Pandora, Randomness, and Game Design

Pandora, Randomness, and Game Design

The first quote in the Faeria trailer is:

“A profound yet accessible game where random fades before player skill.”

This idea is the main reason I was drawn to Faeria. I quit Hearthstone because I did not feel rewarded for my actions, or that I had enough control on the outcome of the game. With the release of Pandora, the amount of randomness in Faeria has become a hot topic. Many players are expressing frustration that reminds me of the thoughts I had when I quit Hearthstone.

I have seen a lot of good conversation about randomness, but a missing piece has been “why” an example of randomness is “good” or “bad”. I see a potential in Faeria that no other card game has, and the developers have shown that they will listen to feedback from the players. So, this post will provide how I think randomness can be used to maximize the excitement of playing Faeria. If you are a fellow player, hopefully the game design philosophy is interesting enough to make this worth reading, as it is quite long.

Examples of recently introduced randomness:

  • Highlander: one card maximum increases deck variation.
  • Pandora: randomness in drafting card selection, Artifacts adding random events.
  • Balance: Magda Queen of Meroval, Tale of the Old Turtle, Luduan, Ruunin’s Champion (Verduran Guardian), and Fire Elemental reworked to include randomness in their design [1].

Impact of recent releases on number of players:

Graph from https://steamdb.info/app/397060/graphs/

With both the introduction of Highlander and Pandora, we see a spike in the number of players followed by a quick return to the norm of about 200 players. Ideally, the retention of players after the release of these new features would be high, resulting in an upward trend in the number of players. Considering that April had many days with more than 400 players, something about these new features failed to capture the attention of Faeria players. My suspicion is that poor game experience due to randomness played a part.

“How can I say players hate randomness? Because I have looked at a lot of market research on our players and they have spoken very loudly about things they perceive as random." [2]

Purpose of Randomness

In 2009, Mark Rosewater wrote an excellent article on randomness and game design in Magic the Gathering. I encourage you to read the entire article, but the points he highlights in bold are:

Upsides of randomness:

  • It creates surprises.
  • It makes the game play differently.
  • It allows players to react.

Downsides of randomness:

  • It can create repetition.
  • It can create frustration (and worse).
  • It can keep the game from advancing.
  • It can make the more experienced player lose.

The main purpose of randomness in a card game is to make the same deck provide a variety of experiences. Imagine if both player chose the order they drew their cards - the game would quickly become an optimized pattern repeated over and over.

Good randomness provides a way to introduce excitement into the game. Like exploring a randomly generated map, the player may discover new surprises and opportunities to take advantage of. Good randomness should tell a story - creating a unique and memorable experience.

Bad randomness takes control away from the player. When the victorious player feels like they did not deserve their win, something has gone terribly wrong. When both players feel confused or frustrated, some hard questions need to be asked about the use of randomness.

Resource Variation

Hearthstone and Faeria differ from Magic in a key way – the amount of resources each player receives is the same. A control deck in Magic will have more “lands” (cards that produce Mana / Faeria) than an aggressive deck, and both players are susceptible to drawing too many or too few lands. This variation in resources helps successive games between the two decks develop differently. In Hearthstone and Faeria, there is a risk of having games play out too similarly because the amount of Mana / Faeria generated is constant. Hearthstone intentionally combats this by introducing a high amount of randomness into the mechanics of its cards to try to make “the game play differently”.

Faeria has a lower requirement to introduce randomness than Hearthstone because land placement and collection from wells provide some variation in the resources generated. If I am playing a red control deck, I may change my initial land placement based on whether or not I draw Bomb Slinger. I may have more resources with Farm Boy in my opening hand than if I did not. This variation provides a dynamic experience that is core to the appeal of the game. This sort of randomness is “good” because it allows the player to react and plan based on their random opening hand.

Play the Odds

“In theory, better players should be better at planning their turns around the randomness that occurs, so the increased amount of in-game luck should theoretically be counterbalanced by an increase in the skill ceiling of the game itself." [3]

“Good” randomness provides an opportunity to increase the skill cap of the game. Conversely, “bad” randomness takes away opportunity for skill and allows opportunity for “the more experienced player to lose”.

Consider a card like Unbound Evolution (transform a creature into a random creature that costs 2 more). This card has “good randomness” because a skilled player knows which evolutions have a good chance of reducing an enemy’s minion strength or improving a friendly minion’s strength. The new design of Tale of the Old Turtle (add 3 random blue cards to your hand, they cost 3 less) is more debatable.

While it creates surprises and makes the game play differently, it does not provide significant opportunity to demonstrate skill (just play the cards you get, if you can). It also can create frustration due to the high variation in value generated. An extreme example would be three Ruby Fish (frustrating for player) vs three Ninja Toad (frustrating for opponent). However, the general impression I get from the player base is that they enjoy cards like Spellwhirl, so perhaps this gain in enjoyment is worth the cost in balance.

In the same vein, Faeria’s mulligan system feels like a missed opportunity. The player makes a binary choice of keep or mulligan, when there is the potential for a more complex decision. For example: draw six cards, choose three to keep. With only a small change, this allows the player to build a plan and reduces the chance of a terrible starting hand.

Board State and Tempo

“…the earlier the randomness occurs in the game, the better." [2]

This point we need to be careful with. I agree that “Responding to early random events is fun. Having the end of the game hinge upon a random occurrence is not. In short, players are much more willing to endure random events if they have time to respond to them." [2] However, this point is not true for random events that affect the state of the board.

Why? In magic, the defending player decides combat. The attacking player declares which of his creatures are attacking, and the defender chooses how combat occurs and which trades take place. In Hearthstone and Faeria, the attacking player decides. This means that Hearthstone and Faeria have a stronger tendency to “snowball”, where the attacking player gains an early lead on board such that there is no chance for recovery.

This makes a Pandora Artifact like Feast (three Faeria, distribute five +1/+1 counters randomly) more dangerous – drawn at the right time, Feast can give the attacking player an unrecoverable advantage, leading to a negative experience for the opponent. Reynad gives some good thoughts on this point and randomness in general in a Hearthstone context.

I was happy to see the balance change “It is no longer possible for an artifact to be the top card of any deck" [5], but it needs to go one step further. When an Artifact affects the state of the board early in the game, it often decides the outcome of the game. I would suggest that no Artifacts other than the shards that have no impact could occur in the first five turns of the game, but more on my suggestions at the end.

Problematic Pandora

Rosewater also provides some “Random Tips” to help implement randomness in an enjoyable way:

  • Make randomness lead to an upside.
  • Give players the chance to respond to randomness.
  • Allow players to manipulate the source of the randomness.
  • Avoid icons of randomness.

Artifacts

Consider Garudan’s Scale, a Pandora Artifact that deals three damage to each creature when it occurs. This event leads to a strong downside (punishing the players for playing creatures) and gives the players no chance to respond to the randomness. This event can randomly decide who wins and leads to a very negative and frustrating experience. The similar “transform all creatures” Artifact often does not lead to an upside, and also does not give the players any chance to respond.

As hinted at in the previous section, Feast and the free Triton Warrior Artifacts are problematic because they give a large advantage to the player who may utilize the free resources first when drawn early in the game. This is particularly true for edge cases where there is no empty land for one player, so only one player receives the Triton Warrior (yes, this has happened to me).

Drafting

The design of locking out a colour at the start of a Pandora draft is a good example of randomness. It provides focus, makes the draft play differently, and allows the player time to adapt. The decision to reduce the number of weak cards is also good. A Magic draft has the player choose 45 cards but only play ~23, so there is much more room for weak cards in the drafting process. In Pandora, you play every card you pick, so it is wise to reduce the chance of seeing five cards that are all terrible.

The idea that “drafting should challenge the player’s ability to maximize synergies between cards” [6] has the right idea; however, it is difficult to draft for synergy without any wiggle room in the drafting process. Players are highly incentivized to choose neutral cards (especially early in the draft), and to withhold choosing a second colour as long as possible until they see a bomb rare or key removal spell.

In a magic draft, due to the nature of a shared pool of cards between the players, it is likely that I will see more cards of one colour later in the draft because that is the colour my counterparts are not taking. Implementing a “drafting pod” is probably not the right path for Faeria, but there should be some way to incentivize a player for choosing cards with good synergy. At the moment, forcing an archetype feels like rolling the dice to see if you get all the puzzle pieces or not.

Pandora Opens

There is a very troubling section in the May 27 Faeria Friday:

“It’s only recently that we came to the conclusion that Faeria simply doesn’t work properly in limited. The general gameplay was way too “controlly,” and as a result - games lasted over 25 minutes consistently. It’s not surprising after all: the Faeria Wells act as magnets, offering an infinite potential of resource advantage. Combine this with an environment where decks tend to lack focus and you get a game mode where all you want to do is to draft the creatures that are best at collecting Faeria, stalling the game until you’ve stockpiled enough value to mathematically be the winner. To be honest: That gameplay is simply miserable, and we refused to admit it for a long time. The problem in a limited environment is that aggressive decks are inherently weaker when you can’t fine-tune your decklist to its maximum potential.” [6]

I must admit – I think removing the wells is completely the wrong solution to this problem. The wells are one of the few core mechanics that make Faeria truly unique, so this approach is removing strategy to enforce shorter games.

A game of Faeria is won by accumulating a resource advantage. If my opponent collects from three wells and I only collect from one, it won’t be long until I am defeated. By removing the wells and replacing with a fixed mana income, the game becomes a slugfest of “who drafted more value cards” because that is the only remaining avenue for an advantage to be found. In the very same article, the developers state “We didn’t want to design a drafting game where a card’s general “value” primarily dictates the cards chosen [6]”. Removing the wells contradicts this intention.

So, how can we accomplish the goal of “rich gameplay in a highly varied drafting environment" [6] for Pandora? I won’t pretend to have all the answers, but I can at least imagine an example that uses the principles we have learned so far.

Suggestions

  • Make Artifacts “discovered” when land tiles are placed instead of when cards are drawn. Artifacts are randomly seeded (hidden) on the board and may not be adjacent to a God or Faeria well.
  • A creature must spend its turn to “excavate” an Artifact (may not move and excavate on same turn).
  • An excavated Artifact is added as an event to the excavating player’s hand.
  • When five Artifacts are excavated, Pandora opens (each excavation site now produces Faeria like a well, is still path-able and does not destroy the tile).
  • Balance change example: Garudan’s scale is now 4 Faeria deal 3 damage to each creature.
  • Draft 35 cards, discard 5 at the end.

Concept image: red tiles may not have Artifacts, green tiles are possible, blue tiles are an example seed.

This structure rewards the player for doing something they already want to do (build lands), gives the player a chance to respond to the randomness (must excavate before event is obtained, opponent sees which event is excavated), and allows players to manipulate the source of randomness.

The amount of steps required to obtain an Artifact slows down their impact enough that they should not decide the game on their own. Since both players see which Artifacts are discovered, both players may incorporate the discovered randomness into their strategy. For example, if I see that you discovered Garudan’s scale, I may hold off on playing my Battle Toads until I force you to use it.

Remembering the problem of long Pandora games, switching the discovery of Artifacts to land placement rewards aggressive land placement and encourages the players to attack. Excavation require a creature encourages both players to be proactive playing their creatures, rather than waiting to respond. When the additional wells spawn, the player with better board control will collect more Faeria, and win the game more quickly.

Conclusion

Ultimately, this is just an idea. More important is the intention behind the randomness. I trust it to the developers to design a system where randomness creates a unique experience, provides opportunity to demonstrate skilled decision making, and does not unnecessarily punish the players. Randomness should create an opportunity, not enforce chaos. If this balance can be found, Pandora and Faeria will truly capture the imaginations of its players.

References

[1] Patch notes: May 23rd, 2016
[2] Kind Acts of Randomness | MAGIC: THE GATHERING
[3] http://blog.prismata.net/2014/07/15/luck-in-games/
[4] Reynad's Thoughts On Randomness In Hearthstone - YouTube
[5] Patch notes: May 27th, 2016
[6] Faeria Friday May 27th, 2016: Artifacts and You

9 Likes

Pelorn,

Thank you for your beautiful and excellent post on randomness. Many interesting points as we are still following very closely how our community experience Pandora on many levels.

It has only been one week and already a ton of feedback to process and some adjustements on the system, but we value complete, in-depth opinions a lot.

Looking forward to see what other players think about it.

2 Likes

Thanks for the detailed, constructive feedback. Posts like this are invaluable to us.

Since you seem to be very interested in our thoughts about randomness, you might want to check out this article we released about it: https://www.reddit.com/r/faeria/comments/48rgoo/dev_speak_the_role_of_rng/

You’ll notice a lot of overlap with the Rosewater philosophies you’ve quoted.

3 Likes

Very good post @Pelorn, I absolutely agree with you!
I would be very happy to see the suggestions you listed implemented to the game.

What I would also like to see is, if we could draft 35 or 36 cards and use the extra cards as a sideboard (explanation).

edit: I like the current mulligan system and dont think it should change. Trying to build a balanced deck is after all really fun (and challenging). Having easy muligans would just make every deck greedier.

3 Likes

Awesome post @Pelorn - its great to see players efforts to giving feedback on making this game as good as it possibly can be (from their perspective!). I agree with the majority of what you say and the “excavate” idea is interesting for artifacts. The randomness can be truly frustrating at times, with wild swings of fortune based on who’s turn it is like when the Triton Warrior gets played, or if your yellow rush deck gets hit with the “3 damage to all”. Having the card go to your hand and not play instantly helps to give the second player time to adapt to the event. Overall thought the mechanic does add variety from the vanilla faerie experience. On your point about the fearia spawns stopping after Pandora opens - perhaps just the 2 closest to the opponent could spawn on your turn? That would force you to play closer to the opponent orb? It would give you the additional and also push you to block those cyclones - if you don’t harvest them then they will be available to the opponent on his turn - more incentive to push up.

I’ve said it a few times but I think varying the board layout is a huge piece of the Faeria gameplay that hasn’t been utilized yet. Having Pandora play on a random board design from a set of 5 or so would add enormous variability to a game. For example, if the fearia spawns were clustered near the middle it would totally change a players strategy.

I agree with Jok3se about the mulligan. If I could design my first hand to get the cards I want it feels the game would be a procession - the unpredictability of the opening hand adds a strategic layer I think.

4 Likes

The article is excellent and hits the nail on the head as to why the current randomness is not working. It’s great to see people invest so much effort into improving the game. It means the game’s concept is solid enough to be defended.

I like the idea of excavating artifacts, I think it adds more lore to the game, but not so much uncovering wells and devastating effects. This can create a unidimensional experience, with everyone placing land tiles frenetically and getting annoyed when opponent is luckier. The person who uncovers an artifact would also have the huge advantage of being able to play the card when it favours him. Additionally, neutral minions would also be highly favoured and having 5 extra wells on the map for a total of 9 would make controlling wells pointless.

What I would do is remove the effects of excavated artifacts (i.e. make them shards) while adding an amount of faeria that gets discovered with them (let’s say 2 faeria per excavated artifact, even though this can be randomized to spice things up a bit and add a sense of discovery).

This way exploration is encouraged, but doesn’t give your opponent the feeling he’s been cheated by RNG, as the advantage gained is fair. So you can go for artifact hunt, or you can go for wells and none of these are necessarily better. The wells are safer resource gains, but artifacts give you an ace in the sleeve, as I will shortly explain.

Upon excavating an artifact, the shard would go into effect immediately and when Pandora is opened, the wells shut down and the board gets cleared. Although this gives the player that excavates it a huge advantage, the first amount of 6 faeria would be given out next turn, to his opponent. This would act as a form of compensation, similar to the extra land tile you get at the start of the game.

How would these adjustments improve gameplay:

  • they would encourage exploration, making discoveries feel like an upside, while not a significant downside for your opponent

  • when you are ahead and have a strong board, you still have the danger of your opponent excavating the last shard and making a comeback, but you have full control on preventing this from happening; this adds new elements of strategy, while preventing that predictable gameplay of whoever controls the wells slowly winning

  • the player who gets behind in the game would still have a shot at a comeback with the 5th artifact, while your opponent will have to battle on 2 fronts - prevent you from excavating, while maintaining control of the wells.

5 Likes