Breeding Season: Alpha Version 5.3 [Update 11/3/14]

[Ironvein]

The equation is:
stat change = (stat growth modifier + WIL and trait effects on that modifier) * (breeding modifier * STR and trait effects) * e^(-0.00690775527 * stat value) * (2 + 2*(partner's stat value - your stat value)/1000) * 75

Egad that looks overly complicated; how did you come up with it out of curiosity?

Yes, 0.001 is extremely low, but this is intentional: getting to 1000 is meant to be hard to nearly impossible. In the final game you will ideally be completing the story-line with monsters that have stats in the 600-700 range. Going any higher is for very dedicated people.

I'm of the opinion that the stat caps handle that particular desire just fine. It takes work to get those caps up. Well unless you wait for a store bought monster to have caps that high I guess. Another reason I think store boughts should only be 'level 1' creatures period (only new traits should be available as they are unlocked).

There is, I realize, a huge problem with my system right now: If a monster has 0 in a stat and a huge modifier (like 200+) it can hit 1000 in a single breeding. Meanwhile, having the same modifier but breeding at 995 it will take a huge amount of breeding to get you to 1000.

I was reluctant to make a system where the stat increases are based on "EXP" values, but you may be right in that using a point system in which the stat values are discrete instead of continuous and just requires ever-increasing amounts of "BP" to increase to each new stat level probably just makes more sense.

The BP example does not require making a new attribute to you stats; you just loop it until your BP is insufficient to raise the level and drop the rest.

Looking at it, I think you just overcomplicating it on yourself and it's bringing out all sorts of weird buggyness. Take a step back and remember the principal of KISS and you should do fine. Keep up the good work.

[Mechaphantom]

I've lost count on the amount of money I've donated and been looking at the list of features I wish to donate. Someone suggested posting images of the ideas for each.

The one I most interested in at moment is Neoteny. I all in favor of younger girls but how young? Are we talking young teens? Okay with that, younger? Getting uncomfortable. Some people use the term cub art of toddlers and the like, not okay with that. How young looking we talking? My cut off point is like 13, I find girls that young cute but not sexy and anything younger not cute and not sexy.

I also interested in your idea for bunnies, foxes and elves. I play a game tera online for example and used to seeing flat chested bunnies, or they going to be big breasted bunnies? I would be more willing to donate if I knew what I was going to see. For the monsters I'd like to see a quick stretch and maybe what perk they'd come with one harvested. There well come a point where you might have to invented more problems to warrant more monsters. I mean by that you already have a monster to make breading more often one to lesson discomfort and blaw blaw blaw, what other kinds of chemicals would we need and want? What would elf goo do? What would dragon spunk do? Do you have any ideas for those or making it up as you go.

I still interested in donating for features like.

1. Option to change clothing.
2. Option to have more humans sexed, I would really like to see the other two female cast members fucked. Their names escape me at moment. The assistant and debt lady.
3. I know it coming down the road but more rpg and story elements. Love the chat test and love when the debt collector comes to collect. More scenes like that.
4. I a big steam punk and machine fan, you once said you'd add golems. I'd love to see machine golems and doll chicks.

Lastly when I donated for player pregnancy it had been hinted and suggested there would be a female monster type in the works. My problem with current build is I get swarmed with male monsters and lacking female monsters. Player pregnancy spawning female types would be a great help to me. I got the girl pregnant a total of 6 times my last play threw had all the hidden traits unlocked but would love to have a chance of a female monster or even human type spawn.

For this month I had budgeted out about 150 for this game, and would be willing to spend it on mech type monsters and lesser amounts to my other ideas and some of the ideas already posted. I like monster pregnancy idea I the one who donated a little for that. I donated a lot towards some of the past features and helped with egg amation, bunny and a few others. Interested in seeing idea for fox girl, bunny and elf.

Thanks for the donations, I appreciate it immensely!

I'm still working on sketches for a lot of the different donation features, but here are a few (extremely rough) sketches. I was also struck by inspiration by the idea of "doll chicks" and drew a new monster idea. Note: these are super rough and not necessarily what the end product will look like.

As for what Neoteny creatures will look like, this is an example of a Kittengirl:

kittengirl.png

Quick sketch of a Bunnygirl:

bunnygirl.png

My stab at a steampunk robot girl:

robotgirl.png

A doll girl:

doll.png

This is the (unfinished) alarune:

alarune.png

I can definitely work on a male golem as well.

Also, there's a ton more storyline-type stuff coming. Pretty soon I'm going to focus primarily on that for a little bit. Sex scenes involving Margo and Delilah will absolutely be a part of the storyline (though, those two characters in particular will be fairly difficult to get in bed ).

Woh i very pleased with most of those. The neoteny looks great, just on the border line of to young for me but I'll take it! As for other ideas I'd donate for most of those images. I really like how you took our idea for butt poses to heart and going to have a few monsters facing backwards, front views are great but for a few would love to see some jiggly asses.

I play a cell phone game that has a lot of the monsters I come to love and when I was thinking dolls and robots this is the kind of thing I was thinking but love your drawings too.



Here some mix robot and golem like chicks from montowers


Here their version of living dolls.

Here an iron robot girl

[HartistaPipebomb]

The equation is:
stat change = (stat growth modifier + WIL and trait effects on that modifier) * (breeding modifier * STR and trait effects) * e^(-0.00690775527 * stat value) * (2 + 2*(partner's stat value - your stat value)/1000) * 75

Egad that looks overly complicated; how did you come up with it out of curiosity?

It's not nearly as complicated as it looks. If you break it down it's just (Growth Modifier Stuff) * (Breeding Modifier Stuff) * (Exponential Decay Function) * (Stat Comparison Function) * 75.

I just solved the decay function for t=1000, f=0.001 to get what the decay constant should be. The stat comparison function is pretty simple as well, it just ranges between 1 and -1.

The ickiness I had before with the dividing by zero and the messed up stat growth was when I was trying to come up with my own function to generate the diminishing returns, I realized that it was much, much cleaner and easier to just use the exponential decay function.

EDIT:
The BP idea is good, though, and I might try it. It's a lot less computationally efficient but that matters literally zero in this situation .

[Ironvein]

It's not nearly as complicated as it looks. If you break it down it's just (Growth Modifier Stuff) * (Breeding Modifier Stuff) * (Exponential Decay Function) * (Stat Comparison Function) * 75.

Isn't Exponential Decay primary used for determining half-lifes? No wonder it drops like a stone (and why the biggest boost would be near zero). Despite the simplicity, that's the main problem with it. Needs to be toned down to say a Geometric or Arithmetic Decay (which isn't as severe).

[Ironvein]

EDIT:
The BP idea is good, though, and I might try it. It's a lot less computationally efficient but that matters literally zero in this situation .

Not computationally efficient? Check this:

// Recursive Function equalivent

double GrowStat(double statToRaise, double modOfStat, double buildPoint)
{

// Basic Use: DEX = GrowStat(DEX, ModOfDex, (MateDex * (MateSTR/100)); // as per my description

If (buildPoint >= statToRaise)
{
buildPoint -= statToRaise; // pays the cost
statToRaise += modOfStat; // adds the increament
statToRaise = GrowStat(statToRaise, modOfStat, buildPoint); // repeat til fail
}

return statToRaise; // all done

}

[HartistaPipebomb]

EDIT:
The BP idea is good, though, and I might try it. It's a lot less computationally efficient but that matters literally zero in this situation .

Not computationally efficient? Check this:

// Recursive Function equalivent

double GrowStat(double statToRaise, double modOfStat, double buildPoint)
{

// Basic Use: DEX = GrowStat(DEX, ModOfDex, (MateDex * (MateSTR/100)); // as per my description

If (buildPoint >= statToRaise)
{
buildPoint -= statToRaise; // pays the cost
statToRaise += modOfStat; // adds the increament
statToRaise = GrowStat(statToRaise, modOfStat, buildPoint); // repeat til fail
}

return statToRaise; // all done

}

This does not make it more computationally efficient . I'm nit-picking though, it's not an inelegant solution.

[Ironvein]

Whoops. Doesn't do negative that way.

... wait, no it does

initial call for negative would be DEX = GrowStat(DEX, (ModOfDex * -1), (MateDex * (MateSTR/100)));

Ought to work that way, just need a check of the breed modifer before calling. Should be enough to get the jest, right?

[IrrelevantComment]

I'll admit that paying off Delilah is much easier now; but I wouldn't call this version complete. There are still a number of bugs to deal with and the determining what really needs to be tweaked to the game balance because it's too easy.

Have you unlocked all the traits or paid off Delilah yet? Those seem to be the biggest goals this time around, maybe one of those open the 6th monster (if there really is one; I hope, but am skeptical).

I meant that I'm done playing thgis version, I have nothing left to achieve, not that the game was finished, sorry for not being clear. And yeah, got all the traits.

[Ironvein]

After opening some other pens, I'm now beginning to see other consumables available for sale, but Cat and Harpie still comes up most of the time.

[Oringlas]

WARNING - MATHS AHEAD

My take on a decay-type function - use a sine curve.

curvy

curve.gif (5.99 KiB) Viewed 6280 times

function used is (1 + cos ( pi() * (statvalue/1000) ^ 1.33 ) ) / 2


Putting this together I played some more with the function and came up with an alternate

also curvy

curve2.gif (6.03 KiB) Viewed 6280 times

1 - (1 + cos( pi() * ( (1000 - statvalue) /1000) ^ 2 ) ) / 2

[stellated]

(stat growth modifier + WIL and trait effects on that modifier) * (breeding modifier * STR and trait effects) * e^(-0.00690775527 * stat value) * (2 + 2*(partner's stat value - your stat value)/1000) * 75

This has some undesirable anomalies as you guys have discussed. One particularly unwelcome feature is that a very strong growth modifier interacts negatively with a negative breeding modifier. This is counterintuitive from a gameplay perspective. In order to derive a good formula, assumptions need to be made about the kind of gameplay desired. We can ask questions about what kind of gameplay is desired to calibrate our design. Let's start with a simple question: How many times should it take for a monster with x breeding modifier and y str modifier to raise a 0 stat monster with a growth modifier and b will modifier to 1000 stat? Should the partner's stats affect the breeding process, or just their mods? Is it optimal to breed monsters with similar stats together or a very large difference in their stats?

I actually kind of dislike the stat system. It is very number heavy but the optimal strategy is to breed monsters until you have a few monsters with +'s in all stats, and then breed with those. It's not hard to get very high stats once you have a good set of monsters to work with and the stat system becomes irrelevant. The most relevance the stat system has is whether or not you have + breeding mods at all and at the very start of the game where you need a few monsters with + Stamina in order to get your farm running efficiently. It is really complicated and seems to be a source of more design problems than fun gameplay.

I would really like to see clients desire personality traits,appearance traits, special physical abilities, and special mental abilities. Breeding monsters to get a certain phenotype is a lot more fun than breeding them to get breeding modifiers and optimal pairings for stat growth. The game already has a rich set of female appearance traits. If it were enriched with personality traits and a wider diversity of male traits, then those could form a much more satisfying basis for client breeding strategies. Stats could be replaced by physical and mental traits that can be gained, lost, and improved through use and strategic breeding. These could complement a class of traits that can only be gained through inheritence by birth.

Just my two cents. I'm sure a lot of people like the number system the game uses now. It's not that big of a deal. It's just not very compelling to me. Most of my strategy is spent trying to get particular looks once I can easily fulfill my debt requirements.

[Werefrog]

HOLD YOUR HORSES (and other animals)

I have something to say regarding the raging math discussion. I was actually thinking about the stat increase function yesterday. I was trying to make a function I thought was being used (I was a bit off), into a function I though was intended to be used. Well that doesn't really matter, the result was an apparently unsolvable equation. What matters is that I should be able to devise a function that fits the intentions of the author. I have some free time tommorow so I will look into it. I expect it will involve logarithms.

BTW, I am against using some kind of leveling system. It's a valid workaround I guess, but this is nothing that can't be solved by the POWER OF MATH.
And Oringlas, this isn't about the decay function. The decay funtion used by HartistaPipebomb is just fine. It just needs some integral love.

[Nuffy]

Quick suggestion. Lots of traits have counterparts (SFR & AYSYNA, Gentle & No Downsides, etc.). Personally, I'd like a counterpart to Albino. I think something like black hair/fur and purple eyes would be more aesthetically pleasing. Just a thought I had.

[kalebon]

@HartistaPipebomb: Just wondering if you have a site or blog that we can go to?

[LoneWolf]

I spent a while playing with formulae for the stat decay rates, and made several discoveries:

A) If you stick with Hart's endpoints, any form of exponential or geometric decay ends up looking exactly the same as Hart's formula. (Which, in hind-sight, is actually kind of predictable. Unfortunately, I still took longer than I should have to realise that.)

B) Hyperbolic or logarithmic functions blow up in your face at x=0 (and, for logs, also at x=1), and:

C) They turn out to be even worse than an exponential decay anyway.

Fortunately, I eventually realised that what we're looking for is not, actually, a general-purpose decay curve; it's a good approximation of a decay curve over a clearly defined range. And, if you only care about values within a specific range, polynomials can do anything.

Half an hour or so tweaking parameters gave me this:

DecayGraph.png (9.13 KiB) Viewed 6134 times

HartistaPipebomb's original formula is in yellow.

Pink is designed to match Hartista's original endpoints, but smooth out the rate at which your rate-of-gain changes rather than having the values dropping quite so dramatically early on. (Growth is around 3x faster at x=300, 4x at x=400, 5x at x=500 and 6x at x=600. The advantage peaks at 6.85x at around x=750 and drops rapidly during the 900s)
The equation is 8.65*10^-10 x^3 + 2.6456 * 10^-6 x^2 + 2.7796 * 10^-3 x + 1 (Parameters chosen such that the equation is equal to the original formula at x=0 and x=1000.)

Since I personally feel that it's a little too easy to gain stats early on, I came up with a second alternative as well. Blue would be my suggested formula, and starts out at only half of the original values when x=0 but still matches the original formula at x=1000. It would lead to slower stat gains than the original formula up to approximately x = 185 and then faster gains through mid- to end-game. (Growth is roughly doubled at around x=400 and x=800, peaks in the middle of those at around 2.5x the original rate, and is actually slightly slower between 900 and 1000)
The equation is 4.35 * 10^-10 x^3 + 1.369 * 10^-6 x^2 + 1.433 * 10^-3 x + 0.5

Both equations behave well back to at least -200 (and probably lower, but I doubt that will be necessary very often...) and the turning points for both first- and second-derivatives are either >= 1000 or don't exist (I.E. growth rates don't start increasing again within our chosen range). (And they're both 'computationally efficient' as well. )

[Mechaphantom]

All that math goes over my head. Hope you egg heads figure that out! Now back to the porns! :0

[Oringlas]

@H-Pipebomb: If you haven't already, it might be easier to define functions such that you can write:
statchange = growth_mod(...) * breed_mod(...) * decay(...) * compare(...) * 75.

From a design standpoint - why the need for the stat compairison? Breeding modifiers give the player an rough idea of the relative strength of any changes to stats.

I do like Ironvein's suggestion of capping (or reducing) the strength of store-bought monsters (say, after a certain number of satisfied clients), which means that the good breeding stock on your farm is what you have bred yourself.

Would a numerical indecation of breeding modifiers work (say a ranking of -5 to +5) be useful, or would it give the player too much information?

[Werefrog]

I'm back! And I have the function with me. After fiddling around a bit with a differential equation based on HartistaPipebomb's original formula, I got this:

newStatValue = ln(q*k + e^(q*oldStatValue))/q

k - what I call RAW BREEDING POWAH, that is breeding modifier timez whatever it should be timezed by, in other words (Growth Modifier Stuff) * (Breeding Modifier Stuff)
q = 0.00690775527 (positive btw)
Just to make it clear, ln() is natural logarithm. A logarithm with base e. (in case our notation differs)
Note that the result is not the stat change but the final stat value.
I intentionally omitted (Stat Comparison Function), if you want that to be a factor, you can include it in k in some way. Simply multiplying should be fine.
Oh and stat decreases (negative k) should work just fine. Of course you have to watch limits of the stat value and the argument of logarithm must be greater than 0.

Alternatively, you can store the total amount of k a monster has received and calculate the stat as:

newStatValue = ln(q*k + 1)/q

The formula is simpler and should be a bit faster because you don't have to do e^(q*oldStatValue) which is essentially an inverse function that calculates the total amount of k the monster has received so far anyway (except multiplied by q).

This could potentially quite dramatically change the game balance because it fixes all the unintentional behaviour. (monsters with low stats getting enormous boosts, stat increase and stat decrease being calculated differently etc.) So I suggest some tweaking.
If you have any questions, comments, or ever want to change the decay function, just say a word.

[LoneWolf]

The argument of logarithm must be greater than 0.

This is the part that worries me.

You're probably safe 99.9% of the time, but single-digit stat values and k = -150 to -200 will make that blow up. From what Hartista said earlier in the thread (as well as in-game observations) modifiers in that range are high but definitely achieveable. It'd mean a monster with a large negative modifier breeding with one that only has 0-5 points in that stat - an edge case, definitely, but probably not one that we can just ignore.

[Werefrog]

This is the part that worries me.

You're probably safe 99.9% of the time, but single-digit stat values and k = -150 to -200 will make that blow up. From what Hartista said earlier in the thread (as well as in-game observations) modifiers in that range are high but definitely achieveable. It'd mean a monster with a large negative modifier breeding with one that only has 0-5 points in that stat - an edge case, definitely, but probably not one that we can just ignore.

That is not a problem at all. It just needs to be checked before using the function. If the argument is less than 1, the result would be negative anyway. Then the stat value is 0. And should he choose the second option, the total k can't be below 0, so the argument would always be at least 1.