Monday, 9 June 2014

BBBen Game Design

This post was originally supposed to be part a general progress report, but I spent so much time talking about BBBen's game design that I’m separating it into its own post. I’m not discussing any of the content of his games, just the game design he used for PAC and PAF and ways in which it can be expanded/improved.

Most of the games I’m working on (I’m foolishly working on more than one project, but many of them are just test ideas to try out design concepts) use what I think of as BBBen’s formula.  I beleieve this boils down to a few key points:
1:  The player has a set number of ‘turns’, even if it is not defined as such. For example, in PAC there are roughly 23 turns where the player decides what to do from multiple options and once he chooses the game advances.
2: The primary choice is to ‘train’ one of three stats.
3: How the player trains the stats determines the overall result of the game (primarily who the player ends up with and what sex scenes occur).
4: There are some other factors, such as a ‘shop’ and money that add a side resource for the player to consider, but I would not say they are *primary* to the game.
5: Many of the ‘traditional’ AIF elements are absent, finding the right object, ‘guessing the verb’, are either not present or not required. Searching drawers, cabinets, or asking the right questions is not required to advance the game.

Overall it is a very simple design. I suppose a player might find it arbitrary and that there is nothing for a player to ‘figure’ out in the long run as he can just play the game multiple times hitting the various tracks to see everything as opposed to figuring out a riddle or finding a hidden object. I find the concept elegantly beautiful in eliminating many of annoyances I have with other AIFs and it is a design I’ve been trying to emulate and improve upon.

It is the improvements I’ve been trying to do that add a lot of work. Below are the ‘features’ I have either tried to add or think could be added to the design.

More stats: This is the most obvious improvement; add more than three basic stats. In my design for my X-Men game, Marvel makes this easy by having six defined skills in their universe (intelligence, strength, speed, energy projection, durability, fighting skills). In addition, I’ve wanted to have skills for ‘classwork’ performance (English/Math) so the player is not totally dedicated to just combat but has some other focuses as well.

More skills means more work on the programming end, which I’m sure is a surprise to no one. It also adds a lot to the balancing of the game. How relevant is the ability to hit something compared to how hard you hit it?

I think more skills is worth it though as it adds up more player options, not just for combat, but to pursue other paths as well.

Different stat ranges: Along with multiple stats, I’ve been trying to add in different ranges for the values. In PAC there are three stats all of which max at twelve. I’m looking to have some ‘minor’ stats that would have a much smaller range. For example: strength might have a range of 1-7, while ‘mechanics’ could have a range of ‘0-2’; it wouldn’t take long to learn, but it would likely only come up a few times during the game.

Multiple Training Forms: Along with multiple skills I’ve tried to implement multiple methods of training. For example swimming would increase strength and endurance, while weight lifting would also increase strength and endurance, but in different increments, while boxing could improve endurance and fighting skills.

Having the stats improved in different increments is what makes this difficult. If weight lifting gets a lot of strength and a little endurance, and swimming gets the opposite, it doesn’t make sense to use ‘1’ as the base statistic for every training action.  Thus for programming the base has to change to something like ‘5’, so that weight lifting would get 5 strength and 2 endurance, while swimming would get 5 endurance and 2 strength.

While adding a lot of complexity to the balance issue, it opens up a lot of choice for the player and allows player to develop their character in many different fashions.

Declining Returns: Something I thought was going to be easy, then turned out to be hard, before going back to easy once I got a handle on it, is declining returns.  As far as I’m aware, PAF used a linear progression with each point netting the same level of improvement (For example, a stat point in a health related skill gives +5 HP every time one is assigned). I’m looking to gear the system so that rewards for pursuing the same stat declines over time (I originally wanted slow growth, then for a sweet spot to be hit with maximum return, followed by declining returns, but I’m looking at just the last one at the moment).

Originally this was hard because there was no clear way to do the math required. My idea going in was to put in multiplying factors. For example, If strength was between 1-3 it would count as just its number, if it was between 4-6 it would be (((Value-3)*0.6)+3), 7-9 would be (((Value-6)*0.5)+5) and so on. There are a couple of problems. The first is that Adrift does not like decimals as results and rounds everything up, but that can be gotten around by thinking with a base ten (or hundred) instead of one.  The second is that it requires a lot of writing of math down and into the program, especially if you want to have more than just 1v1 fights.

How it got easy is when I started thinking in terms of groups. Characters have two strength totals; overall strength and strength level. A character starts at 1 overall strength and 1 strength level. When a character ‘trains’ strength it goes to overall. When he exceeds a threshold it raises the strength level.
1-3 overall strength = level 1 strength
4-9 overall strength = level 2 strength
10-17 overall strength = level 3 strength

The declining returns come from the need to put more time into attaining each additional level.

The downside to this system is that there are clear ‘break’ points and until the character reaches the next ‘point’ there is no benefit. This actually fits the Marvel system, where characters are rated on a 1-7 scale, but wouldn’t be ideal for another game type, but it is a workable option.

Time Use: The most important part of the system is having a balance of how many turns the player has vs. the amount of things to execute. Some choice has to be created, while there shouldn’t be so many options that it becomes possible the player accomplishes nothing. Instead of just improving stats, I would like to add a few other options a player might pursue during a ‘turn’.

Currently, I’m trying to add work, relationship building, and plot progression to things that take a ‘turn’. Work, making money, gets forced in PAC, and it may be the best option as items are kind of a side factor to the game, not the main feature, but I’m trying to balance it in such a way that it could be relevant. Relationship building was already done in PAF, so I’m not touching on any new ground there, but trying to think of ways to expand the option. Adding in plot points as time consumers is the tricky part and requires (I think) a more non-linear game and for the player to get ‘rewards’ from the plot so it feels worthwhile.

As an example, in PAF monsters attack the station at set intervals. I’m trying more for a system where the player elects to go out and fight the monsters (or not). However, wait too long and they might disappear, or maybe you’ll miss out on certain rewards, or some other effect to differentiate when things are done. Alternatively, the game could be designed in such a way that multiple paths allow the character to never engage certain plot lines, in this instance maybe there would be a way in PAF to never even engage the monsters.

Overall: I think of PAC and PAF more as ‘games’ then ‘stories’ (especially PAF with the combat) and it is the game elements I’m trying to discuss. I think the current system that has been used works quite well, that there is a lot room for improvement, but I’m sure many of the improvement ideas I’ve discussed will actually over complicate matters and worsen the game, so I’d like to hear other people’s thoughts on what improvements they think can be made to the design.


  1. There are a lot of more 'fluffy' areas which are arguably more crucial to success in this genre than adjusting the mechanics. For example:

    Transparency - communicating to the player what each stat achieves, and how likely it is to be useful / what it's useful for, so they can make their own choices about how to progress. WITHOUT seeming forced, breaking immersion, or popping up a tutorial screen.

    Hidden scenes - I loved PAF and PAC, but I did find that I spent most of my time doing circuits of the map, checking to see if there were any special scenes at that time of day. (Train -> circle map -> Train -> circle map -> repeat). Again, this is a case of communication. Can you give enough hints to help guide the player without forcing them to check every room? Perhaps by letting them eavesdrop, receive a note, etc...

    Choices - how do you give the player meaningful choices when so much is stat-driven? Part of this comes back to transparency, making sure that if they want to chase girl X, they have a vague idea of how to do it (perhaps after asking her friends). But perhaps there's also room for a few scenes with X - triggered by good stats - where she asks you to make a choice (not necessarily pass-fail!) which affects her personality later.

    Temptation - let's say I chose to go after girl X on day 1 or 2 - right at the beginning. Am I now going to spend the next 23 days going through the motions, ensuring my relevant stats are high enough? That's pretty boring! So as the game progresses, you need to try and supply new interests - which I still have a practical chance of achieving - to tempt me off my path. That way the player's kept squirming!

    Like I said, these are all "fluffy" rather than mechanical factors. However, I think if you got these right then they'd make a world of difference to the game.

    Also, while in AIF these are BBBen games, this style in general is *really* common in Japanese virtual dating games. You might want to try a few out, as research.

    1. There are a lot of more 'fluffy' areas which are arguably more crucial to success in this genre than adjusting the mechanics

      I don't disagree with this at all, but what I'm trying to discuss here is simply the mechanics as they form the base/start of the game. You can lay out the mechanics before you even start designing characters/writing plot, even though the latter are indisputably more important (you don't have to do it in that order, but if you are designing more of a 'game' than a 'story' I think it is important to do so).

      Transparency: This is a good point. Were the uses of the stats in PAC/PAF unclear?

      Hidden Scenes: I think BBBen was criticized for this, I actually think it was the only criticism I sent him. Putting in notes/messages is doable, but can be rather complicated depending upon the events that could cause it.

      I think an easier way to go about it is just not to do the hidden scenes. Have no magic moments were upon entering the library with the right conditions a scene occurs. An easier, and I believe less frustrating, way would be for all such to follow other events (for example, the scene occurs after studying in the library, or the scene occurs after talking to the character).

      Choices: Transparency of what the stats do is quite important. As for how to give meaningful choice, it kind of depends upon how you view the issue. In a stat based system the player controls the player character, and the surrounding world reacts to them. In the choices system you are discussing (there are many different types, so sorry if I got it wrong), the player’s choices don’t really alter the character controlled, but the NPCs and how they perceive him (did the player get me what I want, for example).

      I prefer the former, though nothing prevents both from being done.

      Temptation: This goes to game style and depends upon what the author is gearing toward. In BBBen’s games there is a general harem theme wherein the romance options don’t’ really care if you sleep with other girls (either they never learn or they become involved). This is less than ideal, but from a programming stand point it is probably the best case scenario. A game could be done where it was programmed after sleeping with more than one girl A) they are unaware B) they are aware and angry C) they are aware and into it.

      I don’t think option B adds anything to the game and is incredibly difficult to program unless you want a very limited female cast (if there were say three romance options than I could see it being done). Also, if the games we were talking about were more ‘romantic’ I could see why it would be important, but that’s a style choice.

      Hentai Dating Games: I’m familiar with the genre and believe BBBen was actually trying to play homage to it (or maybe just Japanese hentai in general). I actually see the AIF design and the hentai games as very different. My experience with the majority of the games boils down to two designs.
      A: Occasionally the player gets a choice of which girl to pursue; generally you want to just choose the same girl every time, though there may me a perfect path to get all the girls.
      B: The player is given a variety of breaks in which they make a choice. Sometimes the choices are very obvious, for example which girl to help, and sometimes completely arbitrary (go to the gym or music room).

      I think the stat system is fairly different from those, especially if combat gets added in.

    2. "I think the stat system is fairly different from those, especially if combat gets added in."

      I'm no expert, but there are certainly eroge games that involve stats and/or combat systems. True Love would be an example of the former, and Princess Waltz of the latter.

  2. Beh, the hentai world is very wide and only little has been released in English.

  3. In software design, choice is not always a good thing. Too much and people will "shut down". This is counter intuitive because everyone says the want MORE choice. It is like coffee. Everyone says the want a rich, fullbodied, dark roast coffee. When you look at what people buy, they get weaker, cream and sugar laden coffee.

    I would hazard a guess that most players do not want choice, but instead really want good story, good characters, fun game play, interesting sex scenes, etc. All of those things can be achieved without much real choice.

    The illusion of choice is very important. For a masterclass on the illusion of choice, check out "The Stanley Parable" on steam. It will show many ways to make the illusion of choice. "Mass Effect" is another good one.

    1. On a different note, I'd also suggest taking a look at "Long Live The Queen", which is a Western visual novel (of which there are a fair number, both commercial and non-commercial) that does a lot of what you seem to be describing in terms of gameplay (ie. multiple stats, with the extent to which each is trained effecting how events play out).

    2. This is counter intuitive because everyone says the want MORE choice. It is like coffee. Everyone says the want a rich, fullbodied, dark roast coffee.

      But some people do like dark roast coffee and some people do like heavy choice in games. I tend to find the illusion of choice extremely aggravating.

  4. I have grave doubts about the stat system you describe. It sounds like increasing complexity simply for the sake of added complexity, without it having any real benefit for the player's experience. In this case there's no obvious way in which the extra stats and ranges improve the experience of actually playing the game, which makes me wonder why they're being included.

    I also agree with Anon above; too much choice can be a bad thing sometimes. This is especially true when the choices (and their effects) are not obvious to the player.

    1. It sounds like increasing complexity simply for the sake of added complexity, without it having any real benefit for the player's experience.

      That very well may be true and I do worry that a lot of the things I'm trying could make the game worse. The benefit to the complexity is, for me, personal perception as I kept thinking in PAF 'I wish there were more layers to this. I will note though that a dry discussion of mechanics might not do it is as much justice as an actual system, but it is something I'll keep in mind.

    2. There is a confusion of terms here. There's "depth," which is the additional layers of meaningful choice you describe, and there's "complexity," which is the mental burden the game puts on the player, restricting usability.

      Depth is good, measuring how much the player can get out of the game. Complexity is bad, restricting the player's ability to play the game like a noose.

      Getting the maximum amount of depth out of the minimum amount of complexity is an important component of what the gamer wants. For instance, if there is *never* a time when, say, you use strength but not fighting ability, maybe the two stats should be conflated, since all they're doing is adding one more useless nitpicky thing for the player to keep track of instead of actually playing the game.

  5. "I find the concept elegantly beautiful in eliminating many of annoyances I have with other AIFs and it is a design I’ve been trying to emulate and improve upon."

    I don't think there's much doubt that PAC was a conscious attempt to emulate the style and gameplay of Japanese bishojo games (the readme says as much). Personally, I was strongly reminded of True Love (which incorporates a number of RPG elements) in terms of gameplay and Divi-Dead in terms of plot. However, I think it's a good idea to think about the differences between the visual novel format and a traditional AIF system and, just as importantly, the different expectations the player has for each.

    As a very basic example, visual novels force the player to 'click' through each line of dialogue, partly because there's a limited amount of space for text on each screen and partly because of the need to synchronise each line of dialogue with the spoken line. The former is seldom going to be an issue with traditional AIF, unless you're trying to break up a particularly large block of text. The latter is only an issue if you've gone to the trouble of recruiting voice actors, which most AIF games don't. Additionally, most visual novels give the player various options for skipping to the next decision point, making it more convenient to replay the game multiple times, which is something that the traditional AIF can't do so easily. Consequently, I don't think there's much reason to copy that particular element, other than to evoke the 'feel' of a visual novel.

    "The player has a set number of ‘turns’"

    I actually found this to be a problem because it locks the player into a rigid schedule and thus limits the player's ability to advance the plot. Plot events happen because it's a particular turn, not because the player actually does something, which essentially makes the player subservient to the plot.

    It also doesn't mesh particularly well with having an explorable map (leading to the circuit problem mentioned above). I'd be much happier with having the events happen directly after a 'training' action, or even just be automatically triggered as a cut scene wherever the PC happens to be.

    "The primary choice is to ‘train’ one of three stats."

    The reservation I have about how this was employed in PAC and PAF is that it's a 'meta' choice. That is, it's a choice that the *player* makes, rather than a choice that the player makes *for* the PC. Consequently, it doesn't encourage the player to identify more closely with the PC, resulting in less immersion.

    "More stats: This is the most obvious improvement"

    I wouldn't necessarily describe making the game more complicated as an obvious improvement.

    As has previously been mentioned, providing the player with too many choices can induce paralysis, especially if they have no means of determining which is the 'best' choice for what they want to achieve in the game. Giving the player that information probably means making the background mechanics more obvious (eg. explicitly telling the player that they failed a durability test) with the consequent risk of diminishing immersion.

    1. (continued)

      "Many of the ‘traditional’ AIF elements are absent"

      This ties in with the differing expectations of player agency between the two formats. In a traditional visual novel the player doesn't expect to have that much agency because, as the name suggests, it's more of a story than a game. In traditional AIF there's an expectation that the player will be able to do a lot more (partly because of the text parser), such as explore the game environment and interact with it (the traditional AIF elements you're referring to).

      That kind of local agency makes up for the fact that the plot of most AIF is linear and the player is unable to alter its course (although ideally the game will at worst still provide at least the illusion that it is the player who is driving the plot to its conclusion, and at best give the player meaningful choices that actually decide the outcome). If you take that local agency away, then the game becomes much closer to CYOA (which isn't a bad thing, but does require different considerations than traditional AIF).

      Theoretically, you can use a combat system as a substitute for local agency, but even PAF's combat system (which is the best I've ever encountered in IF) became tedious after thirty or so battles (although the fact that there was no clear reason why the PC was fighting exacerbated that for me). Additionally, if the combat system involves random chance (which most do), that decreases player agency (not to mention encouraging save scumming, using undo, etc.)

      "I think of PAC and PAF more as ‘games’ then ‘stories’ (especially PAF with the combat) and it is the game elements I’m trying to discuss"

      If you're intent on emphasising the game elements even more than PAC and PAF did (and decreasing the level of interaction the player has with the game environment and the NPCs), I would strongly suggest not using a traditional text parser based IF system. Why try to hammer a square peg into a round hole? Something like Twine, or especially Ren'ai (since it's a visual novel engine), would be better suited in my opinion.

    2. Plot events happen because it's a particular turn, not because the player actually does something, which essentially makes the player subservient to the plot.

      Well, good (if I understand you correctly). There should be a world outside the player and that is usually exemplified by time constraints. They don't have to be arbitrary. You have five days to complete the test/get ready for the football game/save the world from an asteroid. Too much constraint is bad, but I think too many games just give the player all the time in the world to do everything possible.

      I'd be much happier with having the events happen directly after a 'training' action

      I agree with that.

      That is, it's a choice that the *player* makes, rather than a choice that the player makes *for* the PC.

      You've brought this up in other threads and while I understand the difference you are pointing out, I don't think it is something that would ever bother me.

      I wouldn't necessarily describe making the game more complicated as an obvious improvement.

      My word choice could have been better here. What I meant to say was that if you were going to attempt to improve upon BBBen's model, the most obvious place to start thinking about where improvements could be would be the three stat base.

      such as explore the game environment and interact with it

      This might be what differs the most between me (and maybe some others) and what counts as immersive in AIF. I never find moving around the map immersive, especially when there is in an inevitable destination. I also never liked have to search through multiple cabinets, beds, wardrobes, etc. to find an object necessary to advance the plot (now if I had to choose between two different objects, that would be important).

      became tedious after thirty or so battles

      Thirty battles is pretty good.

      Why try to hammer a square peg into a round hole? Something like Twine, or especially Ren'ai

      I've looked at Twine and didn't care for it. I'd never heard of Ren'ai until you mentioned it, so I'll look through it. I also considered RPG Maker, but even though I'm discussing mechanics here there is meant to be a lot of text.

    3. For Ren'ai read Ren'py. I was having a blonde moment.

    4. "Well, good (if I understand you correctly)"

      I suspect that you don't. I wasn't talking about time constraints in the sense of deadlines. If the PC has five days to achieve a particular goal, that adds dramatic tension. What isn't quite so fine is if it's impossible for the PC to achieve that goal before Day 5 simply because the plot says it is.

    5. What isn't quite so fine is if it's impossible for the PC to achieve that goal before Day 5 simply because the plot says it is.

      Okay, I could see why that would be bad, but how often has it come up? Are there games that create impossible to achieve goals? Maybe I still don't understand your perspective or reference.

    6. I think that both PAC and PAF suffer from this to a certain extent. For example, in PAC the PC is ostensibly invited to the island to help investigate the mysterious goings on, but there are very few ways he can actively pursue that goal. Instead, events happen around him, and the weird atmosphere builds until it finally climaxes on the final day. With respect to that part of the plot, the PC isn't active or even reactive. He's just a passenger admiring the scenery. But that's balanced out by the fact that he can actively pursue the various female characters (in AIF terms the major part of the plot).

      However, the most egregious example I can think of occurs in a RAGS game called "Dangerous Assets" where there's an object that's freely examinable from the very start of the game, but it's not until the final day (ie. when the plot says so) that the PC thinks to turn it over so that he can discover what's written on the back.

    7. Oh, and what's written on the back is the key to the entire mystery that the PC has been trying to solve.

  6. I enjoyed PAC, but the main issue I have with any stat/training-centric game is that they tend to become exercises in number crunching, where the player skims over the story in order to chase the next available +1. Immersion is harder to achieve because the underlying mechanics are right in the player's face.

    Also, the replayability factor--which is often cited as a strength of these sorts of games--is a bit overstated IMO since most of the time the replay consists of skipping over the text and brute forcing through to the next training junction.

  7. Beh, if someone wanted to use rpg maker there was an incredible opportunity on humble bundle, but now ended.

  8. If you remove all the normal AIF interactivity and make a stats-building game, then your stats system has to be interesting. I like the stories in PAC and PAF, but I found the stats building to be very tedious. The stats system was too simple with no big challenge to optimize. It took too many commands to upgrade a stat. You have to wake up, go to the right room, choose the stat to upgrade, go to another room,... If you play games like True Love or Princess Maker or even Cookie Clicker, they make the stats system to be the center of the game. Making a choice is one-click. They strip out the narrative elements that get in the way of the stats building.

    You don't need a lot of stats. You need choices.

    Stats should be in opposition to each other. You can't improve one stat without getting worse at another. That makes the choices more difficult and interesting.

    Also, things should unlock at different skill levels. More training opportunities. More things to buy. More effective training opportunities.

    Many training games use an exhaustion mechanic. Training a stat makes a character tired. If they get too tired, they become ineffective. It forces players to have to weigh the benefits of intensive training to unlock a new training type early or pacing things out but unlocking things later.

    The stats should also tie into meaningful effects. If your stats aren't at a certain level at a certain time, you will miss the chance to do certain things. There has to be a reward for being better at training. With PAC and PAF, you made one choice at the beginning. You choose which girl you want. Then you train up the needed skills. Then done. There's only one choice the entire game. In a good training game, there must be constant trade-offs at every turn. It requires a lot of content. You have to design these trade-offs early. Bad RPGs have many stats that don't do anything. What is the need for a diplomacy skill if you use it only once? It is better to have fewer skills, all of them meaningful, all of them leading to interesting trade-offs in the story.

    The stats system you describe has complexity but no choices. You should focus on the trade-offs. That's where the interesting stuff happens.

    1. I really hate trade-offs, energy meters, all that other stats-game bulldung, and can't agree. Fact is, he's *not* making a stats-building game, and adding the worst, most time-waste-y elements of that game will only bog it down.

      The thing that lets me enjoy BBBen's games, in spite of their "simplicity," is that, rather than only rewarding the player at the very end, the player gets to enjoy little things along the way. The reason people run the ring in PAF is because they *want* to experience the little cutaways that make the stat grinding bearable, to learn more about the characters and watch them interact with each other and the player.

      I'm glad you're keeping that, and I'm glad that you're removing the "ring-around-the-rosy" aspect by just having it trigger once it's "unlocked." I just hope there are plenty of such scenes, because the stat grinding is a means to achieving them rather than an end in and of itself.

    2. Are we actually mostly agreeing? I liked the story aspects of BBBen's games, but I found the grinding aspect to be off-putting. I could only play each game once or twice because I couldn't bear the idea of having to do all the repetitive wandering and stats building all over again just to see the other content. With PAF, I waited until the walkthroughs came out first because I didn't want to spend a lot of time figuring out the stats system and triggers. My suggestion to Archer Fifteen was that the stats building should be "painless" in that it should just be a couple of clicks or a single command to do the actual stats building.

      By trade-offs, I mostly meant that the game should provide meaningful trade-offs in the story. The stats should be reflected in the game, and the game should play differently depending on the stats. I was suggesting that Archer Fifteen spend time figuring out how the different stats will cause different things to happen in the game. Then, the player has to make choices about what they want to have happen. That's the trade-off.

      We only seem to disagree on whether the stats building itself can be interesting. I think it can be. The important thing is that players shouldn't need to play "optimally" to "win" the game. The trade-offs in making different choices to build up stats shouldn't be win vs lose but should be one interesting direction vs another interesting direction. In a superhero training game, even two stats can be interesting: fighting ability and relationships. You can train up your fighting or you can hang-out and build relationships with the rest of the team. You can be a great fighter who fights evil alone. Or a lousy fighter who is friends with everyone. Or a great fighter with friendly teammates. Or a poor fighter with no friends (perhaps, in this route, you get lured to the opposite side?). During the game, there will be various events that trigger differently depending on the stats. Your teammates might be less friendly if you don't pull your weight in the fighting? People might become jealous if you're too good a fighter?

    3. If you remove all the normal AIF interactivity and make a stats-building game, then your stats system has to be interesting.

      As clarification: I'm not talking about making a pure stats building game. I do think such a game could be done in AIF, and it might actually be easier than a hybrid design. Slavemaker did a lot of what you are discussing, but I want more of a narrative than that game had.

      You have to wake up, go to the right room, choose the stat to upgrade, go to another room

      This is a good point and I had a similar issue with PAC/PAF, but for me it was an incredibly minor complaint. The issue that concerns me is how to do this and keep immersion. Immersion isn't something I had put a lot of thought into, but this thread seems to indicate people care about it a lot. Maybe have a training room or something where everything is located.

      You don't need a lot of stats. You need choices.

      They tend to go hand in hand. You can have a lot stats without choices (which would almost certainly be a poor game design), but it gets difficult to make a lot choices without many stats.

      Of course how you and I define 'a lot' might be very different. To me PAC had 'six' stats: the three main and then three that were of the true/false variety for whether you had given a character a certain item. If I had around ten to twelve, with a mix of stats that develop and true/false variety I’d consider that ‘a lot’ for an AIF game.

      Many training games use an exhaustion mechanic.

      Exhaustion is something I have in, though I think its usefulness varies based on game design. Many of the games I have played with the mechanic make becoming exhausted (or sick) so detrimental that no one should ever logically risk it.

      The stats system you describe has complexity but no choices. You should focus on the trade-offs. That's where the interesting stuff happens.

      I think many of the games you mentioned would just sound complex without seeing the actual game play. If I just tried to describe Princess Maker it wouldn't sound that interesting, but likely incredibly complex.

      I do think declining returns accomplish a lot of what you are suggesting. I could, as an example, do the system so that every time the character trains strength, charisma goes down. Or I could have that every time the character trains strength it becomes worth less than the last time the character trained strength, thus making the use of the 'turn' on the strength versus charisma a tradeoff.

      Having trainings have detrimental in addition to beneficial effects would also be substantially more difficult to program in (though that is only relevant to the creator).

      I just hope there are plenty of such scenes, because the stat grinding is a means to achieving them rather than an end in and of itself.

      I agree that this is a necessity to have some 'rewards' and story progression along the way (unless it was a rather short game). There are kind of two ways that it can be done. In the easy way, when the player reaches a certain threshold(s) an event triggers (simple example: character reaches five strength, the cheerleader comes in and compliments him on his appearance). In a more complex system multiple requirement play into the scene (elaboration of above example: character doesn't train strength until late in the game, cheerleader comes in and asks him what took him so long, if character is already in a relationship with the cheerleader's rival she comes in and snidely asks him if all her rival cares about is muscles, if character is already in a relationship with the cheerleader she comes in and a sex scene occurs).

      The latter I think makes for a deeper game, but is obviously more difficult to implement.

  9. Personally I just enjoy playing the game, if I like it I play it and recommend it, if I don't, then I delete it and forget about it. It doesn't matter on the format, it doesn't matter if its got graphics, or text or video, or if its adult or PG13, Its all about the experience of playing it.

    I like how there is 2 tiers of gameplay to PAC and PAF, that you can romp through the game, see some graphics and end it winning in the end somehow without the often confusing syntax. But the seasoned player that realizes there is something more, can replay and find the little nuggets that allow for enhanced experiences and get even more out of the game then the previous player. Its almost the best of both worlds. I don't have to "look under bed" to find a specific key item that leads to a specific "ask so n so about something or other" But at the same time I can enhance the play experience with my knowledge of traditional AIF plot arcs and find more content that way.

    The balance is pretty hard to beat.


  10. This comment has been removed by the author.

  11. I wrote a big, long response to this thread and then blogspot just discarded it, which is a real kick in the teeth. Anyway, this is probably a lot less detailed and well expressed than the first time, but maybe more succinct. Also I'm in a more ill-tempered mood now, thanks to blogspot.

    I'm aware that some players feel the need to run around and find relationship scenes in every time period. I feel like PAC and PAF are getting conflated here, however, possibly because Super PAC was just released and also because some players who had played PAC went into PAF playing it the same way. Actually, the scenes in PAF (largely in PAC, too) are designed to appear in places you're largely going to walk to anyway. In PAF all the scenes are in the main ring, the training room or Kenji's bedroom.

    It's not totally perfect because I put a few slight outlier scenes in there, but I think if you tried playing PAF by just building stats and letting the game play out you wouldn't have any real trouble with missing those scenes. The biggest problem was probably just the ring shape, which meant you could choose to go east or west, and maybe miss a scene. I don't think a lot of players actually had runs through the game where they missed scenes and failed to get a girl, however; I think most people had an issue with this were just worried they would miss scenes and acted preemptively.

    If you try to solve that issue by making all the scenes appear immediately in response to flirting actions, however, then you do really limit the types of relationship scenes you're writing. Either you have to have them all happen in the same, static location, or you need to find some awkward reason every time to have the characters wander around the place. I much prefer to write scenes that have a more organic, unpredictable quality. I also can't really speak for players on this point, but I do think there's some joy to be had in happening across a new scene when you're walking down a corridor, with a couple of characters in a discussion that you overhear, or with someone getting dressed and leaving their door open, etc.

    The best way, I think, to resolve the issue is to just further refine the design of putting scenes in places where the player will be going anyway, and not making them all that miss-able in the first place. It comes down somewhat to map design and somewhat to scene design, although I can't guarantee this will stop some players from going out and doing the rounds every time period anyway.

    1. By the way, I do think this is a really interesting thread, and I've been reading everyone's ideas with interest. I like to know what people take from the game because I really have no player data other than these sorts of testimonials and the beta process. Playing the game myself isn't the same.

      Anyway, the second thing I wanted to bring up was about the idea of needing temptations to veer off the main path, because otherwise you're just choosing a path at the beginning and then following it. I think this doesn't really fit the game design, because it kind of assumes that you already know what the paths are and how to follow them before you've ever played the game.

      On the first run of the game I find a lot of players (that have given me feedback) are finding their feet and getting acquainted with characters, etc. They don't have a really clear sense of the best tactics to easily get one of the girls - in fact I have had a fair number of reports of people playing through the first time and failing to get any girls.

      On replay, of course, things change and getting a girl becomes much easier. The goal-posts are moved at this point, however, and the harder objectives of combo endings, card collection and the objective of getting the harem mode/ending are introduced. That does increase the challenge. It would be easier to make things more unpredictable if there were some way to introduce more randomness, but unfortunately it's not that easy, as the 'undo' command prevents a lot of that.

      I've toyed with ideas like making relationship increases from flirting much more variable but also masking the exact score, so that you just know that the score has increased, and maybe the girl has gone from "not interested" to "somewhat interested", and maybe I'll experiment with that kind of thing a little. I did find it took a lot of people quite a while to get to grips with the way the PAF system worked, however, with a lot of people not realising how important it was to have a good eye score before giving a compliment, for example.

    2. Oh, one last thing! The text formatting with ellipses and a waitkey command breaking up lines


      like that actually serve three purposes. The first was just to attempt to mirror the button-clicking style of visual novels. The second emerged later in PAF, when it became a great way to make audio and images appear at the right time in a block of text.

      The third unexpected point, however, is enough that I would probably use this method (or a form of it) regardless of whether I was including multimedia or making a hentai-style game. That reason is breaking up the flow of text blocks, and making them feel more interactive and alive. Okay, not interactive in the sense of choice, just in the sense of pushing a button.

      It's a point that I wasn't so sure about at first; I wondered if people might just find it annoying to have to keep pressing a button through the scenes rather than just reading them like a book. I found I personally liked the feeling that things were happening in real time, and found the small bites of text far easier to digest, but I couldn't judge this for myself until I got player feedback.

      As it happens, those players that have actually commented on it have consistently liked it, and even been very enthusiastic about it. Most don't think about it, of course. Better still, there's something people aren't complaining about in the Pervert Action games - text dumps. And I put in loads of huge text dumps! The games are full of them, and I really think that if not for the line breaks heaps of people would be put off by that. I did get comments about them back when I was making the Crossworlds games, but not any more, even though my linear scenes are longer and more frequent.

    3. I think the game would feel more organic and unpredictable if the women weren't directly coupled to your stats. Instead of "if stat > 5, then you get a scene with woman A" it might be more interesting to have "on day 3, event X happens where woman A trips. If your agility stat is > 5, then you stop her from falling and your relationship with her improves." And you can have things like "on Wednesdays, she always struggles with her homework in the library. If your intelligence stat is > 7, then you can help her." The women become more like real people that you have unpredictable experiences with instead of "the intelligence woman", "the looks woman" and "the athletic woman".

      I noticed the text dumps. They were fine. Interactivity would be better, of course.

    4. PAC has relationships directly coupled to the stats, but PAF already changes that idea.

      Your proposed alternative sounds interesting, but also would make it difficult to juggle when and where the sex scenes crop up, creating as it does a much more restrictive plot flow. I think it's more appropriate to a CYOA style of game rather than a dating sim style. It can be done to a limited extent, of course, and actually PAF does have some scenes like that.