Uncertainty and RNG in games
Note: Just starting a new thread as requested to discuss the effect of uncertainty and RNG in games, in particular with regards to the RNG nature of Elven Mercenaries.
NovaBlast;n8446640 said:
One might say its a negative thing for the game as well. . RNG is an inherent quality of CCG's
RickMelethron;n8447930 said:
so..? diseases are an inherent part of life, doesn't mean i want more of them
WARNING: WALL OF TEXT
Yes and no. On the one hand, you have Hearthstone, which has over time become more and more RNG based as they have added many cards that are both strong and viable despite their RNG nature. This means that there are many RNG cards in constructed that can thrive and be effective, and hence add more and more RNG to any particular game.
RNG is generally a downside to any card if it's incorporated into its effect. That's why when you observe the Piloted Shredder (4/3, spawns a random 2-cost minion when it dies, compared to the basic 4/5 of Chillwind Yeti), you're in effect trading 2 health for the deathrattle, which is okay if it spawns something like a Nat Pagle, but a really good trade if it spawns a proper 2/3 or 3/2 body. Thus, the cost of the RNG has been in built into the card which makes it both cost-effective and RNG.
Put enough of these cards into the game, and you incentivise players to include all these solid RNG cards, and push the meta into something which is very RNG, overly RNG.
On the other hand, if you remove the RNG element of the Piloted Shredder and give it a consistent effect (say spawning a no ability 2/3), and this card becomes absolutely borked. You're basically getting a 4 drop that can spawn a 2 drop (or maybe 1.8 equivalent drop) when it dies. It's totally the RNG effect which keeps the card in line.
Similarly with Elven Mercenary, the effect is absolutely ridiculous if there's no RNG to it. Imagine if you could choose which card to play with your Elven Mercenary: I'm pretty sure people would say this would be incredibly OP, and I'm glad that this isn't the case. In fact, many ST efficiency decks include so many ways to lessen the effect of RNG: Brouver, Nature's Gift (to promote better First Light chains), King of Beggars to name a few. These are all Silver options, and if you make Elven Merc Bronze-only you're basically removing the need for ST to run these Silver efficiency options because suddenly their Elven Merc chain will be a 50+% thing rather than the severely diminished version in post patch which makes current patch ST look like a laughable runt of its former glory.
You might say, if the Elven Merc ability is so strong that it requires an RNG element, why not give it a point penalty instead e.g. 2 power? That seems fair right, combined with a more reliable Bronze-only pull? Well, not necessarily. A big part of the reason that Elven Mercenaries got increased from 3 to 4 was to reduce their interaction with King of Beggars. Before, you had a really simple tiering of ST strength, 2 for Hawker Healers, 3 for Elven Mercenary, and 4 for BMC. This was a huge reason why KOB was such a hugely efficient card in ST, because you knew 100% what you'd draw with KOB, and then decoy it. Right now there are plenty of other useful 4s: BMC, Hawker Healer, Hawker Smuggler, which compete for KOB pulls, diminishing the potential to thin your deck 100% reliably every game.
The state of Elven Mercenary resists a simple solution that solves all problems. I presume the OP was asking if Elven Merc being Bronze-only was a good idea is because pulling a silver spell for free could be seen as too strong. On the other hand I have shown that Bronze-only would be a significant buff as opposed to the proposed nerf, which makes it a poor change to achieve what was intended.
Now I will move onto your point, which is less RNG in a CCG is inherently a good thing. Hearthstone is proof of this, you might say. Well, again, yes and no.
What makes a competitive game interesting and fun? I would argue that in any game that involves PvP, it has to involve a combination of two things:
1. It is a game of skill, and
2. That given equal skill, both players have a roughly 50-50 chance of winning, i.e. uncertainty, which means neither player is sure to win. You may argue that this is simply fairness, but this also means the result of any match will involve some degree of uncertainty. If the result of any given match was certain before you started said match (for example, match fixing and declaring the match result), nobody would watch this or play this. There is little entertainment value for the player, and perhaps none for any observer.
Feel free to contest these, but I will take it that these are a given.
I will also make a claim for an addendum to 2, which is:
2*. Given unequal skill, there is always a chance, however small, that the underdog will win.
This is potentially controversial. You may argue that the better player should always win, especially given a big MMR differential. However, if you have ever followed any competitive sport, you'll know that most of the most memorable games (you know, apart from the 5-4 wins in injury time) are the ones between the big teams that had a huge impact on the state of the league, but also in particular the scalps that much smaller teams take from the top of the league. The David vs Goliath games that goes in David's favour, and the way that fans and pundits alike rave about those underdog stories really shows that the uncertainty of results and random nature of the game add flavour to said game. It's the stories and drama that people lap up and add a layer of emotion. These games are fundamentally more interesting than the boring drawish games of the middle of the league table.
Given these 2 or 3 qualities of an interesting competitive game, let's look at some examples:
Poker:
1. Hold'Em has really taken off. Players and fans constantly defend Poker from anti-gambling laws by claiming that it's a game of skill, which are as a class exempt from the anti-gambling laws. However, note that this game has a huge amount of RNG: in fact, Omaha is a variant which is noted for being particularly exciting for player and observer alike for being high variance (RNG). Case in point, the leader on the flop in Poker is often in the 50%-60% range as favourite, whereas in Omaha Hi the leader in any particular hand can be as little as a 36% dog. Yet people find this game interesting despite that.
2. Poker is by nature an uncertain game, no matter if it's a community card based variant like Hold'Em or Omaha, or a pocket-only variant like Stud. There is no way to tell what the next draws will be, the only thing a player can do is play the odds, and more importantly, play the opponent. Pressuring the opponent in the correct way will win a hand more often than simply going to the showdown, when RNG i.e. hand quality is the only determining factor of who wins any particular hand. WSOP Europe champ Annette Obrestad went so far as to show the importance of player skill, position and pressure by playing an online tournament without looking at her hand, except where she was calling all-in situations where the only options are Call or Fold. In these cases she looked at her hand to make that decision. She won said tournament.
The excitement of poker as a spectator sport and player is that the next card and next development is never certain, not even on the river. Both players can make a grab at the pot even after both players have seen all the cards to be dealt. The player with the losing/behind hand wins a hand as often as the player with the winning/ahead hand, because it's a game of decision making and partial information.
The role of RNG in Poker: Poker is a classic example of a game that is labelled a game of skill, which has taken off in the competitive scene with millions of viewers and millions of dollars worth of sponsorship and prize pool, but which has a huge factor of RNG in built into the game. Prized player skills include psychology and reading a player, nerves of steel, balancing of odds (remembered and calculated), and clear accurate memory of player actions, both during a hand and over the course of a table/tournament. The effects of RNG are balanced out over the course of many hands, and as such you try to win out percentage-wise over hours rather than minutes.
Chess:
For my second example, I have chosen a game which has no RNG in it at all. Let's compare it to my criteria.
1. A game of skill? Oh most certainly. I'm distinctly intermediate at chess, and I make no claims to being good at it.
2. This game is a completely fair game (minus the coin flip. I mean, which side you start with.) Both players start with completely even resources, and the board is symmetrical. There is no RNG. In fact, some may argue that Chess has been solved, and we now have huge endgame databases, or at least that the game will be entirely solved.
https://en.wikipedia.org/wiki/Solved_game
The most obvious example of a solved game is Tic-Tac-Toe, which is so simple it's basically boring- a forced draw is inevitable so long as neither player throws the game. So let's consider a weakly solved game for now, Checkers. With deep computer calculation, they've weakly solved this game so that a forced draw can be provided with proofs of every single step being optimal. Imagine if humans played like this. Pretty boring, huh? Regardless of whether it's forced wins or forced draws, both scenarios are uninteresting.
The reality is that Checkers is only interesting because players make mistakes, by making the suboptimal plays (inaccuracies) or blunders, and hence the better player wins, more often than not. The key phrase here is "more often than not", which is the uncertainty quality of 2. Similarly, the Chess World Championship is only interesting because Vishy Anand had a chance (actually 2 chances) to take the crown from Magnus Carlsen, even though most people saw Carlsen as a huge favourite: the result was uncertain.
At this point though, it's worth quoting wiki:
"Whether a game is solved is not necessarily the same as whether it remains interesting for humans to play. Even a strongly solved game can still be interesting if its solution is too complex to be memorized; conversely, a weakly solved game may lose its attraction if the winning strategy is simple enough to remember (e.g.
Maharajah and the Sepoys). An ultra-weak solution (e.g.
Chomp or
Hex on a sufficiently large board) generally does not affect playability."
The poignant question is though, where does the uncertainty in Chess come from? This is a computer-calculable game where AIs can beat the best grandmasters quite reliably (Kasparov vs Deep Blue). Surely with such a stable and calculable game there should be less uncertainty?
While there is definitely less uncertainty in a game like Chess than Hearthstone, Poker or Gwent, it is the uncertainty that has allowed the game to flourish as a competitive sport. Top grandmasters are able to calculate many branching forced lines, yet fall back on general positional considerations when the position isn't forcing enough to be calculable. They have to calculate different lines because their opponent may play one of several.
Yet this uncertainty of exactly what is going to happen creates space for human endeavour: The depth of calculated lines, the adaptation of positional reasoning. The brilliancies, the blunders, the predictability yet unpredictability of human play. This is part of the reason that playing against humans will be more dynamic and more interesting than playing any of the current AIs, no matter the game.
Role of RNG: Non-existent. This game is wholly centred on player decision making and ability. Prized player skills include psychology, stamina, calculation, concentration, intuitiveness, inventiveness, memory and visualisation.
Gwent:
Where does Gwent fit into all of this? Well, it's a CCG which has a smaller emphasis on Draw RNG (fewer draws per game). It does have RNG cards but they're relatively few and far between, or at least their RNG is more limited (buffing a random unit, pulling a random card from deck) as opposed to the RNG of Hearthstone (Spawn any random 2-drop in the ENTIRE GAME).
1. Is it a game of skill? I would argue so. The fact that the same names are invariably at the top of the leaderboard shows that it's a game of skill.
2. The result is uncertain, which is a good thing. Check. However, the result is too RNG, you might say. Firstly, I would dispute this, since I feel that this game is less RNG than Poker is. In Poker, you often spend 60+% of your game waiting for a decent position, a good hand (considering the position), or a combination of the two. In Gwent, you're always in it to win it. No spam folding, just playing. This makes you feel the effects of wins and losses much more, despite the fact that laddering in Gwent and playing a poker table is exactly the same: if you play the percentages game, and win more than you lose, you're in the money.
Imagine if there were no RNG in Gwent's cards. Then the game would be entirely calculable. Armed with a deck tracker, any intermediate player with a brain could calculate the consequences of playing any card or line. Opponent's moves would be predictable and likely calculable by the best players with little to no uncertainty. The game would look much more similar to a game of AI weak solved checkers, or worse, Tic-Tac-Toe, with every game panning out the same way. Decks like old ST Ciri: Dash where both players would throw out disgusting 5 merc chains back-to-back would make the game stale and predictable, and stunt innovation and other archetypes. Ew.
Since we are playing with such thin decks, with such thinning mechanics, Gwent has been designed in a way that allows both players to field very consistent decks with themes and a consistent plan in mind. 25 cards is super thin for any kind of CCG, which means people are able to consistently assemble combo pieces required for their plan. If you ever remove the RNG of cards entirely, the game would likely move towards a solved state which is wholly uninteresting.
Role of RNG in Gwent: Considering that all interesting competitive games require a degree of uncertainty, removing RNG from cards entirely would move the game towards a weakly solved state, regardless of meta differences in between patches. This would be incredibly unhealthy for Gwent, unless you somehow found some other way of introducing uncertainty back into the game, but I believe that uncertainty is the whole thing you're railing against. Prized skills include bluffing, calculation, visualisation, probability, understanding multiple different archetypes and creativity.
Over the course of these examples there has been no consistency in RNG. You've got everything from no RNG to high RNG, yet all of these games are competitive and interesting for players and observers despite that. I believe that the reason for this is that the "healthy" amount of RNG depends on the type of game you have. Games with complete information tend to have little to no RNG, whereas games with incomplete information tend to have higher RNG. Card games need more RNG than most because they need to be less calculable to interact with the medium (i.e. cards) better. Chess pieces with their strengths and limitations on the other hand don't need RNG to be interesting and skillful to use.
In general, however, any game archetype will have some kind of sweet spot for uncertainty, which is generally in the middle. Too much uncertainty, and the result will be defined by it and dominate the influence of player skill, making it unfun for players (Hearthstone). Too little, and the game becomes too calculable, and depending on the complexity of the game might cause it to be played by humans like omniscient bots (Tic-Tac-Toe, Weakly solved Checkers.) It's important that any competitive game you're crafting hits that sweet spot of uncertainty.
In any game, it is the role of the player to manipulate the effect of uncertainty. In chess, a game of no RNG, the player that is ahead will try to minimise uncertainty, to secure the win in the safest manner, whilst the player who is behind will try to maximise uncertainty, in order to contrive a situation where they can make an insane comeback. This is true of any game. However, since the role of RNG in CCGs is higher, the role of the player is generally to weigh the odds and try to minimise RNG through careful sequencing and balancing of probabilities. That's not to say the whole losing/uncertainty - winning/certainty dynamic doesn't apply to CCGs: it does, just to a lesser extent. In Gwent, your role is to balance probabilities, but also figure out your opponent and their deck, applying pressure to them to make suboptimal plays where possible. Never forget that.
To summarise:
-RNG is neither inherently good or bad, and a "healthy" amount depends on the type of game in question;
-Uncertainty (which can come in the form of RNG) is fundamentally healthy for games;
-It is the role of the player to manipulate uncertainty in their favour: this is part of player skill- uncertainty is often provided by RNG in CCGs, but there are other ways too;
-I argue that uncertainty is generally on a spectrum: a game can have too much uncertainty (for example, RNG determining games more often than player skill, Hearthstone, ahem), or a game can have too little uncertainty (Tic-Tac-Toe.) It's just that RNG is a very helpful way of dialing up or down the uncertainty of any particular game, especially in the format of card games.
TL;DR: RNG is neither inherently good or bad, so you're both right/wrong.
