that skill is going to be hampered significantly for quite some time (if you’re F2P), and might just choose to avoid the TCG genre altogether. In the second case, the skill comes into play during the deckbuilding portion, deciding whether or not you’re close enough to the top of your ZPD to need a good RNG comeback mechanic to increase your winrate and claw your way up to the next Division break.

Either way, the point – that at high skill levels, RNG will almost always decide the victor whether or not it comes from a specific card or from the nature of TCGs themselves – is kind of missed by your analogy.


Well just as a counterpoint, I am mediocre at best at the game and I want as little RNG as possible. Why? For the exact reason you mentioned. How am I going to get better if I can win close to half my games by luck?

Your post is well-written and reasonable. It basically makes the argument that “we already have RNG so it doesn’t matter if we have more”, which I tried to address in the prior thread. This is mostly a matter of opinion but as I see it, how is more of it a good thing? If I have a chance to lose based on a bad card draw, that’s better than having a chance to lose on a bad card draw OR a good card draw + bad RNG effects for me / good RNG effects by my opponent.

More simply put: we already have some RNG (this isn’t chess) so why do we need ever-increasing amounts of it on top?



  1. It appeals to a (not insignificant) part of the playerbase,
  2. It allows more people to reach a higher rank than they otherwise would, which adds to 1),
  3. It allows for more dramatic moments in-game, which adds to the excitement of watching streams and vids, which adds to both the word-of-mouth spread of the game and the degree of immersion it’s possible to achieve (I can’t, for example, play while I cook breakfast, but I often watch Mogwai when I cook, and there would be a lot less people like Mogwai if the game wasn’t as dramatic.)

From Counterplay’s standpoint, RNG is a good thing. The fact that beyond a certain skill level it’s an annoyance is…not all that important, really.


Unless I misread it, your argument boils down to “good/veteran players hate this but it’s good for business so Counterplay is doing it anyway”? Irrespective of the merits of that argument, it’s pretty cynical, don’t you think?

As for the argument itself, I think it’s much more dramatic to see someone come back and beat the opponent based on skillful play than good dice. Just yesterday I got beaten by someone who made some superb moves I never saw coming. I gave him an honest “Well Played”, he deserved it. If he had just topdecked an easy lethal, that would have been considerably less dramatic IMO.


Well, he is right though. Cynical or not, RNG increases the accessibility (how dafaq does one spell this word ;_:wink: of a game. The subjective amount of rng someone wants differs wildly, but the silent majority of players play videogames casually. It’s just a fact that easier games are more popular though. And RNG adds to simplicity.

jesus christ, all that talking about RNG and simplicity makes me want to play dark souls again lol.


And yet there is this constant claim that people are leaving HS because of too much RNG. Many people who play Duelyst say this. to what extent this is actually happening I do not know, but clearly there’s a tipping point.


Thanks for continuing this thread. I didn’t get a chance to get in on the action in the first one.

I used to think I hated RNG. As someone who attempted to transition from MTG to Hearthstone, I got a very bad taste in my mouth for it.

However, I have come to the conclusion that there are types of RNG that I love and types that I don’t care for so much.

For example, I love drawing cards from my deck. Even though there is RNG here (as in any card game), it teaches me something about my deckbuilding skills every time I draw a card. Why? Because while the card is randomly selected, it is from my deck and thus I’m the one who chose to put it in there. This is i guess “probabilistic RNG”. While I have no idea which card I’ll draw, I know what I can draw and therefore I can plan.

Another form of RNG I like is one’s that allow certain cards to exist. Take Lightning Blitz for example. If it were “choose where each minion goes on the opponent’s side of the battlefield” that would not only be broken but it would be tedious. RNG that reduces tedium is good. We (I at least) want this game to be more like chess. I do not, however, want it to be more like Risk. Reducing tedium is a good use of RNG. Can Lightning Blitz make or break a game based on where each unit falls? Yes. But not likely.

Forms of RNG I don’t like are one’s that pull in cards from outside your deck (Grincher, Lkian, Reaper, Nature’s Confluence, etc.). I don’t think there are many of these that are a problem, however, since with few exceptions they are either balanced or underpowered. Gaining a random card that you didn’t plan to have reduces the “fun of deckbuilding” and thus I’ve never used them. I do wish that they didn’t exist, just because I like fighting against a deck of cards that a player designed and planned rather than cards he “rolled”. It’s a personal thing though. I don’t think the game is broken by having these cards.

Also, as is the case for the battle pet “gain a random” mechanic, the range of RNG cards should be tighter. Nature’s Confluence for instance can spawn for minions that can’t move…or 4 GROs. This is the type of huge swings that I don’t find very interesting. Although…as others have mentioned, watching streamers such as Mogwai react to the outcomes of such cards is indeed entertaining.

All this to say, I find RNG uninteresting as a deck building option so I never use it myself. However, there are times when RNG has its place. And while I did lose to my friend’s getting a Sky Phalanx from a L’kian (which I knew he had none of in his deck) recently, I would say that excluding the innate RNG in card draw, I only lose about 1-2% of my games to RNG (currently rank 9).


Confluence could IMO be greatly improved while reducing the RNG swings simply by having it always summon 4 different battle pets, or at least making the selection random per slot rather than one random selection x4.


In my opinion, that would make the card weaker (as the opponent could remove 1-2 of them and leave the junk ones alone). It wouldn’t, however, change the “best roll” or “worst roll” (though it would reduce the chance of rolling them). So I’d only vote for this if Confluence becomes broken (too soon to tell, but my hunch is it won’t). Otherwise, no sense in making a non-broken card worse.

When in doubt, the status quo is the way to go.


It doesn’t make the card any weaker on average. It simply reduces the standard deviation of the random results to make them more consistent.

This would allow the actual power of the card to be more reasonably assessed, which is impossible now because of the huge swings (“4 Gros” vs. “4 Roks”.)


So let’s the take the center case: 2 Gros and 2 Roks. Most players will deal with the Gros and ignore the Roks. I can see your point. The odds of there being something the opponent has to deal with (which for 5 mana, there should be), which actually makes the card more reliable.

Point accepted.


I want to reiterate something I’ve posted in the previous thread. IMO, one of the key skills for the CCG players is RNG manipulation. It means calculating the possible outcomes in your head wheighing them against each other and choosing which play is preferable in the current situation. This requires knowledge of the overall card library, common deck archetypes and their counters, tracking your opponent’s plays to figure out what cards they may have left and many other things. In high-level play, when games are unaffected by the mistakes and straight out bad plays (like vomiting out your hand of two-drops on a turn your opponent might cast a Plasma Storm), RNG manipulation skills become key to winning games.

Thus, a certain degree of randomness within the cards themselves does not ruin the game for skillful players. They know the possible outcomes and can determine which one will be the most devastating in the current situation. With that they can play around it, at least to a degree. Of course, there are fringe scenarios where RNG tilts the game in favor of one of the players so heavily the other has no way to retaliate. But those cases are unavoidable in any CCG, regardless of how random the cards themselves are.

But there is also another case. When RNG is too wild and unpredictable it can lead to a lot of bullshit. The prime example of that would be Unstable Portal. There is no way you can play around Unstable Portal since the variation is way too wild. In such cases, the cards simply add more factors that can decide the outcome of the match based on pure luck, becoming detrimental to the game as a whole.

So far I can see no huge problems with RNG cards in Duelyst. Even the hated Reaper has a reasonably narrow range of variables and you can prepare yourself for the worst outcome. I actually think that it is not the RNG that makes that card so opressive - it is the tremendous value it generates, almost like a Zen’rui which can target anything. Another card that has caused a lot of rucus is Grincher, but it is actually pretty easy to play around: if you stay close to your enemy and replace for ways to remove the artifact you can reasonably deal with it without too much of an issue. Also, posititioning around Ankh is a good idea. Other RNG cards also have absolutely reasonable variablity. You can expect them, you can prepare for them, you can adapt to each possible outcome. Even if your NC rolls Roks you can still use them as an obstacle for your opponent.

The long and short of it - RNG in Duelyst is fine, you just need to learn to manipulate it in your favor and accept that fringe cases of bullshit are unavoidable in any card game. Analyze your plays carefully, anticipate every possible outcome and learn not to rely on dice rolls to win you games.


Finally, someone who appreciate the nuance (or rather, the un-nuanced use) of the “RNG” meme…I share all your concerns (heaven forbid card/ deck-milling is ever introduced into Duelyst shudder) and add my conclusions as follows:

That last thread got de-railed/locked because posters got “triggered” by each other into a cyclical feedback of knee-jerk reactions and refused to accept/understand what each other was talking about in terms of “RNG”. So I’ll use TCGs with a completely different system as an example, so that I won’t trigger anyone:

In TCGs like Magic/ Hex/ etc. where the mana/resource needed for playing “costed” minion/spell/etc. cards must also be drawn and played, players must accept the possibility of mana-flood (only drawing resource/mana cards) or mana-screw (not drawing enough mana/resource cards)-- which causes clueless players to instantly fall back on the “(too much) RNG” meme.

But the inclusion of the no. and type of resources/mana cards is part of the deck-building process under the player’s control, which actually adds skill/complexity into planning of the mana/cost-curve of costed cards. With auto-mana/resource generation (and without other game mechanics) the game could easily be reduced to “deck-building + top-decking” because you can largely plan for the limited scenarios that can happen in-game.

This is where I pause and ask any reader here to remember (and ACCEPT) that there is NOTHING inherently “good/bad” about RNG-- it’s all about the differences and nuances in vision and execution in each particular game/instance

Hearthstone went heavy/ crazy with RNG-effects in their cards because their deck has only 30 cards and their board has only 7 minions a side-- so there is literally nothing else to prevent the game from becoming stale/solved as they are determined to keep up the tempo and brevity of their game.

OTOH, Hex with a deck and board of unlimited size (your PC capability is the limit) has precise card text multi-paragraphs long because on top of deck-building and mana/ resource-generation planning, it has already given players plenty to do with multi-phase turn-taking (player declare attack, opponent declare defense, etc.-- not to mention “interrupt” cards available to both sides) as well as a discard pile to interact with…

The theme I getting to here is this: the kind of interactivity/control a game gives its players determines the kind of players they attract (no, I don’t mean “aggro/control/etc.”-- it’s much more nuanced than that)-- sort of like the way FPS and 3rd person shooters tend toward different players, or at least different kinds of gameplay.

All cards/ mechanics, including RNG-effects, introduced in a game will be seen by players as a challenge to be “mastered” in their attempt to beat the game or other players-- an RNG-effect too crazy/weird to be “mastered” (or cannot be “answered”) will simply be rejected by top-tier players who can get to where they are only with consistency, consistency, consistency… and a game-changing card/mechanic rejected by such players will cause them to reject the game.

Which is not to say there isn’t a place for special needs/ interest cards like Groovy Lion in gimmicky/ fun decks like Solo Vaath.

TL;DR: An expansion is just that, an expansion-- whatever fails to work with the existing game/ cards (and trust me, people will be experiment/ testing the hell out of it) will be rejected with the equivalent of an “ignore/block”-- so the game developers will know by the game stats (and popular strategies) they track, how well a new card/ mechanic introduced has actually been “integrated”… if pure theory-crafting can tell developers how much RNG is “too much”, there would never have been any patches/bans/etc.


I, too, missed out on the thread the first time around. I’m glad to see this has been a civil discussion so far and I hope it stays that way. This issue is truly with the AMOUNT of RNG, not because it exists.

Life itself is the ultimate game we play, and as angry as I may get at the weather, there is never going to be a day when shouting my displeasure at it is going to make it do what I want. Now, of course, life’s RNG we have to deal with the best we can, and I believe we can all agree that some of humanity have seen more fortunate outcomes than others; more often than not with non-trivial investments of talents and conscious effort.

But then again, we are talking about Duelyst here, which is just a game made by people. We hope that they are willing to listen to our concerns and feedback, but we are ultimately responsible for our own happiness. I figure if it gets to the point where Couterplay’s development of this game incorporates more RNG than I am willing to put up with AND/OR start completely ignoring the majority of the player base’s constructive criticisms, I will move on.

So, to focus more on the topic, there is no simple answer because there are so many factors at work. If everyone had to play with the exact same cards and with no board it would be pure luck of the draw that would significantly impact the gameplay choices between players. Add the board into the mix with no other changes, now players must use more skill and experience to manage the RNG element of the draw. When we move on to adding the ability to choose our own cards for our decks, we explode the variances between players and the marriage of RNG and skill. And THEN, we have to deal with the types of RNG introduced by each card’s ability. It gets pretty complicated.

No one is ever going to agree completely on the correct level of RNG, but at this point we can still give our opinions and stay constructive and have faith that CPG is listening.

See you on the ladder!

Edit: It took me too long to write this because I got interrupted. I am now an echo of a previous post! Lol


God, this thread makes me feel so nostalgic. I remember arguing over this back in beta. It was interminable then, too.

People don’t tend to have a particularly good definition for RNG. The term represents a concept, certainly, but the concept isn’t just randomness, but also uncertainty, which is subtly distinct, and unpleasant surprises, and the inability to make a choice that feels satisfying. Words tend to end up like that, with connotations beyond their direct denotation, muddying any conversation in which they’re involved. It’s also worth noting that there’s no real objective and universally applicable measure of value in Duelyst. You can’t look at the different outcomes of a card with a random effect and assign a number to the difference between them, and say that all cards must keep the differences between their possible outcomes below a certain number. This makes it more or less meaningless when people suggest that Duelyst should have “less RNG.” What’s less?

I won’t presume to know what others think when they say they’d like less RNG, but I find that the random elements of Duelyst that are most prone to frustrate me are the ones that have a very concrete difference in value and the ones that I have no way of predicting. I’ll address the latter first. Most random effects in Duelyst have a relatively small range of outcomes. Many cards summon a token in one of eight spaces, minus whichever of them are occupied. This is few enough outcomes that you can plan for each of them. On the other hand, you have cards like Grincher, which has, I think, eighteen possible outcomes? Admittedly, you should have been planning for three of them, to some extent, already, but it’s hard to accommodate all the cards Grincher could give your opponent, which can make decision-making very stressful and uncertain. Regarding the former case, most cards with random effects in Duelyst don’t have large differences in objective quality built in to the outcome of their random effects. If they did, they would be much more frustrating, both for their player and the person they’re played against.

I also find it frustrating when I come out the worse for making a decision that was, as far as I can tell, correct. This necessarily happens occasionally. Duelyst is not designed to have games that last a long time, which means it’s unlikely for games to last until someone makes a mistake, at least with good players. It still feels bad when this happens. I don’t think there’s a way around this. As has already been mentioned in this thread, randomness is a fairly necessary component for games, especially turn-based ones. There’s an article about it here that I’ve had sitting in my bookmarks for ages, which I think might have been linked during this discussion’s equivalent in the beta.


Posting just to point out the incorrect statement in the OP:

“That’s because, as you get higher and higher in skill level, of course the games are going to be determined less and less by differences in skill or significant mistakes…which means of course the games are going to come down to RNG elements.”

This is not true. When play skill is EQUAL then RNG has a greater effect, however players can be equally skilled at any rank.

It would be more correct to say that “once a player has ranked up to their general skill level, RNG will have a more significant impact”. This is, of course, true.

However that really depends on how DEEP skilled play is in the game. If it is significantly deep such that no two players are ever really on the same level, then it becomes less significant again.

Ie: If skill is on a spectrum or 1-10 then more people will find themselves on the same skill level, and thus be more RNG effected, than if skill is on a spectrum of 1-1000.


No, I think the impact of RNG on a game between people making no mistakes is much larger than if there were multiple mistakes on both sides…

It’s like once you reach that skill level it’s all in the deck, the draw, and the RNG - and not messing up.

Otherwise, RNG is like another mistake among many.


False dichotomy. There is nothing about RNG that prevents skillful play from being a meaningful or even necessary part of a comeback. In fact, skillful manipulation of RNG is often what makes RNG so dramatic: clearing spare targets to increase the chance a White Widow will strike for a kill, knowing that there’s a 1/7 chance that Vespyr’s Call will give you the Iceblade Dryad you need to seal the game before you die (and thus not conceding until you’ve seen the result), and recognizing that killing that Piloted Shredder has the potential to spawn a Doomsayer and save you even if it is a ridiculous longshot.

RNG isn’t the opposite of skill – it just benefits the less-skilled player if they’re skilled enough to use it properly.