Modifying subcode to affect stats and monster type in Monster Rancher 2

Monster Rancher Metropolis: The Laboratory: Technical Research On The Monster Rancher CD Read Process: Ongoing Research : Modifying subcode to affect stats and monster type in Monster Rancher 2
By Enigma on Sunday, July 8, 2001 - 08:41 pm:

Greetings all, I've done more research along the lines that I started in the "The quest for..." thread.
This time I've used Monster Rancher 2, since it's newer and more people are interested in it, and I've done a larger range of samples.
I'm still using the subcode from a Nirvana Nevermind CD (BMG edition, does not produce the Swimmer normally) taken using CloneCD v.2
As was shown under my Monster Rancher 1 investigations, it was possible to pull the rare from this disc, and I was able to do so in MR2 as well with the same methods.

As with last time, I have modified the PMin, PSec and PFrame values from [Entry 2] to effect the desired changes, and PLBA recalculated to match accordingly with the known formula.

Below are my results, after the table I will give a brief synopsis of my findings.
(The bottom entry is the original subcode from the CD unmodified, the second Swimmer is from the values modified to match the Swimmer CCD file posted on this site)

Pmin 2Psec 2Pframe 2PLBA 2monstermainsubLifPowIntSkiSpdDeftotalNatureLikesDislikes
41590188775Spot SaurianZuumHare1331668115715696789GoodPlayPotato
415974188849Spot SaurianZuumHare1331668115715696789GoodMeatMilk
4200188850Rock HoundTigerGolem80122159128162119770GoodMilkPotato
42230190575Rock HoundTigerGolem71125163125151115750GoodPlayCup Jelly
422370190645Rock HoundTigerGolem71125163125151115750GoodPotatoCup Jelly
423774191699Rock HoundTigerGolem88127154135150112766GoodMeatPotato
42380191700SwimmerMew???14690105143129126739NeutralMilkMeat
423870191770SwimmerMew???14690105143129126739NeutralPotatoCup Jelly
423874191774SwimmerMew???14690105143129126739NeutralBattleMeat
42390191775Rock HoundTigerGolem76128155107148127741GoodTabletCup Jelly
423974191849Rock HoundTigerGolem76128155107148127741GoodPotatoCup Jelly
42400191850Rock HoundTigerGolem79126162125161108761GoodMeatPotato
424074191924Rock HoundTigerGolem79126162125161108761GoodShoesMilk
42410191925Rock HoundTigerGolem85123140115157130750GoodCup JellyPotato
424174191999Rock HoundTigerGolem85123140115157130750GoodRestCup Jelly
42420192000Rock HoundTigerGolem85111157121163133770GoodShoesBattle
42430192075Rock HoundTigerGolem96106156120145128751GoodMeatBattle
42440192150Rock HoundTigerGolem70132164126163111766GoodFishMilk
42450192225Rock HoundTigerGolem87133146129166110771GoodMilkTablet
42590193275Rock HoundTigerGolem93126141117156116749GoodPotatoBattle
4300193350TyrantGolemDragon10024214910892149840BadRestTablet
43380196200TyrantGolemDragon1152291379497158830BadMeatBattle
433870196270TyrantGolemDragon1152291379497158830BadRestCup Jelly
59230267075Draco HopperHopperDragon6112516317514185750BadPlayPotato


As you can see, when I modified the PMin, the result was a new Main and Sub type monster (possibly just new main, with corresponding subtype tables for that main type).
Modifications of the PSec resulted in a new set of stats for the monster (provided PMin remained constant).
PFrame did not seem to have any affect when modified. I did not try manipulations of PLBA that do not fit with the calculated values from PMin, PSec and PFrame.

Also, the rare monster that can be found on this disc only showed up during a 1 second range.

I know it has been mentioned by Quincunx that likes and dislikes are generated randomly at the shrine, I just felt like including the ones I got with my monsters here. (some of the Rock Hounds seem to like to chew on shoes.)

Hopefully this helps the rest of you who are working along this line.
Some areas of interest that I did not get to with this research, and would like to see investigated or investigate myself are modifications to the TocEntries under [Disc] and PMin under [Entry 1].

Good Luck and Happy Ranching "


By Enigma on Saturday, July 21, 2001 - 07:24 pm:

Greetings all, I've been back at it again.
This time the disc in question is my copy of Beck: Mellow Gold (BMG edition) which produces the rare Magnet (Henger/???) in MR1, but a Mum Mew in MR2.

This time I only modified the PSec value from Entry 2. I was unable to manufacture the mythical Moo, but the data gathered can still be used for comparing starting stats with altered .CCD files.

TocEntriesPmin 1Pmin 2Psec 2Pframe 2PLBA 2monstermainsubLifPowIntSkiSpdDeftotalNatureLikesDislikes
1512473060213660Mum MewMewPixie10471115156152129727BadWool BallBattle
1512473160213735Mum MewMewPixie1027898156170103707BadFishMeat
1512473260213810Mum MewMewPixie11383112140160116724BadCup JellyMilk
1512473360213885Mum MewMewPixie1097411216415699714BadPotatoTablet
1512473460213960Mum MewMewPixie11282113159149117732BadRestMeat
1512473560214035Mum MewMewPixie1138799142168119728BadCandyMilk
1512473660214110Mum MewMewPixie1167797162159118729BadMeatCup Jelly
1512473760214185Mum MewMewPixie11789100163158104731BadFish BoneBattle
1512473860214260Mum MewMewPixie1247193142175120725BadFish BoneTablet
1512473960214335Mum MewMewPixie10590101135156119706BadFishPotato
1512474060214410Mum MewMewPixie10888108153169100726BadRestMeat
1512474160213485Mum MewMewPixie1148586143165122715BadPlaymatePotato
1512474160214485Mum MewMewPixie1148586143165122715BadRestPotato
1512474260214560Mum MewMewPixie11473103149171125735BadCandyMeat
1512474360214635Mum MewMewPixie12568102148153120716BadPlaymateMeat
1512474460214710Mum MewMewPixie9994110154171103731BadMeatPotato
1512474560214785Mum MewMewPixie1169592157174102736BadTabletMeat
1512474660214860Mum MewMewPixie1097388147173127717BadFishBattle
1512474760214935Mum MewMewPixie12669105158147108713BadBattleMeat
1512474860215010Mum MewMewPixie1187787146158129715BadPlayPotato
1512474960215085Mum MewMewPixie9996116138163102714BadRestPotato
1512475060215160Mum MewMewPixie10772104147173111714BadPlayMilk


The purple entry above corresponds to Lisa's copy of Mellow Gold that does not produce a rare in either game.
Blue is for my copy of Mellow Gold.
And the red is for where I messed up and used an incorrect value for PLBA, but that had no affect on the monster created (as can be seen by the entry immediately below it with identical stats).

Unfortunately, it does not seem likely that the Moo monster can be gotten from this CD, as I doubt the different pressings would vary in length by up to 10 or more seconds in any direction (mine and Lisa's varied by 2 seconds), without additional or fewer tracks. And if that's the case, the odds of finding the rare by this method are astronomically small.

Perhaps some day I'll finish out the full minute for a complete set of Mum Mew's off this disc.

G¥d luck and happy ranching "


By Quincunx on Sunday, July 22, 2001 - 03:34 am:

Enigma: You said "I know it has been mentioned by Quincunx that likes and dislikes are generated randomly at the shrine, I just felt like including the ones I got with my monsters here. (some of the Rock Hounds seem to like to chew on shoes.)"

Somehow I missed this bit earlier. Is it possible that likes and dislikes are generated from a list of possibilities, one of which is [breed-specific item] and each main breed (in the case of shoes, Tiger) has its own breed-specific item? For instance, just as possibilities, Tigers could have shoes as a possible like, while Pixies might have perfume as a possible like.

Just a thought.


By Quincunx on Sunday, July 22, 2001 - 05:25 am:

Methinks I previewed a post and forgot to send it. Anyway, the point of it was that another thread says that the MR1 Shrine is specific to the frame when checking rares (a disc producing Magnet will produce a Suezo with one frame of difference in either direction), while from your data, the MR2 Shrine appears to be specific only to the second (as evidenced by the three Swimmers). So the two Shrine programs are different in that way, apparently.

And I was looking at those statistics for the Mum Mews trying to find some sort of pattern. I looked at differences between sequential stats, differences between those differences, and differences between those differences. I looked at the remainders in mod 2, mod 3, mod 4, mod 5 and mod 7. I made parallel one-dimensional scattergrams...

...and I can't see a pattern. Either there's a pseudorandom number table involved at some point (in which each original number gets translated to a different number), or there's some tricky math (can a PSX do trig functions on the fly?), or I'm just missing something which would be painfully obvious if someone told me the answer.

In line with my first research post, I notice that the highest value of any stat for the 21 different Mum Mews is never more than 30 higher than the lowest value for that stat. And what still baffles me and defies logic is that this holds true even for the composites; they range from 706 to 736.

Something to sleep on, I guess.


By Enigma on Sunday, July 22, 2001 - 10:08 am:

Actually, Quincunx, the MR1 shrine is also only exact to the second, not the frame when choosing rares.
This was demonstrated in my "The quest for...." thread, where I modified the Nirvana Nevermind subcode to produce the Karaoke monster in MR1.

And for the stat range, perhaps there is a set starting value for all stats, and some algorithm for adding pseudorandom amounts to each stats up to a given pseudorandom amount (that never varies by more than a set amount, apparently 30 for Rock Hounds, Mum Mews, Melon Suezos and KnightMocchi's, about 45 for Frozen Gaboos, according to yours and my stat charts).
But as for determining what those stats are, that's for someone with a lot more free time on their hands, I guess.

Monster likes and dislikes do seem to be from a predetermined list, at least for each main, with a lot of generic items that can be on all lists, and some specific to the race. My Rock Hounds never asked for a Playmate or Fish Bone, and the Mum Mew's never wanted some Shoes to chew on.

Perhaps the next big project would be to create a completely controlled .CCD file and see what all monsters can be created from it.

G¥d luck and happy ranching "


By Quincunx on Sunday, July 22, 2001 - 02:12 pm:

OK. I was just referencing the post by John Hawley on Friday, July 6, 2001 - 07:11 pm from the thread The Laboratory: Ongoing Research : How the MR2 Shrine Program Generates Stats for Monsters from CDs: Initial Research. As I do not own MR1, I can neither confirm nor disprove anything.

As for stat numbers, if there's a pseudorandom number table of, say, 50 entries (i.e. 0 becomes 17, 1 becomes 38, 2 becomes 6, etc.) involved, this is going to make cracking the stats ridiculously hard.

Even worse if it's not a one-to-one function (I forget the mathematical term) where, say, inputs of 4 and 37 both yield an output of 16, because then working backwards, that 16 that we see could be either a 4 or a 37 originally.