Developing proper Concepts for Card Evaluation
Like many (if not all) of the people on this forum I am super excited for Gwent. I love collectible card games and I sunk 150+ hours into the Witcher 3 (a decent amount of which was playing Gwent). The problem is that while we know a lot about the game already we can't play it yet. We don't all the cards so we can't really talk about balance. Since we can't play the cards/factions we don't know what the experience of playing them will be like and if it matches our expectations. In short: we can only speculate. That can be fun, but to me it is boring and unproductive (if it is not for, good for you!). So I thought I would propose a project to shorten the wait while doing something useful in preperation for the BETA: developing concepts for card evaluation.
In most other collectible card games or tradeable card games over their life span the playerbase develops concepts in order to help them understand how good/bad a card really is. If you are familar with the genre you will have heard of them. Examples are "Vanilla test", "Card Advantage", "Tempo" among a myriad of others. These concepts are often applicable to almost all games of the genre since the games are often similar enough (Magic: The Gathering and Hearthstone are prime examples of this). Gwent however is a another case, the framework is fundamentally different in a couple of ways:
- Cards are your only ressource (there is no such thing as Mana)
- You can only play one card per turn.
- You do not draw a card at the beginning of your turn.
- Cards on your side of the board can not interact with the cards on your opponnents side of the board (there is no such thing as creature combat)
- The basic game mode is a best of three, but played with one hand (no such thing as player life count etc.)
- others (honestly there are too many to list them in a )
Since Gwent is so different concepts that were developed for other games do not work for Gwent and as such we have to develope our own. We know that the rules will not change massively from the Witcher "Mini Game" so we can make some progress in developing those concepts prior to the Beta. I do have some ideas which should be a good starting point but I am sure they can be improved upon, which is where the rest of the community (you guys) comes in. Here we go.
I believe that two core concepts should suffice for very basic card evaluation. I label them as "strength differential" and "turn cost" for now (if anyone has a better name, feel free to suggest them ) These two concepts should make many different cards comparable, even cards of different types (such as Spells and Units).
Strength Differential
The goal of a game of Gwent is to gain more strength in 2 out of 3 rounds. So the first important part of a card is to consider how much strength it adds to your side or how much strength it detracts from you opponents side when played. The more it does of either of those things the better the card is.
The simplest example is plain hero card such as Hjalmar. Hjalmar adds a 15 strength to your side so he has a basic strength differential of +15 when played.
The other way to affect a cards strength differential is by decreasing the number of points your opponnents has on his side of the board. Imagine you opponent has 2 Griffins (2 plain 8 strength units) in play , you have none. If you now play Scorch you remove 16 (2*8 )strength from your opponents side which results in a strength differential of +16 for you (-16 for your opponent).
Scorch is a great example to show that strength differential is not static. If you play Scorch on an empty board it has a strength differential of 0. So strength differentialbe is variable. Two concepts that are important to consider here are the "floor" and the "ceiling" of a card.
The "floor" of a card refers to its worst case scenario. For Scorch this would be -X if you control the creature(s) with the most power.
For Hjalmar the floor is +15 since he is a hero card and can not be debuffed (strength can not be modified by other cards).
The ceiling for Hjalmar is also +15 since he can not be buffed. For scorch the ceiling is theoretically infinite, which shows that the value of scorch is highly dependent on the micro context (board state) and the macro context (type of decks being played).
A case that can be confusing are cards that have the "Resurrection" such as Priestess of Freya. The strength differential for cards like this is calculated by adding the strength of the card played and the card that is brought back. Imagine you play a Priestess of Freya and then bring back a Griffin. That would be a strength differential of 8 (0+8 ).
Turn Cost/Advantage
In Gwent every card in your hand represents a turn since you can only play one card per turn and you have to play a card if you don't want to pass the round. Having more turns available to you is generally advantageous. Imagine you have won the first round and you have more cards in hand than your opponnent. You can essentially run him/her out of cards since they can't pass without risking losing the game on the spot.
A good way to think of this is that the majority of the cards have a "Turn Cost" of -1, by playing them and decrease the number of cards in your hand by 1. But there are some cards that add cards to your hand. The good old Decoy for instance has Turn Cost of 0, you play it (--1) and then get a card back to your hand (+1).
Just ike strength differential turn cost should factor in your opponents hand size. If you play a card that makes your opponent discard a card that results in a Turn Cost of 0 you play a card (-1) and your opponent discards a card from their hand (+1).
These concepts should suffice to evaluate cards. In order to compare Card Combos and Groups of Cards you just have to calculate the values for both concepts and add them up. How important Strength Differential in comparison Turn Cost is (and vice versa) is hard to tell without playing the game so we will have to wait till October to answer that question.
Properly evaluating cards is important for understanding the game, deck building and tactical decision making. However it is also important when talking about balance. In order to understand how card evaluation factors into balance. To make this easier the next section is about the basics of balance and what questions are interesting when talking about the balance of Gwent.
Basics of Balance
Balance is a complex topic and an attempt to explain it fully would take far too long (and probably bore you to death). So we are going to focus on capturing and understanding the core idea of balance.
Most people base their conception of Balance on the idea of Fairness. Fairness is concept that is used in many different situations but when it comes to games it is mostly used in relation to competitive games (e.g. Counter Strike, League of Legends, Chess). In those games players directly compete with each other and such a game is fair if every competing player/team has an equal chance to win the game. This is easily achievable in symmetrical games. In these games every player has the same set of starting options. It gets more problematic in asymmetrical games where players have different sets of starting options (e.g. in chess white moves first). So far so good but why should (or do) we care about fairness/balance?
A good place to start the search for an answer to this question is a quote from Sid Meier (creator of the Civilization Series, Pirates, Alpha Centauri and other great games):
If you are making a decisions you choose in between at least two options. If only one option is viable then the decision becomes redundant since you will choose that option every time. Such a decision can not be interesting.
So for a decision to be possibly interesting you have to have at least two viable options. This does not mean that if you have two viable options the decisions is interesting but that if you don't the decision can not be interesting.
What makes an option potentially viable ? It has to be as good or better as any other option in at least one situation from a certain perspective. That perspective is often the "playing to win" mentality, whose primary goal is to win the game (in competitive games).
So how many options need to be viable for a decision to be interesting ? All possible options ? 90% ? 55.58% ? At this point we get into difficult territory. One thing we can say for certain though: If an option is never viable it is redundant and if all options are equally viable in every situation the choice becomes meaningless. If an option fulfills this condition it is not necessarily viable but if it does not fulfill this conditions it can not be viable
Gwent (like almost all games) has several levels of decisions. Before you play a game you choose a faction, then you choose a character, then you choose which units to put in your deck. The top of those levels is the faction choice. Here the answer to question raised above is simpel. Ideally all factions should be equally viable. From here you can go down the chain. It is important to remember that is most useful to compare different options of the same choice since they are in direct competition.
To sum this section up: Balance is about identifying the viability of options. Identifying is the key here. Balance poses problems but does not answer them, that is the job of design as a whole. To clarify this issue think of a game as an equation with many (and I mean many) variables representing all the parts of the game that influence a certain balance issue. When thinking of balance you want the equation to have a certain outcome lets call it: "Balanced". Often when people make suggestions that aim to make a game more balanced they act as if a ton of those variables are constants and their suggestion is the only way to achieve the wanted outcome. The assumed constants however are in fact variables and worse they also represent parts of the game that influence other important aspects of the game (such as "Thematic" or even the most important of them all "Fun"). Balance is a condition that is necessary (to some degree) but not sueffecient for a good game.
Side notes:
Leader Cards - Different Leader Cards of the same faction are comparable with the metrics described above. The usefulness of Comparing Leader of different Factions should be fairly limited since the viability of a Leader is in general closely tied to its faction. Or in other words: Comparing Leaders of different Factions is not useful because they are not options for the same choice. (see the Balance section of this post for further eplanation).
Like many (if not all) of the people on this forum I am super excited for Gwent. I love collectible card games and I sunk 150+ hours into the Witcher 3 (a decent amount of which was playing Gwent). The problem is that while we know a lot about the game already we can't play it yet. We don't all the cards so we can't really talk about balance. Since we can't play the cards/factions we don't know what the experience of playing them will be like and if it matches our expectations. In short: we can only speculate. That can be fun, but to me it is boring and unproductive (if it is not for, good for you!). So I thought I would propose a project to shorten the wait while doing something useful in preperation for the BETA: developing concepts for card evaluation.
In most other collectible card games or tradeable card games over their life span the playerbase develops concepts in order to help them understand how good/bad a card really is. If you are familar with the genre you will have heard of them. Examples are "Vanilla test", "Card Advantage", "Tempo" among a myriad of others. These concepts are often applicable to almost all games of the genre since the games are often similar enough (Magic: The Gathering and Hearthstone are prime examples of this). Gwent however is a another case, the framework is fundamentally different in a couple of ways:
- Cards are your only ressource (there is no such thing as Mana)
- You can only play one card per turn.
- You do not draw a card at the beginning of your turn.
- Cards on your side of the board can not interact with the cards on your opponnents side of the board (there is no such thing as creature combat)
- The basic game mode is a best of three, but played with one hand (no such thing as player life count etc.)
- others (honestly there are too many to list them in a )
Since Gwent is so different concepts that were developed for other games do not work for Gwent and as such we have to develope our own. We know that the rules will not change massively from the Witcher "Mini Game" so we can make some progress in developing those concepts prior to the Beta. I do have some ideas which should be a good starting point but I am sure they can be improved upon, which is where the rest of the community (you guys) comes in. Here we go.
I believe that two core concepts should suffice for very basic card evaluation. I label them as "strength differential" and "turn cost" for now (if anyone has a better name, feel free to suggest them ) These two concepts should make many different cards comparable, even cards of different types (such as Spells and Units).
Strength Differential
The goal of a game of Gwent is to gain more strength in 2 out of 3 rounds. So the first important part of a card is to consider how much strength it adds to your side or how much strength it detracts from you opponents side when played. The more it does of either of those things the better the card is.
The simplest example is plain hero card such as Hjalmar. Hjalmar adds a 15 strength to your side so he has a basic strength differential of +15 when played.
The other way to affect a cards strength differential is by decreasing the number of points your opponnents has on his side of the board. Imagine you opponent has 2 Griffins (2 plain 8 strength units) in play , you have none. If you now play Scorch you remove 16 (2*8 )strength from your opponents side which results in a strength differential of +16 for you (-16 for your opponent).
Scorch is a great example to show that strength differential is not static. If you play Scorch on an empty board it has a strength differential of 0. So strength differentialbe is variable. Two concepts that are important to consider here are the "floor" and the "ceiling" of a card.
The "floor" of a card refers to its worst case scenario. For Scorch this would be -X if you control the creature(s) with the most power.
For Hjalmar the floor is +15 since he is a hero card and can not be debuffed (strength can not be modified by other cards).
The ceiling for Hjalmar is also +15 since he can not be buffed. For scorch the ceiling is theoretically infinite, which shows that the value of scorch is highly dependent on the micro context (board state) and the macro context (type of decks being played).
A case that can be confusing are cards that have the "Resurrection" such as Priestess of Freya. The strength differential for cards like this is calculated by adding the strength of the card played and the card that is brought back. Imagine you play a Priestess of Freya and then bring back a Griffin. That would be a strength differential of 8 (0+8 ).
Turn Cost/Advantage
In Gwent every card in your hand represents a turn since you can only play one card per turn and you have to play a card if you don't want to pass the round. Having more turns available to you is generally advantageous. Imagine you have won the first round and you have more cards in hand than your opponnent. You can essentially run him/her out of cards since they can't pass without risking losing the game on the spot.
A good way to think of this is that the majority of the cards have a "Turn Cost" of -1, by playing them and decrease the number of cards in your hand by 1. But there are some cards that add cards to your hand. The good old Decoy for instance has Turn Cost of 0, you play it (--1) and then get a card back to your hand (+1).
Just ike strength differential turn cost should factor in your opponents hand size. If you play a card that makes your opponent discard a card that results in a Turn Cost of 0 you play a card (-1) and your opponent discards a card from their hand (+1).
These concepts should suffice to evaluate cards. In order to compare Card Combos and Groups of Cards you just have to calculate the values for both concepts and add them up. How important Strength Differential in comparison Turn Cost is (and vice versa) is hard to tell without playing the game so we will have to wait till October to answer that question.
Properly evaluating cards is important for understanding the game, deck building and tactical decision making. However it is also important when talking about balance. In order to understand how card evaluation factors into balance. To make this easier the next section is about the basics of balance and what questions are interesting when talking about the balance of Gwent.
Basics of Balance
Balance is a complex topic and an attempt to explain it fully would take far too long (and probably bore you to death). So we are going to focus on capturing and understanding the core idea of balance.
Most people base their conception of Balance on the idea of Fairness. Fairness is concept that is used in many different situations but when it comes to games it is mostly used in relation to competitive games (e.g. Counter Strike, League of Legends, Chess). In those games players directly compete with each other and such a game is fair if every competing player/team has an equal chance to win the game. This is easily achievable in symmetrical games. In these games every player has the same set of starting options. It gets more problematic in asymmetrical games where players have different sets of starting options (e.g. in chess white moves first). So far so good but why should (or do) we care about fairness/balance?
A good place to start the search for an answer to this question is a quote from Sid Meier (creator of the Civilization Series, Pirates, Alpha Centauri and other great games):
"A (good) game is a series of interesting decisions"
- Sid Meier
If you are making a decisions you choose in between at least two options. If only one option is viable then the decision becomes redundant since you will choose that option every time. Such a decision can not be interesting.
So for a decision to be possibly interesting you have to have at least two viable options. This does not mean that if you have two viable options the decisions is interesting but that if you don't the decision can not be interesting.
What makes an option potentially viable ? It has to be as good or better as any other option in at least one situation from a certain perspective. That perspective is often the "playing to win" mentality, whose primary goal is to win the game (in competitive games).
So how many options need to be viable for a decision to be interesting ? All possible options ? 90% ? 55.58% ? At this point we get into difficult territory. One thing we can say for certain though: If an option is never viable it is redundant and if all options are equally viable in every situation the choice becomes meaningless. If an option fulfills this condition it is not necessarily viable but if it does not fulfill this conditions it can not be viable
Gwent (like almost all games) has several levels of decisions. Before you play a game you choose a faction, then you choose a character, then you choose which units to put in your deck. The top of those levels is the faction choice. Here the answer to question raised above is simpel. Ideally all factions should be equally viable. From here you can go down the chain. It is important to remember that is most useful to compare different options of the same choice since they are in direct competition.
To sum this section up: Balance is about identifying the viability of options. Identifying is the key here. Balance poses problems but does not answer them, that is the job of design as a whole. To clarify this issue think of a game as an equation with many (and I mean many) variables representing all the parts of the game that influence a certain balance issue. When thinking of balance you want the equation to have a certain outcome lets call it: "Balanced". Often when people make suggestions that aim to make a game more balanced they act as if a ton of those variables are constants and their suggestion is the only way to achieve the wanted outcome. The assumed constants however are in fact variables and worse they also represent parts of the game that influence other important aspects of the game (such as "Thematic" or even the most important of them all "Fun"). Balance is a condition that is necessary (to some degree) but not sueffecient for a good game.
Side notes:
Leader Cards - Different Leader Cards of the same faction are comparable with the metrics described above. The usefulness of Comparing Leader of different Factions should be fairly limited since the viability of a Leader is in general closely tied to its faction. Or in other words: Comparing Leaders of different Factions is not useful because they are not options for the same choice. (see the Balance section of this post for further eplanation).
Last edited: