7 ms·

Replayability in game design

85 points by amichail 4y ago

setr 4y ago
This article almost entirely focuses on content variation, but I think far more important is game system complexity and inter-complexity. The ultimate problem stopping replay (assuming you like the game systems in the first place) is the realization that you’ve “solved it” — you understand the various inputs and outputs of any action or mechanism and can trivially predict the results. This usually occurs because the game systems themselves are trivial (often a set of independent features stuffed behind the same UI, but A has no influence on B), because one or a few of the systems are overwhelmingly influential (e.g. melee mechanics don’t matter if guns are trivially superior for all scenarios), or because the game itself doesn’t incentivize the player to actually exploit those systems to any real degree (e.g. the game is too easy or the levels don’t introduce enough new constraints, so the first solution found applies globally; often fixed by player-induced artificial challenges like no X runs). Multiplayer however cheats this because even trivial systems can become a huge exploration space when incentivized by direct competition (analysis and understanding is pushed to the limit; e.g. sm64 requires a certain level of understanding to play effectively, but speed runners and TAS take that understanding to the nth degree) and the players themselves bring metagaming as an additional system worth exploring (eg any competitive multiplayer) It’s also confounded by multiplayer games that are less “played” and more used as a social space — replacing the shopping mall — like many mmo’s, Minecraft, etc. You’re not replaying so much as visiting and killing time between social interactions (so even highly repetitive actions often suffice). The content variation is largely necessary to give reason and space to explore the game systems under novel constraints.
- forgotpwd16 4y ago
  Note though that a game can have emerged complexity arising from simple rules. Classic examples will be chess and go.
  - Jensson 4y ago
    Chess and Go are only popular for legacy reasons, they shouldn't be used as examples of good game design. People care about the complexity in those games only because of the legacy culture and competitions, people would never care about them as new games without all that history.
    - stcredzero 4y ago
      People care about the complexity in those games only because of the legacy culture and competitions I think this is less than half true. People wouldn't care about those games, if they didn't have the amount of emergent complexity that they do. It would be more accurate to say that they soak up so much attention because of legacy culture and competitions. But those wouldn't exist, if those games didn't have certain qualities which made them eminently interesting and replayable in the first place. Could they have been replaced through different accidents of history with other games entirely? Sure. But those games would have had many of the same qualities.
      - rcxdude 4y ago
        Indeed. Chess and go have survived a long time, and they would not have done so if they were not interesting in their own right. We just don't see the games lost to history.
        hoseja 4y ago
        We do see some, check out "Royal Game of Ur" for example.
    - spywaregorilla 4y ago
      They're only POPULAR for legacy reasons, but popularity is a very poor proxy for game design. Chess and Go are great games. Fixed start, no hidden information, turn based games are pretty rare anymore. They're not eyecatching, but they're still good concepts.
- hypertele-Xii 4y ago
  > The ultimate problem stopping replay For me it's time. If I know in advance the game will take 10 hours to finish then no matter how much variety it has, I'm not gonna play it again.
  - Jensson 4y ago
    Why did you play it the first time? Why would the second time be different? Your statement doesn't answer this fundamental question about replayability.
    - egypturnash 4y ago
      The first time you play it, you will be encountering the game’s art, music, and design for the first time. Will it be good? Bad? Just okay? Will there be one surprise that takes your breath away? It’s full of novelty and possibility. The second time, you mostly know what you’re in for. There may be different paths to take, there may be new stuff unlocked by your first play through. But if you made it to the end then you’ve probably exhausted most of its unique content. There may be other things to come back for (DLC, competitive play, 100%, achievements, etc) but it’s easy to put it down for good. How many books, movies, or TV shows do you come back to? How many do you go through once and never again? Why should a game be any different?
      - boomlinde 4y ago
        > How many books, movies, or TV shows do you come back to? How many do you go through once and never again? Why should a game be any different? You can design a game such that it's a somewhat unique experience every time you play it. Even games with limited variation in setup allow different approaches from the player. That's not possible with traditional novels and film. At best, maybe you can review a film or a book and pay more attention to some detail you didn't notice that much in the first view, but in the end you are rather passively consuming a fixed narrative. Of course, how meaningful this is depends on whether the game mechanics stand out on their own, or if it's mostly a vessel for a fixed narrative. If the game is just a mechanically bland walk from one cutscene to another, with little agency, it might not seem worth it. A lot of AAA titles are like this IMO. If the game generates a unique set of challenges every time you play it, or rewards new approaches, there is much more fun to be had replaying it.
SSchick 4y ago
Got baited by the headline into thinking this would be an article about determenism and implementation of 'replay' functionality but this is interesting as well. I personally like to think the most 'replayable' games are the ones that have deep emergent gameplay. Eg. interactions between many different mechanics that when combined create unique/challenging/entertaining situations without being programmatically forced to occur. Unfortunately capturing this is notoriously hard and (from my experiences) primarily out in multiplayer games since humans are generally unpredictable and tend to find interesting edge-cases in game mechanics.
kibwen 4y ago
One topic that I didn't see it address: should your game be replayable? For many (most?) of my favorite games, movies, and books, I don't really care about experiencing them again. The first time through was delightful, and then I'm happy to move on and try other things. It's great if your game is replayable, but not all games need to worry about replayability. And even if your game is replayable, it should still be fun even to people who only play through once, without locking core mechanics behind endless repetition. And the pursuit of replayability is no excuse for trying to get your players addicted; players should feel free to stop when they've had their fill, without any pressure to consume more.
- mediaman 4y ago
  And implementing replayability often comes at a cost: it is much more difficult to create an interesting and compelling story, for example, or witty dialogue, if it must also have the sorts of variation to allow for novel experiences upon replay. That’s not important for all games, but for some, it absolutely is the difference between a deep and meaningful experience and one that has better replayability mechanics but lesser emotional impact. I’ve seen a number of high potential games get ruined because the developers thought they had to make it more replayable, which precluded other aspects of the game.
  - Gigachad 4y ago
    Two different kinds of games. Some of them are almost interactive movies with little replay value and some, especially multi player competitive games are entirely replay.
- xdrosenheim 4y ago
  Just look at the recent release of Stray. You can sink about 5 hours into it and get the whole story out of it. Sure, you can spend double that and get all the small collectables and achievements, but it is not needed for everyone. https://store.steampowered.com/app/1332010/Stray/ https://store.steampowered.com/app/1332010/Stray/
adamesque 4y ago
As pointed out in other comments, this article doesn't discuss the fundamental prerequisite for replayability, which is to build a core loop that is fun and rewarding. The fact that there are so many definitions of "rewarding" (challenging, relaxing, aesthetically pleasing, competitive, collectability) makes the solution space huge, and loot tables and biomes are a very small fraction.
- njyx 4y ago
  Agreed that you need a rewarding core loop and also one that becomes richer over time (more characters? more actions, more power, less power...). But, I think you also need a longer term engagement element beyond the core loop: being engaged with the goals of game. If there's only one (e.g. only one ending) then if you love the game you'll try to reach it at least one. If there are more you might be tempted to try them all. Another thing that helps and isn't discussed much in the article is the "meta" game around a game. Especially if the game has liveops and new characters come out, there are events etc. Most by themselves probably won't make someone replay. But if there's enough new stuff and a community which is constantly discovering new stuff to do then that can spark the desire to play again.
  - adamesque 4y ago
    My favorite way for a core loop to become richer over time is when the loop has a very high skill ceiling the player can move through in a semi-controlled way. For multiplayer games, this means good matchmaking, but for other games this is layered in via optional challenges/achievments/bonus goals etc. The stated goal of the game (aka reaching the ending) may or may not be along this path.
maltyr 4y ago
Games actively train a player's "skill". I define the player's skill as the player's ability to use the mechanics in the game to meet the challenge the game is presenting. The way to make a game more replayable, then, boils down to 3 things: 1) Make the game more challenging. This forces the player to change their approach to problems (because their old solution no longer works) Multi-agent games do this by having a variety of opponents. Single-player games do this by ramping up the difficulty, generally by making the execution requirements tighter. More complexity, or less leeway for mistakes. Leaderboard-type games do this by using a "best score" as the goal. (e.g. Rhythm games, racing games, etc.) 2) Adding and/or removing mechanics that force the player to change their approach to problems. Again, this is either because the old solution doesn't work, or you must use the new mechanic in the solution. 3) Add randomness, mimicking either of the above. (e.g. Roguelikes/lites, deckbuilders, etc.) There's also the possibility of extending play-ability by "creating new content". I think this isn't really "re"playability, since the generated content hasn't been "played" before, but procedurally-generated content and sandbox games have this sort of code-generated-novelty-as-replayability.
- orbital-decay 4y ago
  3) Add randomness, mimicking either of the above. (e.g. Roguelikes/lites, deckbuilders, etc.) Not sure that the randomness in roguelikes is just mimicking the challenge. Nethack is a canonical example of randomness adding to depth. The game is insanely random, but the mechanics are unbelievably deep and diverse, and if you are skilled, you can always avoid death in a myriad of creative ways unavailable for a newbie. The downside of such a design is its combinatorial complexity, which is also apparent in Nethack, as it also allows plenty of "abusive" behavior. In contrast, DCSS devs aggressively cut features from the game to "streamline" it (aka simplify it enough to make the task of preventing abuses of mechanics easier) This can be applied to anything skill-based that contain randomness or incomplete information, down to battle royale FPS games - the skill is in to use the knowledge and game sense against whatever the RNG (or a game situation) throws at you to get predictable results.
proc0 4y ago
This is a good analysis. I think there is a high level perspective that abstracts most of those points. For this context, games fall into two main categories or a combination of both: the ones created with a designed solution, and the ones created without a solution that the players have to discover and/or solve. A game is "solved" when the player has accomplished the main goals of the game, which usually means having gone through the main experience it's offering. A highly replayable game is one that was not designed with the solution out of the box. These games are usually "sandbox" games, and the designers created systems and tools for the player to use, but they did not design the specific experiences the player will go through. Sanbox games have emergent gameplay. In contrast, games that are designed with a solution in mind, don't have as much replayable value. Games that are story driven, like the The Last of Us, or Resident Evil, have less replay value than Factorio or GTA which are sandbox games. Many games are in between, having elements of both.
smiley1437 4y ago
Funny, the game I replayed the most was 'Hack', a Rogue-like. I can't quite put my finger on why it was so addictive, but I think it was the randomness - discovering which potions were which, which scrolls were which, and of course hoping for a merchant with a massive shop full of goods to pillage kept me playing. The difficulty was quite high too, forcing quite a few replays before getting that Amulet of Yendor
flobosg 4y ago
(2018)
hoseja 4y ago
Has anyone been working on using AI for content generation for games? Seems like a very promising space. Like you have a "traditional" game but some part that would traditionally be hand-crafted or boringly procedurally generated would instead be provided by DALL-E/GPT/...
lazlee 4y ago
The quickest way to kill replay ability in games: cut scenes that you can not skip. If the player has seen the scene once, there should always be the option to skip it.

85 points by amichail 4y ago

setr 4y ago

This article almost entirely focuses on content variation, but I think far more important is game system complexity and inter-complexity. The ultimate problem stopping replay (assuming you like the game systems in the first place) is the realization that you’ve “solved it” — you understand the various inputs and outputs of any action or mechanism and can trivially predict the results. This usually occurs because the game systems themselves are trivial (often a set of independent features stuffed behind the same UI, but A has no influence on B), because one or a few of the systems are overwhelmingly influential (e.g. melee mechanics don’t matter if guns are trivially superior for all scenarios), or because the game itself doesn’t incentivize the player to actually exploit those systems to any real degree (e.g. the game is too easy or the levels don’t introduce enough new constraints, so the first solution found applies globally; often fixed by player-induced artificial challenges like no X runs). Multiplayer however cheats this because even trivial systems can become a huge exploration space when incentivized by direct competition (analysis and understanding is pushed to the limit; e.g. sm64 requires a certain level of understanding to play effectively, but speed runners and TAS take that understanding to the nth degree) and the players themselves bring metagaming as an additional system worth exploring (eg any competitive multiplayer) It’s also confounded by multiplayer games that are less “played” and more used as a social space — replacing the shopping mall — like many mmo’s, Minecraft, etc. You’re not replaying so much as visiting and killing time between social interactions (so even highly repetitive actions often suffice). The content variation is largely necessary to give reason and space to explore the game systems under novel constraints.

forgotpwd16 4y ago
Note though that a game can have emerged complexity arising from simple rules. Classic examples will be chess and go.
- Jensson 4y ago
  Chess and Go are only popular for legacy reasons, they shouldn't be used as examples of good game design. People care about the complexity in those games only because of the legacy culture and competitions, people would never care about them as new games without all that history.
  - stcredzero 4y ago
    People care about the complexity in those games only because of the legacy culture and competitions I think this is less than half true. People wouldn't care about those games, if they didn't have the amount of emergent complexity that they do. It would be more accurate to say that they soak up so much attention because of legacy culture and competitions. But those wouldn't exist, if those games didn't have certain qualities which made them eminently interesting and replayable in the first place. Could they have been replaced through different accidents of history with other games entirely? Sure. But those games would have had many of the same qualities.
    - rcxdude 4y ago
      Indeed. Chess and go have survived a long time, and they would not have done so if they were not interesting in their own right. We just don't see the games lost to history.
      - hoseja 4y ago
        We do see some, check out "Royal Game of Ur" for example.
  - spywaregorilla 4y ago
    They're only POPULAR for legacy reasons, but popularity is a very poor proxy for game design. Chess and Go are great games. Fixed start, no hidden information, turn based games are pretty rare anymore. They're not eyecatching, but they're still good concepts.
hypertele-Xii 4y ago
> The ultimate problem stopping replay For me it's time. If I know in advance the game will take 10 hours to finish then no matter how much variety it has, I'm not gonna play it again.
- Jensson 4y ago
  Why did you play it the first time? Why would the second time be different? Your statement doesn't answer this fundamental question about replayability.
  - egypturnash 4y ago
    The first time you play it, you will be encountering the game’s art, music, and design for the first time. Will it be good? Bad? Just okay? Will there be one surprise that takes your breath away? It’s full of novelty and possibility. The second time, you mostly know what you’re in for. There may be different paths to take, there may be new stuff unlocked by your first play through. But if you made it to the end then you’ve probably exhausted most of its unique content. There may be other things to come back for (DLC, competitive play, 100%, achievements, etc) but it’s easy to put it down for good. How many books, movies, or TV shows do you come back to? How many do you go through once and never again? Why should a game be any different?
    - boomlinde 4y ago
      > How many books, movies, or TV shows do you come back to? How many do you go through once and never again? Why should a game be any different? You can design a game such that it's a somewhat unique experience every time you play it. Even games with limited variation in setup allow different approaches from the player. That's not possible with traditional novels and film. At best, maybe you can review a film or a book and pay more attention to some detail you didn't notice that much in the first view, but in the end you are rather passively consuming a fixed narrative. Of course, how meaningful this is depends on whether the game mechanics stand out on their own, or if it's mostly a vessel for a fixed narrative. If the game is just a mechanically bland walk from one cutscene to another, with little agency, it might not seem worth it. A lot of AAA titles are like this IMO. If the game generates a unique set of challenges every time you play it, or rewards new approaches, there is much more fun to be had replaying it.

SSchick 4y ago

Got baited by the headline into thinking this would be an article about determenism and implementation of 'replay' functionality but this is interesting as well. I personally like to think the most 'replayable' games are the ones that have deep emergent gameplay. Eg. interactions between many different mechanics that when combined create unique/challenging/entertaining situations without being programmatically forced to occur. Unfortunately capturing this is notoriously hard and (from my experiences) primarily out in multiplayer games since humans are generally unpredictable and tend to find interesting edge-cases in game mechanics.

kibwen 4y ago

One topic that I didn't see it address: should your game be replayable? For many (most?) of my favorite games, movies, and books, I don't really care about experiencing them again. The first time through was delightful, and then I'm happy to move on and try other things. It's great if your game is replayable, but not all games need to worry about replayability. And even if your game is replayable, it should still be fun even to people who only play through once, without locking core mechanics behind endless repetition. And the pursuit of replayability is no excuse for trying to get your players addicted; players should feel free to stop when they've had their fill, without any pressure to consume more.

mediaman 4y ago
And implementing replayability often comes at a cost: it is much more difficult to create an interesting and compelling story, for example, or witty dialogue, if it must also have the sorts of variation to allow for novel experiences upon replay. That’s not important for all games, but for some, it absolutely is the difference between a deep and meaningful experience and one that has better replayability mechanics but lesser emotional impact. I’ve seen a number of high potential games get ruined because the developers thought they had to make it more replayable, which precluded other aspects of the game.
- Gigachad 4y ago
  Two different kinds of games. Some of them are almost interactive movies with little replay value and some, especially multi player competitive games are entirely replay.
xdrosenheim 4y ago
Just look at the recent release of Stray. You can sink about 5 hours into it and get the whole story out of it. Sure, you can spend double that and get all the small collectables and achievements, but it is not needed for everyone. https://store.steampowered.com/app/1332010/Stray/ https://store.steampowered.com/app/1332010/Stray/

adamesque 4y ago

As pointed out in other comments, this article doesn't discuss the fundamental prerequisite for replayability, which is to build a core loop that is fun and rewarding. The fact that there are so many definitions of "rewarding" (challenging, relaxing, aesthetically pleasing, competitive, collectability) makes the solution space huge, and loot tables and biomes are a very small fraction.

njyx 4y ago
Agreed that you need a rewarding core loop and also one that becomes richer over time (more characters? more actions, more power, less power...). But, I think you also need a longer term engagement element beyond the core loop: being engaged with the goals of game. If there's only one (e.g. only one ending) then if you love the game you'll try to reach it at least one. If there are more you might be tempted to try them all. Another thing that helps and isn't discussed much in the article is the "meta" game around a game. Especially if the game has liveops and new characters come out, there are events etc. Most by themselves probably won't make someone replay. But if there's enough new stuff and a community which is constantly discovering new stuff to do then that can spark the desire to play again.
- adamesque 4y ago
  My favorite way for a core loop to become richer over time is when the loop has a very high skill ceiling the player can move through in a semi-controlled way. For multiplayer games, this means good matchmaking, but for other games this is layered in via optional challenges/achievments/bonus goals etc. The stated goal of the game (aka reaching the ending) may or may not be along this path.

maltyr 4y ago

Games actively train a player's "skill". I define the player's skill as the player's ability to use the mechanics in the game to meet the challenge the game is presenting. The way to make a game more replayable, then, boils down to 3 things: 1) Make the game more challenging. This forces the player to change their approach to problems (because their old solution no longer works) Multi-agent games do this by having a variety of opponents. Single-player games do this by ramping up the difficulty, generally by making the execution requirements tighter. More complexity, or less leeway for mistakes. Leaderboard-type games do this by using a "best score" as the goal. (e.g. Rhythm games, racing games, etc.) 2) Adding and/or removing mechanics that force the player to change their approach to problems. Again, this is either because the old solution doesn't work, or you must use the new mechanic in the solution. 3) Add randomness, mimicking either of the above. (e.g. Roguelikes/lites, deckbuilders, etc.) There's also the possibility of extending play-ability by "creating new content". I think this isn't really "re"playability, since the generated content hasn't been "played" before, but procedurally-generated content and sandbox games have this sort of code-generated-novelty-as-replayability.

orbital-decay 4y ago
3) Add randomness, mimicking either of the above. (e.g. Roguelikes/lites, deckbuilders, etc.) Not sure that the randomness in roguelikes is just mimicking the challenge. Nethack is a canonical example of randomness adding to depth. The game is insanely random, but the mechanics are unbelievably deep and diverse, and if you are skilled, you can always avoid death in a myriad of creative ways unavailable for a newbie. The downside of such a design is its combinatorial complexity, which is also apparent in Nethack, as it also allows plenty of "abusive" behavior. In contrast, DCSS devs aggressively cut features from the game to "streamline" it (aka simplify it enough to make the task of preventing abuses of mechanics easier) This can be applied to anything skill-based that contain randomness or incomplete information, down to battle royale FPS games - the skill is in to use the knowledge and game sense against whatever the RNG (or a game situation) throws at you to get predictable results.

proc0 4y ago

This is a good analysis. I think there is a high level perspective that abstracts most of those points. For this context, games fall into two main categories or a combination of both: the ones created with a designed solution, and the ones created without a solution that the players have to discover and/or solve. A game is "solved" when the player has accomplished the main goals of the game, which usually means having gone through the main experience it's offering. A highly replayable game is one that was not designed with the solution out of the box. These games are usually "sandbox" games, and the designers created systems and tools for the player to use, but they did not design the specific experiences the player will go through. Sanbox games have emergent gameplay. In contrast, games that are designed with a solution in mind, don't have as much replayable value. Games that are story driven, like the The Last of Us, or Resident Evil, have less replay value than Factorio or GTA which are sandbox games. Many games are in between, having elements of both.

smiley1437 4y ago

Funny, the game I replayed the most was 'Hack', a Rogue-like. I can't quite put my finger on why it was so addictive, but I think it was the randomness - discovering which potions were which, which scrolls were which, and of course hoping for a merchant with a massive shop full of goods to pillage kept me playing. The difficulty was quite high too, forcing quite a few replays before getting that Amulet of Yendor

flobosg 4y ago

(2018)

hoseja 4y ago

Has anyone been working on using AI for content generation for games? Seems like a very promising space. Like you have a "traditional" game but some part that would traditionally be hand-crafted or boringly procedurally generated would instead be provided by DALL-E/GPT/...

lazlee 4y ago

The quickest way to kill replay ability in games: cut scenes that you can not skip. If the player has seen the scene once, there should always be the option to skip it.