Every 7 kittens born, 6 are normal size and 1 special size!

[Jiyulie Resident]

I think different avatar has different data. After looking at my pedigree, I realized that my special sizes tend to come from special sizes parents.
Like from my teacup couple, I've gotten bigger de big, mega, toy and teacup and a petite. And it only has 7 offsprings so far.
Same goes to my 2 other pairs of toy couple and teacup-toy couple.

Therefore I conclude that for some, the sizes are "bred", some has a pattern and others are just random. As long as my couples are of special sizes together, I get more special sizes than normal.

So I think size variation is dependent on the avatar, the time they join kittycats and what system they get for breeding.

[Tad Carlucci]

Actually, the results depend upon the human examining the results drawing incorrect conclusions from limited data.

Size is random. KittyCatS has said so, repeatedly.

Size does NOT depend upon genetic factors or heritage.

Size does NOT depend upon the individual breeding (however defined: IP address, email, avatar UUID, etc)

Size does NOT depend upon time (however defined, except as below).

When examining historical results, however, size does, in fact, depend upon time. Quite simply, when examining old data, certain sizes will be absent because they did not yet exist.

And, yes, every time anyone other than KittyCatS officials posts any contrary statements, no matter how firmly they hold to in their (false) belief, I will follow up with a post correcting the error. Quite simply, I believe that is is extremely important that these myths be put to an end and none of us should allow the last word a new user might see to be one promulgating such myths.

[jwenting Resident]

Tad, just because it's not breedable doesn't mean there's no pattern programmed into the cats.
The two aren't mutually exclusive.

But apart from that, yes, it's pretty much nonsense to think there's some conspiracy or hidden agenda on the part of KC that makes certain sizes appear at certain exact intervals.

[Ivy Norsk]

Talk about repeating annual threads. Last year, probably nearly to the day, I posted very much the same thing as jwenting.
About algorithms and what data could possibly be used by KC HQ for size. At the end of that thread, it had pretty much been established that KC uses a simple randomizer and does not track extra factors like avatar, box history etc., which would be vastly more complicated for them.

The issue persists though, and I have often wondered if the gap between "NO, there is nothing but random!" and "YES, I'm seeing special size boxes on a more than random basis" could be closed by something like number theory. I am not a mathematician.

But reading up (on wikipedia - yes, I know, not the most learned or technical source) does suggest that simple randomizers can and do fall into local patterns. Discussions on the subject often distinguish between global randomization and local randomization, noting that randomization theory often talks about "global" randomization (and this is what Tad has generally presented), that is to say what perfect true random is, and "local" randomization, which deals with what happens with smaller sample sizes, such as KittyCats.

A few possibly relevant quotes:

wikipedia Random Number Generation Wrote:The second method uses computational algorithms that can produce long sequences of apparently random results, which are in fact completely determined by a shorter initial value, known as a seed or key. The latter type are often called pseudorandom number generators. These types of generators do not typically rely on sources of naturally occurring entropy, though they may be periodically seeded by natural sources, they are non-blocking i.e. not rate-limited by an external event.

A "random number generator" based solely on deterministic computation cannot be regarded as a "true" random number generator in the purest sense of the word, since their output is inherently predictable if all seed values are known. In practice however they are sufficient for most tasks.

Wikipedia Pseudorandom number generator Wrote:A pseudorandom number generator (PRNG), also known as a deterministic random bit generator (DRBG),[1] is an algorithm for generating a sequence of numbers whose properties approximate the properties of sequences of random numbers. The PRNG-generated sequence is not truly random, because it is completely determined by a relatively small set of initial values, called the PRNG's seed (which may include truly random values). Although sequences that are closer to truly random can be generated using hardware random number generators, pseudorandom number generators are important in practice for their speed in number generation and their reproducibility.

PRNGs are central in applications such as simulations (e.g. for the Monte Carlo method), electronic games (e.g. for procedural generation), and cryptography. Cryptographic applications require the output not to be predictable from earlier outputs, and more elaborate algorithms, which do not inherit the linearity of simpler PRNGs, are needed.

Good statistical properties are a central requirement for the output of a PRNG. In general, careful mathematical analysis is required to have any confidence that a PRNG generates numbers that are sufficiently close to random to suit the intended use. John von Neumann cautioned about the misinterpretation of a PRNG as a truly random generator, and joked that "Anyone who considers arithmetical methods of producing random digits is, of course, in a state of sin."

Both of those articles are worth reading. The subject is highly technical and over my mathematical ability. But I do think that the implication Could Be that whatever generator KittyCats HQ uses, could possibly fall into patterns.

[fabioazevedo Oh]

Ivy

Thank you for sharing articles on wikipedia. I love math, and I am a senior system developer.

I'd like to read this your post about this same subject. Come join the theories. I have not used the forum, when you discussed this. But you can opposing my thesis. I show in graphs, and pedigrees, my results.


All systems based on 'random numbers' or 'random choices' must have a rule to guide. The system needs to be told what he must choose randomly.

If the answer is a number between 0 and 1. Then, the system will choose 'random' between 0 and 1

If it is a number between 1 and 100, then the system will choose random numbers between 1 and 100. There is a rule before applying the 'random choice'.

If you have only 7 options, for example a 'random system' can not choose 'option 10' because it does not exist. The system needs to be instructed to 'randomize' between 1 and 7.


If you want to prove that the 'KC' system, is a 'fully random system', we are left only say that it is configured this way:


1.Normal / 2.Normal / 3.Normal / 4.Normal / 5.Normal / 6.Normal / 7.TEACUP / 8.Normal / 9.Normal / 10.Normal / 11.Normal / 12.Normal / 13. Normal / 14.TOY / 15.Normal / 16.Normal / 17.Normal / 18.Normal / 19.Normal / 20.Normal / 21.PETITE / 22.Normal / 23.Normal / 24.Normal / 25.Normal / 26.Normal / 27.Normal / 28.BIGGER BIG / 29.Normal / 30.Normal / 31.Normal / 32.Normal / 33.Normal / 34.Normal / 35.MEGAPUSS

that is, the system chooses 'randomly' a number between 1 and 35. But, these options 30 are 'normal sized'. This would be a way to circumvent the 'random' system, offering more 'normal size' to be chosen. Thus explains the large number of occurrences with 'normal size' (85%).


If the system were just 'basic random', choose between 6 different sizes, ie, a number between 1 and 6. Never we would have 85% of the 'normal size'


Fabio

[Tad Carlucci]

(10-31-2014 11:15 PM)jwenting Resident Wrote:  Tad, just because it's not breedable doesn't mean there's no pattern programmed into the cats.
The two aren't mutually exclusive.

Actually, yes they are.

Proof of a 'pattern' is sufficient (but not necessary) to disprove 'random'

There is provably NO pattern to size, which is necessary (but not sufficient) to prove randomness.

Yes, I can prove there is no pattern. The proof requires an understanding of number theory, and information theory, which I doubt many here have and even fewer would like to plough through.

Personally, while I take KittyCats assertion that it is random as sufficient, I have tested a small dataset to verify the claim appears true. The OP dataset is large enough for a quick test, btw, if we presume no selection bias (as was claimed).

Further, as a programmer, I understand the ease of producing a random dataset, and the difficulty of producing a systemic pattern which does not yield easily to analysis (that field of number theory is called 'cryptography').

---

Just because the odds of an event are not 'even' .. all outcomes are equally likely .. does not mean it is non-random.

The odds of a caesium atom decaying are less than 1 in 10,000,000,000, completely random, yet, given enough such atoms, we can use that very randomness to calibrate our clocks to amazing accuracy. Just because 'time' is regulated and systemic (linear, in fact) does not mean the decay of caesium which produces 'standard time' must also be non-random.

---

I know the PRNG KittyCats uses. It is very, very good. The odds of it producing a pattern, and becoming non-random are less than the odds of that caesium atom decaying. The solution, to avoid that problem, is to use a good entropy source to re-seed after every few million uses. I also know the entropy source, and it is very good, too.

If you want to talk 'could happen' .. well, the odds of the heat-death of the Universe occurring in your lifetime are about equal to the odds of KittyCaS's PNRG breaking down .. 'aint gonna happen' is a good-enough approximation of either event.

[fabioazevedo Oh]

Hey Tad,

The KC uses MySQL. Please bring your theories based on the system.

You know: 'random choice based on data addicts'?

Please choose randomly among the options below:

1.Normal
2.Normal
3.Normal
4.Normal
5.Normal
6.Normal

I gave to you, 6 options. choose only one.


No system errors with the random choice. There is only the addiction to options to normal size.

[Ivy Norsk]

@Fabio. I think you've misunderstood my post. I never suggested that there were equal odds to get each of the sizes ( and I've never heard of anybody in KittyCats who thought that). I understand how the distribution of sizes can be created in an easy way, and customized by KittyCats, using a 35 point distribution or a 100 point distribution or whatever.
I am talking/asking about the "calculator" or algorithm that is used to Do the randomizing and stick each kitty box into a Size "slot".

Tad has understood what I was trying to say. And he says that the PRNG is very very good. And that the entropy source is also very good, which is good to hear, because I didn't know if anybody knew if KittyCats was using a PRNG that utilized an entropy source.

[fabioazevedo Oh]

hello Ivy.

I went for two or three time the system google translator. one of my trips to google translator, actually its text was understood differently. and I had two interpretations of his text. one was negative and the other positive. if my answer is different from what you wrote, I apologize.

[Kayleigh McMillan]

I am positive it is random I've heard multiple of the owners and CSRs stating that.
I think there is a sort of sequence ofcourse when the server hands out an odd size just as with specials vendors; the rare can come to you after one try.
It must be so though, due to the sequences, some are extremely lucky while orhers aren't: random at it's best.