animal-data

[animal-data]

lilies

Lilies are fairly straightforward- redxyellow=orange, redxwhite=pink, redxred=black and pink. No hidden colors, no real surprises aside from that rxr pink which I don't think I've ever seen included in any guide.

31 red x 0c yellow = 14 orange, 15 yellow 31 red x 03 white = 11 pink (and I guess 13 white but I don't think I actually got that?) 31 red x 31 red = 33 pink, 30 black.

Other notable combos: 0c yellow x 03 white = 05 white, which is useful for basically nothing 15 yellow x 15 yellow again is theoretically capable of producing every possible value. 30 black x 0c yellow is guaranteed to produce 14 orange, if those extra rxy 15 yellows are bothering you for some reason. 30 black x 33 pink produce default 31 reds, if you need that for some reason 33 x 33 pink will only produce more 33 pink; 11 x 11 pink will produce a wide range of values (including reds, whites, pink and black) and 33 x 11 pink will produce a smaller range of values.

Full color map:

I don’t see any clear rules associated with these ones...

[animal-data]

How flower hybridization works: some sort of guide (wip)

Short story: It's magic, Joel

Long story: In Animal Crossing New Leaf, each individual item has two types of values associated with it: an item ID that determines what it is and a set of flags that store various other values. For furniture these flags store things like customization state, and for all items the uppermost couple bits store things like whether it's buried or watered (outside) and how it's rotated (inside). For flowers, flags store... uh, something.

Pop open any ACNL savefile in a save editor and you'll notice every single flower in your town has a value attached to it. There's 0cs on those yellow lilies, 31s on those red pansies, whatever. These values are used to determine what offspring a flower can produce. Seemingly what the game does is take these values, run them through whatever algorithm it has, and look up the color associated with the result for that species.

To start off, each flower that's spawned "normally" has a flag value specific to its species and color. By "normally" I mean most of the ways you get flowers- spawned spontaneously in an open spot, planted by your villagers, bought from Leif, kidnapped from an island tour... basically anything that isn't breeding. This means that with fresh flowers you can guarantee the outcomes you get, and work from there for rare hybrids like purple pansies. Or you can make jape flowers, like a pair of yellow lilies that can produce every other possible value, and thus color, of lily. There's a value for white tulips that can produce reds, for example, but those whites and reds can't produce pinks. However, while you can have some idea of what results you get by controlling what parents you use, there's still an element of chance.

The nitty-gritty of it... One byte of flags is used for each flower. For breeding purposes, these seem to be interpreted as four pairs of bits- there doesn't seem to be a name for that, so I'll call them "bibbles".

These are the possible outcome for each pair of bibble values- 11 with 11 can only produce 11; 00 with 00 can only produce 00. 11 and 00 can only produce 01. 11 and 01 can produce either 11 or 01; 00 and 01 can produce either 00 or 01. 01 and 01 can produce 00, 01 and 11.

10 doesn't seem to be an attainable bibble value under any circumstances.

An example, using lilies: A red lily's default spawn value is 31, while a yellow lily's is 0c. Here's the bibbles: 00 11 00 01 red 00 00 11 00 yellow EQUALS 00 01 01 00 orange (14), or 00 01 01 01 yellow (15).

So here you can see 00x00, 11x00 and 00x01 at work. The former two only have one outcome while the latter has some variability, producing two different results. Incidentally, as you can see those 15 yellows have all bibbles set to 01 (except the first one, which is always 00 in most species). As mentioned before, pairing 01 with 01 can have all three bibble values as results... so pair two of these up together and watch them spit out all sorts of crazy results. (Except it seems like they'll mostly produce more yellows anyway. It seems like 01 is a more likely outcome for 01x01 than the other two, so outcomes will sort of tend towards the center of the distribution. For whatever reason yellows and oranges tend to be towards the middle of the values range...)

[animal-data]

So while I was working on the tulip color map, I had a good chance to peek into the RNG system as applies to flower flags. I took two of the 15-yellow tulips I got from trying to breed oranges and put them together. Now these tulips should have been able to produce all possible values for tulips, since all their bit pairs are 01- and 01x01 can produce all three possible pairs. But something rather interesting happened.

Here's a list of the bit pairs of flowers I got from this arrangement:

00 00 01 01 00 01 00 01 00 01 01 00 00 01 01 01 00 01 11 00 00 01 11 01 00 11 00 00 00 11 01 01 00 11 11 01 00 11 01 11

If you'll look at the second and last columns, you'll notice a little something. Now I didn't bother to record frequency of each individual value result (hah, what am I, a fucking nerd), but I can definitely say that "raising" in the last column and "dropping" in the second were both much less likely than the other way around. This clearly isn't what happens with all flowers- white roses (one bit set in the third column) and white pansies (one in last column) are much more likely to raise than drop.

So here's my best guess- likelihood of dropping versus raising is determined by relative position. The first non-zero bit pair from the left is more likely to raise, pairs further on are equally likely to do either and then more likely to drop. This greatly lowers the chance of white roses or pansies dropping and producing completely useless 00 flowers (though we have seen that happen with 10 red pansies to 00 whites?). ... No, wait, if this is true, black roses should be more common than pink roses from red x red, since pink requires raising the last bit pair. Uh, maybe it's the leftmost 01 pair?

I do want to note at this point that all my hybridization has been done with the help of jacob's ladders. I, uh, hope that doesn't change the results too much.

Some other notes:

3d x 3d: have yet to produce 3f, mostly more 3d, one 3c/more now but still fewer. 3dx3f produces 3f more easily however.?

0c x 1f 00 00 11 00 x 00 01 11 11 ... this got an 07. ? 00 00 01 11

1d x 1d: 3f, 1c, 1d, 3c, 3d 00 01 11 01= 00 01 11 00 00 01 11 01 00 11 11 00 00 11 11 01 00 11 11 11

Produced a 3f faster than 3dx3d, though still only one(?). No 0d/0c so far, no 1f... same outcome as 15x15 then.

[animal-data]

Full map of all flag value-color associations for tulips. Ugly teal colors are invalid/unobtainable values due to containing a 10 as a bit pair, grey squares are unobtainable due to no tulip naturally having a bit set in the leftmost pair; white flowers are represented with light pink and everything else is exactly what it looks like.

Version with those invalid values cut out:

Just looking at this it’s easy to see there’s certain rules in place:

- Black tulips are 00 11 00 00 and 00 11 01 00- ie 00 11 0x 00

- Red tulips are mostly 00 11 xx x1... plus 00 01 00 00 for some reason

- Orange are 00 01 x1 00, purple are 00 11 11 xx

- Pink is 00 01 00 01... just that one value.

[animal-data]

pnsies

regular 01 white x regular 30 red gives 10 and 11 red.

0000 0001 0011 0000 0001 0000 0001 0001

11 red x regular 01 white gives regular 03 blue...

11x11 red pansies give 10 red and 13 blue (0001 0011), very similar to 01 white pansies producing 03 blues. - Can also produce 31 reds, 01 whites(?)

0001 0001 0001 0000 0001 0011 0011 0001

10x10 red pansies give more 10 reds, normal 30 reds(?) and 00 whites. 00x00 whites seem to produce nothing but more of themselves. So I guess these are all useless.

31x31 reds produce- OH

Well, we did it Reddit. We found the elusive special reds. Of course this path to getting purple pansies is almost entirely useless since you can't tell any of the mutant reds apart from each other, but at least we've proved the concept. These purples have 33 flags, as opposed to the 3f of the seedbag purples, but as we've seen with pink roses flowers having a different flag setup depending on their parents is entirely normal. 31s also produce more 31s and ordinary 30 reds.

33 purples seem quite happy to produce more of themselves.

Incidentally, those 13 blues produce 03 blues and- OH

More 33 purples! This is maybe slightly more useful than 31 reds since you can actually tell you have a result after a point, but still not really. Depends on what 11x10 reds produces.

So looking at this, you can see multiple cases where same-breeding causes a 1 to become a 3 (ie, the next highest bit is ticked), as well as cases where the ones bit vanishes. Same as r/r pink and black roses. Whitexwhite = purple roses have a similar pattern but with 04 to 0c, so it might be more accurate to look at flower flags as four pairs of bits rather than two sets of four. There's no outcome there where the fours-place bit is set to zero instead though, even though that can happen with 10 red pansies to 00 whites. This may shed some light on why 00 white pansies don't seem to produce any form of blues, and why those 00 golds produced nothing at all (since nothing can hybridize gold roses obviously). You need at least one flag to determine where more flags go. On the other hand, you'll never see a 3 or c lose a flag unless it's bred with both flags zeroed on the other flower, and then you'll only see the higher bit cleared- ie, 1 and 4 but never 2 or 8.

30 red x 03 blue produces 11 red, one flag from each.

13 blue x 33 purple produces more of either.

[animal-data]

In the game's code there exist seed bags for all(?) varieties of flower, including hybrids, which obviously aren't normally sold or found. However, they apparently function normally, soooo I hacked in a couple to see what flags they end up with. The results: blue roses: fc (same as the ones in my town that I didn't breed myself) purple pansies: 3f Aside from being the only flowers I've found so far to have an F in their flag values, these are the reverse of each other (1111 1100 vs 0011 1111), for whatever the fuck that tells you. And the gold roses, oddly, have 00 flags. I'm not really surprised, given how different the spawning method is for them compared to normal flowers.

Now the blue roses and purple pansies both bred more of themselves quite easily (which matches up with the usual reports that once you've finally hybrid-bred these things, they'll multiply like rabbits), but the gold roses would not produce *anything* over the course of like a month. No yellow roses, no buggered white roses, just straight up nothing. Not even a jacob's ladder would coax anything out of them.

[animal-data]

The upper two bits of the second flag in furniture is used to store how it’s rotated when placed in your house... which has some fairly amusing outcomes in Marc’s editor.

At any rate:

00 is not rotated, 40 (01xx) is right-facing, 80 (10xx) is up-facing, c0 (11xx) is left-facing.

This leaves the remaining 14 flags to store customizations.

[animal-data]

For funsies, let's regroup these by similar values.

white pansies: 01 (0000 0001) white cosmos: 01 white tulips: 01 white violets- 01 white lilies- 03 (0000 0011) white roses: 04 (0000 0100)

yellow cosmos: 0d (0000 1101) yellow pansies: 0c (0000 1100) yellow tulips: 0c yellow lilies- 0c purple violets- 0c

yellow violets- 10 (0001 0000)

yellow roses: 30 (0011 0000) red pansies: 30 red cosmos: 30 red tulips: 31 (0011 0001) red lilies- 31

red roses: c1 (1100 0001)

So what we find is: - No species has two (standard) colors with similar values. - Most flowers have only one or two flags set, with a couple of the few exceptions having only one or two per "digit" (nybble? whatever it's called). The only complete exception to this rule is yellow cosmos, which still only have one digit with flags set. - White flowers are the most homogenous group, with all having low values; lilies are the only ones with more than one flag set - Purple violets have a value more similar to most yellow flowers, while yellow violets have a unique value. You'd think it'd be the other way 'round. -- This might have something to do with the fact that yellow violets have no hybridization potential. - Yellow roses have a value more similar to most other red flowers, and red roses have an entirely unique value. Since the value for white roses is slightly different than most other white flowers, this makes roses the most "unique" family; this is possibly related to roses having the largest number of hybrids possible (I think?)

I'd love to do an overall look at the family trees of each flower species, but given that most of the guides are... confused at best, it'd be hard to get into details.

- In most species, redxred produces black, and also pink in roses, which can also be achieved with redxwhite -- Exception: carnations, which obviously don't have a black version; instead redxred apparently produces pink? - Yellow x red produces orange in all species that have both parent colors. This is probably the most consistent rule. - White x red is the next most consistent, producing pink in all species that have all three colors (pansies having the parent colors but not pink). Pink roses can also come from red x red. - In two species, whitexwhite produces a hybrid- blue in pansies and purple in roses. Ironically the reverse (blue roses and purple pansies) are the two most elusive hybrid colors. - Wilted black x golden watering can produces gold in roses; gold x gold roses produce yellow. - In violets, white x purple (which has yellow-like flags) produces a blue hybrid; in no other species does white x yellow produce anything. -- Violets (and carnations?) are the only species where no two flowers of the same color can produce a second color.

So there's definitely some consistent rules underlying the flower hybrid system, and I suspect the flags mechanic is meant to facilitate this- take the flags of the two flowers doing the breeding, do some arcane math on them, and then compare the result to a lookup table for the flower's species. That way the function could be simplified, as the species doesn't have to be considered while calculating the outcome.

[animal-data]

roses

- c1 red roses, 04 white roses, 01 white pansies and 30 red pansies all produce more of themselves with the same values... - c1 red roses produce also produce c0 black roses and c3 pink roses

Considering those the "default" values for their respective hybrids, let's compare the flags-

1100 0001 (red) 1100 0011 (r/r pink) 1100 0000 (black)

Basically, only one changed- one on for pink and one off for black. Meanwhile, pink roses from red + white parents have different flags (41). So:

1100 0001 (red) 0000 0100 (white) 0100 0001 (r/w pink)

Like some of the other two-parent hybrids it has two flags both from its parents, but in this case both flags are from the same parent (red). Or you could think of it as red with one flag off, in a different pattern from r/r pink or black.

White + white roses = purple roses also just adds a flag:

04 (0000 0100) 0c (0000 1100)

[animal-data]

So something odd happened while I was breeding for red + yellow = orange roses... The game spawned a yellow rose with unusual flags.

Red + yellow roses = yellow roses with 51 flag... ??? I thought one of my villagers must've planted it, but apparently not- the other times I've caught villagers planting, their flowers had normal flags. And this was set up with only one yellow rose, so it's not a yellowxyellow.

c1 1100 0001 30 0011 0000 51 0101 0001

I suppose it follows the pattern of only having flags set that its parents did. The actual orange roses are 50 (0101 0000).

[animal-data]

pansies

01 (0000 0001) white x self = 03 (0000 0011) blue

30 (0011 0000) red x 0c (0000 1100) yellow = 14 (0001 0100) orange

Breeding red (30) x white (01) pansies just to see what happened netted me red 11 (0001 0001), and then something (not whitexwhite, since there was only one white involved) got me 03 blue pansies. What?

[animal-data]

- flower seed bags have no flags, in pockets or on ground. Planting said bags produces the usual flags. -- Note there's unused bags for the hybrid flowers that apparently work normally... will check those out when I start save editing. - Can't find the data on what the player's wearing to check hair/face flower flags.

So across two towns, I've confirmed flowers have the same default flag values across all the following: - new village, randomly spawning in the open in a village - islands and islands tours (inc. butterfly tour for lillies/violets) - flower seeds from Leif's/lost and found - being planted by a villager

And here's the list:

white violets- 01 (0000 0001) yellow violets- 10 (0001 0000) purple violets- 0c (0000 1100)

red lilies- 31 yellow lilies- 0c white lilies- 03

yellow pansies: 0c (0000 1100) yellow cosmos: 0d (0000 1101) yellow tulips: 0c (0000 1100) yellow roses: 30 (0011 0000)

white pansies: 01 (0000 0001) white cosmos: 01 white tulips: 01 white roses: 04 (0000 0100)

red pansies: 30 (0011 0000) red cosmos: 30 (0011 0000) red tulips: 31 (0011 0001) red roses: c1 (1100 0001)

Incidentally, the blue violets in my town are (all?) 05 (0000 0101). Looking around, I can also find some purple violets with flags of 14 (0001 0100) and 04 (0000 0100). These are both only one flag off from the "default" value; both have the fours place flag set in the second digit, and the hybrid blue violets have the fours place flag from the purples and the ones place flag from the whites.

[animal-data]

The second flag for map items

Looking at flowers, it seems the only use the second flag has is to be set when the flower has been watered (or rather, when the tile it sits on has been watered). Notably, there's no difference between the three different watering cans or even when it's been watered by a villager- all of these just result in the second flag being set to 40 (0100 0000). I've no idea how gold roses work with this.

At a glance the only other object I can find using the second flag is rocks- the breakable rock has 04 set, while some of the permanent rocks have 08 set with 00-03 set in the first flag- presumably coin rocks + which player character gets it.

- Also, fruit trees that've been shook recently, likely counting days until the fruit spawns again.

To research: - Watering the tile a rock is on (probably just results in 48 or what have you in the second flag) - Hack in some wilted black roses to water and see what marks them to turn gold

[animal-data]

Cypress plant customisation (dark)