Messages

1

Final Fantasy XII / Re: RNG and ZS Chest Appearance

« on: June 26, 2010, 10:24:37 PM »

Hey, FrozenJosh!

Have you built an RNG spreadsheet?

No, sadly. It's been a long time since I've used Excel, and although it would save time if I knew the appropriate magic ranges for that level, I don't know exactly how the game handles rounding, so I was going to do it manually to cut out the uncertainty. If someone does know how to make sure the numbers are exactly right, it wouldn't be necessary, but I still like the consecutive Cure value idea better because it's so accessible. If that does work as well as I think it may, I'd be interested in helping make a new guide for the Cure Twin method or whatever you're going to dub it, haha.

Another thought: if you do play through various levels, when you're at one of your target levels, save your game.  Next time you save, start a new file

Yeah, the problem was more that I didn't intend on having to have these target levels until I wanted to expand the method to be usable by more people. I didn't have any files with any of the target levels. Instead of having multiple files though, I just decided to keep Vaan at level 1, level someone else to level 25 and stop, level someone to 50, etc. Have all my target levels in one file. But I didn't really feel like doing it, especially now that I don't play FF12 for myself much anymore and I have so much less free time.

Even so, just verifying that 640/641 is a reliable double at every combination is daunting.

I hadn't considered testing every combination, honestly. I was thinking that the variance in Cure value for the same two consecutive multipliers would be greater at higher levels and magic values and lower at lower levels and magic stats, so if it works for level 99 and 99 magic, it would be the same consecutive Cure value for all combinations, and at the very least, if you manually tested the lowest values and highest values and they matched in both cases, that should cover the entire range of variance due to stats. Now that I think about it, I don't know if that's a valid theory though.

By the way, I don't know much about the "line XXX" notation because I don't use the spreadsheet much. Are you talking about the two values that produce 1596 twice in a row for the level 99 and 71 magic chart? That's the one I thought you were referring to and the one I was going to test.

And looking for nearby occasional doubles at every combination completely blows my mind.

It shouldn't be necessary to create a program for it. There won't be tons of consecutive doubles at all, and I wouldn't look for them by manually checking spreadsheets. Instead, I'd focus on how to confirm the doubles you got are the ones you want. In the same line of thought as before, if the variance is least when the level and magic is the lowest, you should catch more doubles at those minimums, right? I'm not sure, but that's what I'm thinking. But there really shouldn't be many of them, and like I said, there are easy ways of telling if your consecutive doubles are the ones we want. I'd test using lowest stats and get to your marker. Then, I'd try to find some other marker that isn't based on stats that can tell you if you've found the right one. It's a little more complex, but it will at the very least tell a player if they've found the right doubles and it only has to be done once. Using a high combo weapon found early in the game should work pretty well, and the higher the combo rate, the lower the chance of having the same pattern follow after two different sets of twins, whether perfect twins or LIARS >:O!

BTW, knowing how RNG numbers spawn chests makes it really easy to locate numbers that will spawn specific chests. There seem to be plenty of them between 641 and the earliest combo in GM's method.

Yeah. I haven't been up on the news, but as far as I know, the chests in a zone are numbered and they each pull one value for their spawn. If you knew the ordering of the chests in the zone, you could theoretically tell just by looking at a given chunk of RNG and tell which chests in any zone are going to show up and which aren't. Crazy awesome in theory, but I only know of one zone where someone has gone through the effort to confirm the game's chest ordering. But like I said, it's been a while, so if more's been discovered and I'm in the dark, by all means...

No doubt that's why you'd have to play through to various Levels. How does FF12 determine Magic Power, anyway?

I thought it varied playthrough to playthrough, at least a little. Pretty sure there's a chance of the stat going up every level [EDIT: guess not, thought it was like HP, but apparently it's fixed], which means I'd do a normal playthrough and equip some equipment to increase magic to a point that will be able to be reached by those that have naturally higher magic than I happened to get (by equipping lower boosting or no equipment) and by those that have lower stats that I had (by using higher boosting equips). Arbitrary, just like 71, but the intent is just to hit a number that everyone can hit without the randomness of stat increase upon leveling being a problem.

My gut feeling is that this method will work. I feel guilty because I don't have much time, but I think this will work and I want to confirm it and then present it to anyone that still cares about FF12, heh.

2

Final Fantasy XII / Re: RNG and ZS Chest Appearance

« on: May 25, 2010, 08:47:52 AM »

@ Gramps: I've been really busy with real life stuff, but I've started to come back to this. My current problem is I don't really want to play FF12 right now, but to create Cure lists for various levels requires progressing the story to the point where I can power level at least one character up to 75. I was planning on adding charts at levels 1, 25, 50, and 75 to the level 99 one, and the Knot of Rust chart and then redoing the entire guide from the ground up with a section to determine what sub-method you should jump to. The problem, other than me not really wanting to do it, is in how cumbersome one chart is and then multiplying that by 6 and probably having to include more theory so people understand why things went wrong. Ugh, just the opposite of what I wanted. However, your marker might just work. I will try it pretty soon. My biggest concern is other consecutive values in the RNG that are so close to one another that they will be the same value at lower level and Magic. I think there may be some of those earlier in the RNG, which is also easy enough to check (just have a level 1 character go a-casting, and then create ways to check the RNG after consecutive doubles, like a high combo weapon that is easily obtainable early on, or anything else that isn't so heavily affected by stats).

Another concern is whether or not all models start before the consecutive cures in the RNG. I feel as if they will though, given the spot in the RNG.

You'd also need to re-locate solutions that are earlier than the double cure marker, which is easy enough. Just takes time. It's pretty darn accessible, obtained early, and although explaining how to cut corners might be a little difficult to grasp, it's nothing compared to wading through charts and having to learn the basics of the entire method to figure out what's going wrong. If this works, this might be the answer to making RNG chest manipulation as simple as FFGM's punching yourself in the face method.

EDIT: Bravo.

3

Final Fantasy XII / Re: RNG and ZS Chest Appearance

« on: February 04, 2010, 04:35:54 PM »

ZERO: Don't have the time to quote, but I was wondering...

You said you killed the same enemy at the same level at the same time and the loot was still different. Did you try <i>collecting</i> the loot at the exact same time?

4

Final Fantasy XII / Re: RNG and ZS Chest Appearance

« on: January 24, 2010, 04:30:55 AM »

Been gone a little while.

Wow, seems like a lot has been figured out. I had a suspicion that the game used one value for each chest spawn, but my problem with mapping things was always trying to figure out which one was chest #1, which was #2, etc.

So, some questions:

Does the RNG use a value for a spawn% if that trophy game is dead and cannot spawn? I would expect that it doesn't, but it could potentially take the value anyway and handle the lack of a spawn on the back end. I was told that there was evidence to suggest that %spawns were on a different RNG than other things, namely in that entering a particular zone moved the RNG the same amounts regardless of whether a particular rare game was dead or not, but I started to doubt all that when I started seeing how nonstatic the RNG could be even when entering the same zone (must have been the weather).

Is the nature of item drops known now? Is there anything more we know about monster drops? I was thinking about some possible ways to narrow in on figuring out more about item drops and determining whether or not it will be manipulable and if so, how.

It does still seem like there are a different set of RNG values for drops though. Is that the case? Because if you get the 5 hit combo and then kill something without moving the RNG, the drop doesn't use a new value and spoil your chest contents.

EDIT: Sorry about the 10 patterns thing and the Knots of Rust confusion. I realized that the 10 patterns is an arbitrary definition for the purpose of visualization (and for matching the variance in output of FF12GM's method), and the idea of Knots of Rust was to use something that was controllable, not relying on level and requiring no conversion. I haven't been playing FF12 or working on RNG stuff with it though because I don't know if I want to map out Knot of Rust values for those that want a simpler approach than using Excel and spreadsheets and converting things themselves, especially since using Knots of Rust like that would either:

1) require the player to just get enough of them to determine starting range and then substitute Cure for them

2) require buyable Knots, only late game

EDIT: Also, there is at least one room in Garamsythe that does move the RNG itself. It's Cloaca Spur 4. I doubt it's character dependant, as I tested that specifically with other RNG tests, and character selection has thus far seemed to be a nonfactor.

5

Final Fantasy XII / Re: RNG and ZS Chest Appearance

« on: December 23, 2009, 08:47:39 AM »

I don't understand what you mean. All rare steals have the same success rate which is 3%/6% (no thief's cuffs/thief's cuffs equipped).

Oh, do they? Oh... heh. Right, okay. Silly me.

For testing purposes I chose a zone where the RNG doesn't move (where enemies don't cast anything). If I'd be in a zone where enemies may cast spell (such as the Water-steps in Giruvegan) I'd need to kill the off as fast as I can, before the RNG passes the rare steal value.

And you just stay far enough away from the mark until you get the RNG where you want it and then run up and steal before they get an action that messes it up? Or even start charging it from outside theire aggro range so that you get it immediately upon moving into steal range?

6

Final Fantasy XII / Re: RNG and ZS Chest Appearance

« on: December 22, 2009, 05:05:41 PM »

ChaosBeelzemon i think it is better if you manage to find lines without the Thief's Cuffs

I agree (although you could collect values and then check them for success against other factors). Also if you aim for the steals that are the lowest steal percentage in their category (for example, the lowest rare steal). If you find a value that works for the lowest percentage, it may work for all steals regardless of percentages.

Also, how are you neutralizing the enemies doing things and moving the RNG along? I'm wondering about what it would take to remain in control of the RNG if you're going for rare steals from rare game, bosses, etc.

7

Final Fantasy XII / Re: Most Valued Type of Weapon?

« on: December 17, 2009, 12:25:34 AM »

The thing is you could dual wield almost everything in FFT;

I wouldn't say that. Spears, guns, bows, crossbows, poles, cloths, harps, books, axes, bags, and Fell Swords (in WotL) cannot be dual wielded. It's easy to forget that there are only 9 weapon types (including fists) that can be dual wielded compared to the above 11 because those 9 are much more common weapon types with more weapons of that type. Overall, in WotL, there are 77 weapons that can be dual wielded (counting fists, not counting the Onion Sword) compared to 65 weapons that you can't. A fair portion of those 65 are trash weapon types, but I still don't think it qualifies as "almost everything".

But you know that already, in general. I knew I recognized your username.

ninja blades weren't special in that regard.

It's a Ninja ability and Ninja Blades are the Ninja's iconic weapon in FFT (only Ninjas can use them besides Orlandu, not counting OK or Mime, and they are the most obvious weapon for a Ninja). Ninjas have the ability innate, and although you can apply the ability to other classes and weapons, you have to be a Ninja first and you'll probably be using Ninja Blades unless you either unlocked the class before they were buyable or if you know better and go unarmed.

As with many of the other things in FF12 that are light references to FFT in some way, they aren't exact matches. There aren't classes in this version, so it would make most sense to be a property of the weapon itself (although it could work as a license, but would have an even looser connection to FFT).

8

Final Fantasy XII / Re: RNG and ZS Chest Appearance

« on: December 15, 2009, 05:12:01 AM »

Billygoku the RNG can go from 1 ~ 1.125 (as formulas shows) this mean 125 different values for randomness
the KoR have only 10 steps

At first glance, that's what I thought as well from seeing the formula for Knot damage, but it is not necessarily true (actually, it probably isn't). Those formulas are derived, after the game has taken the RNG value and used it. They show the bounds of the randomness multiplier, but we don't know the divisions for sure. I believe I've seen more than 125 different Cure values, implying the smallest distance between two consecutive Cure multipliers (in numeric order) is less than .1, although I could be wrong.

In any event, the true upper limit of variability lies within the variability of the RNG values themselves. Those values are completely hidden to us - they are the values that the PS2 generates and I don't even know what they look like. The game takes those values from that list and runs them through functions to see what the "random" result is, but without actually seeing the values themselves, we can only derive the variability in the formula's randomness, not the variability in the actual RNG values. I assume at least that two different values in the RNG can produce the same randomness in one function and different randomness in another. It's all about how the game is currently using the actual RNG value.

if you convert the cure list with numbers like 1,2.....124,125 then you are going to have your global list

It would be nice, but it won't work for practical use because that would require people to do math to match their numbers with mine. You absolutely can make a Cure Shift chart that lists the Cure multipliers (I know for a fact that it would work, because you can compare to the Cure Shift chart even when you're not level 99 with 71 magic and find your location based on the value going up and down and how much compared to other numbers, which is basically comparing the multipliers), but people would have to recalculate that entire list using their own level and magic, and as we have been shown, we don't know how the game handles the rounding, so it might be off by +/- 1 or so.

(BTW, I have reached Golmore, and could make a “fleeing” dash to the right spot, but I don’t think I can get in to the village until the story brings me there.)

The story is about to bring you there, pretty much. Right after the story sends you to Jahara, you head off to Golmore. You don't get all the way through Golmore before needing to get through the barriers. Fran steps in and you visit Eruyt.

“BGT (Billygoten) Universal Starting Number,” (a number selected just past the starting range of all PS-2 models.)  Past and future discoveries of spawn points would all be plotted from the BGTUSN.  The player would never need the chart again (at least for spawn points) – just a list of hit/cure formulas for various chests.

Hm... yeah, a starting number that everyone learned how to get to on their model and the solutions would all be given from that same starting point. That would simplify things. You really wouldn't need the chart except to find your distance to the BGTUSN (lol, I love the acronym - I'm iffy on attaching my name to it, but considering the crazy acronyms on FF12 challenge runs, this is something I support 100%, haha).

Once you get the Diamond Armlet, Knots are in chests everywhere, even in Low Town!

I know, but you need a bunch of them and a continual supply. Well, you really don't, but some people will surely have trouble learning to use Cure instead of Knots, even with instructions and explanations of what you're doing, and Cure values change as you level up.

Imagine that the large Cure value table I posted was converted to Knot values.  Once you get your chest to spawn, move to the chest, cast three more Knots and find that three number sequence on the Knots chart. (The area you’d have to search should be relatively small.  Make an educated guess based on zones you crossed and whatever else happened.) Once you found your sequence, count down for hits and across for cures to get your rare item.

Yeah. Just shift into the no-brainer and let it go.

maybe there are no treasures available before Eruyt that would be worth the trouble anyway!

There are. In general, the most important are the Rod, the Celebrant's Miter, and the Dragon Mail. I RNG spawned/farmed all of them this time through. There are a bunch of other things that are also pretty good for that point in the game (in the Sandseas), but you can just run straight through the Sandseas, complete Raithwall's Tomb, and be on your way to the really good stuff, so it's not really worth the time to farm from the Sandseas and such.

Perhaps the Level and Magic Power progression would have little or no variation in the early part of the game.  For example, maybe in every game Balthier hits Level 4 Magic Power 7.  A single Cure chart could be developed for that level, and players could watch for Balthier to reach it, then go collect those early rare items.

Yeah, maybe.

As far as cheating, it's going to come down to an individual basis - where you draw the line and most importantly, what is and what isn't fun for you. This changes with multiple players, of course, but if it's just you, do what you want. This game was explicitly designed to get people to seek outside information on it. The designers wanted you to talk to people and seek outside help to unlock all its secrets. It also sounds like a lifetime (perhaps more) of tedium trying to find everything without using any outside information at all, and it sounds like a terrible experience too. I can understand others having qualms with an exploit that is so very abusable, but I don't miss any sleep because "farming all day for one chest isn't fun" is an excuse that works for me, and in that this is a game that is meant to be enjoyed, is also pretty relevant.

I think the published formula for Cure values is reliable.  It seems clear that actual RNG values fall in the range from 1.000 to 1.125, but  it’s probably a mistake to assume that each RNG number is a perfect 3-decimal-place figure, and that therefore the RNG has only 125 different numbers.

Exactly. There's a difference between the multipliers that we see (which is a workable derivation of the actual RNG value) and the actual RNG value itself.

What I can't grasp is how the RNG would generate exactly 184 different Cure values, unless there were exactly 184 different numbers in the RNG.

There's more than that, probably. There might be 184 different multipliers for Cure. It doesn't matter much, because it's just a number that happens to be what it is; the number 184 doesn't offer us much insight anyway. Also, you're assuming every single number in between those Cure values are accounted for. They likely are, but not necessarily.

If there are more than 184 different RNG numbers, is it possible that two or more different RNG numbers could calculate cure values (out to who-knows-how-many decimal places) which all round off to the same whole number?

I assure you, it's not just possible. Just consider the cases of output overlap with other things. There's the lower variability / greater overlap of Knots of Rust with lower HP. And that won't be the only example. The two values in the Cure Shift list that are 1651 and 1652 are almost always the same number with lower level and magic. There are undoubtedly values for level 99 and 71 magic that overlap but wouldn't with higher magic. And it is probably the case that there are numbers that will always appear the same even at max magic and level that come from different actual RNG values.
 

I suspect there are other places on the RNG that would also yield rare items from chests.  To find them we need only to persuade Grand Master to come back to FF12 and spend several thousand more hours locating them for us.

Oh come now! You are fully capable of doing so yourself, given your knowledge of the way the RNG works. We can even move along the RNG trying values two at a time until we get a rare item and then checking it for higher gil percentages. With a big enough map, this wouldn't be hard at all. I found a value that works for most normal items myself this time through (so I could farm the aforementioned Dragon Mails and Celebrant's Miters).

Since zero369 reports finding the Cure value 1629 at around line #1330, it seems that all of the numbers in that range, and no numbers outside that range, appear as a Cure values.

Wow! Neat!

9

Final Fantasy XII / Re: RNG and ZS Chest Appearance

« on: December 13, 2009, 12:15:56 PM »

This might be more useful in the IZJS version because you can get Knots of Rust right away. Of course you won't have Infuse though so early.

Still I do know you can use the self-hit method as soon as your characters can do themselves no damage. I've seen FF12GM use the hit method in IZJS to get the rare Meteorites from chests in Giza just after Raithwell.

You can do 0 to yourself even earlier, potentially.

But that method can't really be used for spawning the chests 100%. You have to determine your location in the RNG and get to the right spot all before passing FFGM's first marker. To use combos as markers would require determination of place / pattern convergence earlier than the first marker or it would require a different method that gave you more room to spawn the chest before getting to the 5 hit combo, or even another target location.

You can even use Cure shifting earlier, but it's more about understanding the theory behind it, as I don't have spawn solutions for chests that early in the game and you'd have to do mapping on your own, based on your own numbers (I did for the Barheim chests). Anyway, I know you can't use it super early, but it's independent of level and magic, at least.

At least people at the end of the game could use it to spawn the ZS chest without needing to be level 99. That was kinda the point in the first place anyway, although it's a shame because this kind of manipulation is so easily applied to any chest, potentially.

10

Final Fantasy XII / Re: RNG and ZS Chest Appearance

« on: December 12, 2009, 01:20:39 AM »

I finally got to the point in the story where I could test Knots of Rust with characters of various levels. For the record, that point in the story is right after walking into Eruyt for the first time (you don't have to do anything there, just enter the village) because that is the first point at which Infuse is available.

The method involves buying the Nihopalaoa from the Clan Provisioner once you've defeated 10 marks. You equip it and use a Vaccine on yourself to inflict yourself with Disease (Vaccines are in a repeatable bazaar package - you can essentially farm them from Buers in Zerti Caverns). Then, you use Infuse, take a couple steps (I went to 1 MP) and use Infuse again. This takes your HP down to 10 x your current MP and due to Disease, that becomes your max HP, lowering your Knot of Rust damage (and making it a standard that anyone can match).

The damage you get out of Knots of Rust does indeed follow the RNG in a completely predictable way, moving one spot at a time, and it is the same pattern flow like using Cure just changing depending on your starting range. The amount of damage in the pattern remains the same regardless of stats or level. The same flow was reproduced with two different files using Vaan at level 1, Penelo at level 14, Balthier at level 27, and Ashe at level 99 with all licenses. This means if Knot of Rust damage became the basis for a RNG chart, anyone at any level with any stats could use it to determine their spot in the RNG and manipulate the RNG as long as they had a Niho, Infuse, Vaccines, and Knots of Rust.

Problems:
This method would still use a chart, and due to all the possible starting ranges, it would still be rather large. You wouldn't need it except to determine your place and figure out the instructions to get a target chest to spawn, but you would probably need to refer to it if you decided to go after another chest and someone posted the "solution" online.

This requires Knots of Rust. In my newest file, I have only 10 of them. I didn't farm any, but I did go out of my way to make sure I got a couple during story events. In early parts of the game, it is a real pain to get them - right now, I can only farm them from Barheim for the most part. I will get at least one almost every trip, and often two, but it is a rough zone unless you have Dragon Mails and Celebrant's Miters or unless you abuse the Immobilize glitch (you still have to be pretty familiar with it, due to how the enemies respawn).

Late game, you can buy them. You need to be Clan Rank Knight of the Round, which means defeating 32 marks (!) and getting 500,000 clan points.

Also, it is possible that you won't need Knots of Rust except to initially determine your starting range. What you would essentially end up doing is using a bunch of Knots of Rust to compare that to the chart to get your starting range, and then you'd cast Cure on yourself and figure out how Cure overlaps with Knots of Rust (since they both behave the exact same way as far as the RNG is concerned, but Cure expends no resources).

Another problem is when this is available to use, although it is a minor problem. You can't use Knots to farm the good stuff in Barheim chests when you first get the chance and some of the requirements are much easier to meet if you already have that stuff. You can use Cure instead though (although the Miter is tricky because of a zone in the middle that constantly changes the RNG, requiring you to make it through the room at approximately the same amount of time each trip, but I have never failed to do so).

So, what do you think? I would like to hear your opinion as well, pprincess.

EDIT: Also, I forgot to mention that at 10 HP, you only see 5 different numbers. It is enough variability, I believe, to be able to detect your spot given enough numbers, but I did confirm that there is higher variability with higher HP. In other words, at 10 HP a lot of the 1 damage spots are actually using different multipliers but getting rounded to 1. You can see a greater number of different damage values at 100HP. If 2 or 3 HP eliminates some of that output overlap, it might be the best number to use, although honestly, the damage from 100 HP didn't get close to killing my characters that had leveled up (but it might be a problem for a Low Level Game party, as max damage for a Knot is equal to your max HP).

11

Final Fantasy XII / Re: Most Valued Type of Weapon?

« on: December 09, 2009, 11:03:13 PM »

Weren't ninja blades supposed to be able to be dual wielded in the planning stages of the game? Like in FFT with ninjas?

Lots of love for Poles here. Hm, I always hated Poles. I tried to like them, but I can't justify their use over other weapons. Their damage is beaten out by other weapons and I'd rather use other options against tough enemies with low magdef.

Sorry to be cliche, but I like Greatswords a lot. I also find Daggers (the Danjuro mostly) and Unarmed attacks to be sorely underrated (late game). The Masamune is also amazing.

I thought I'd like ninja swords, but I can't get past the dark element. It's just too much of a pain too often. I like guns for much of the story, but I use the Sag Bow over guns late game. I can't deny how awesome guns are in LLGs though.

I don't like Staves, except the elemental boosting ones. They can be pretty amazing mid game. Late game, I find magic to be too much of a pain to use over physical attacks most of the time, and if you level your characters up, Staves offer little to nothing over going Main Gauche + Shield. You can meet the magic and damage while having better defense against blockable enemies. Rods are just sad in how bad they are late game. I thought the Faith Rod would be amazing like in FFT, but it's just terrible.

Measures are awful, bombs are awful, and hammers and axes don't beat sword damage over time and swords are more consistent. And Unarmed attacks even outdamage swords over time if you have the Strength to support them.

I never used Maces that much. Didn't see a point in using them, really.

Crossbows are convenient for the speed of their attacks, but I don't ever use them seriously.

EDIT: Oh, duh. I really like spears too.

12

Final Fantasy XII / Re: RNG and ZS Chest Appearance

« on: December 02, 2009, 08:00:39 AM »

(But alas, every time you get my hopes up, you dash them in the next paragraph.)

Yeah, but I don't think there will be many major problems with the Knots of Rust idea. I don't have it in me to start today and I have things to do in the next few days, but I think it will work. It won't be tied to level or stats and the numbers will be much easier to look at.

You will either have to get Knight of the Round rank and buy them (most convenient if you've done it), or you'll have to farm Knots from a chest. Once you figure out your starting position (you don't have to save, so you don't lose any Knots), it will only take 2 Knots every time you load (probably, it might take 3 to be sure depending on output variability, but I have ideas for that).

I looked over every area in the game that has a chest with a possible Knot of Rust (thanks again, ff12maps). The best places for farming will be some balance between high chances of getting the Knot (high spawn, low gil, high Knot as the item), distance to reset the zone, and number of possible Knots per trip.

The best places seem to be:

Barheim

Special Op Sector 3 - overall 77% chance every trip for one, and a moderate chance for another, fairly quick to 3 zone

West Annex - overall 90% chance for a Knot every trip without the DA (it's the chest with the Sweep that spawns on top of the ZE chest), but the zone is rough and it's a bit of a way from the crystal or three zoning. The good news is you only need two Knots to get the Dragon Mail from the same zone, and you can even get the Celebrant's Miter in the Zeviah Span too.

Lhusu

Site 3 - Overall 90% chance for one Knot, plus a few more lower chances on the way. Just takes a little bit of walking each trip

Site 6 South - A 61% chance that's really close to a crystal, and also on the way to the chest below.

Site 5 - Overall 90% plus a slim chance at another chest on the way besides the 61% above.

Sochen

A path from the crystal to the Mirror of the Soul and into the Falls of Time - Two slim chances and five 60% chances in the same path. You can save at the crystal every trip or three-zone. Not a long walk to reset, but not short.

Feywood

Antiquity's End, up and around - Low spawns, but 0% gil chance and 90% Knot if they spawn, and not too bad on the reset.

Pharos Subterra

Highest level of the Subterra (the name escapes me) - There's an 81% chance chest in each of the four quadrants.

I have long overlooked the value of Knots of Rust – they were a nuisance and collected like lint – so I would be glad to find another use for them.

Wouldn't it be great if they ended up being used for a method to guarantee items? Turning Knots of Rust into chest contents suddenly makes them not the worst thing ever.

And speaking of chests, the idea of using them as markers is intriguing.

The big problem is you'd be zoning multiple times.

I’ve also been wondering if there could be a way to use blocks (as in blocking an attack) as a marker.

Tried it today (along with using weapons with a high combo rate, trying techs, using counter attacks, and using dodging). Blocking seems to change the number of spaces the attack shifts. If the attack is unblocked, it seems to work as normal, but if it is blocked, it shifts the RNG fewer spaces. That's contrary to the thought that an attack uses more RNG values by default if you can block it. I think the game uses an earlier value to determine the block and if it did block, it doesn't need the rest that normally come with it. As opposed to an unblocked unarmed attack moving 10 spaces down and most other weapons moving you down 11, a blocked unarmed attack moves you 6 spaces (I believe).

- there may be more than one reason for large advances (though I can’t think of even one yet), so a theory for one zone may not be wrong even if it doesn’t fit other zones.

True.

- thinking about RNG uses not initiated by the player, (e.g. enemies and neutrals casting buffs or using Regen); such things could throw off any system of RNG manipulation.

Absolutely. I'm thinking that it will be worthy to note how some zones are constantly moving the RNG, so someone doesn't attempt to Cure Shift (or whatever) in them and have no idea why it didn't work.

When I first tried Grand Master’s method, several times things went strangely amiss.  Now I wonder if while I was counting, some nearby enemy cast a buff and derailed my RNG progression.

Things went amiss for me too. I haven't had any problems since, and I usually could figure out what I did wrong, so I don't think it was an enemy messing things up.

(Is it possible for enemies at the far side of a zone, not yet visible, to cast buffs and use RNG?  And imagine what those enemy-on-enemy battles in Henne could do to the RNG!)

I think it's possible, yes. I don't know for sure exactly how close they have to be to actually affect the RNG, but I believe that they can do it when not in your sight. I'm pretty sure they can. They have to be moderately close though, I think. Enemies you left behind don't ever seem to mess up my getting the chest in Henne or Cerobi. Or Barheim or any place I've done this, actually.

Gizmaluks fighting Abysteels can affect the RNG. At the chest, they're too far away to be doing things, which is why it's best to get to the chest before using FFGM's method and why it's nice to clear the area when you get to the chest with the Immobilize glitch.

13

Final Fantasy XII / Re: RNG and ZS Chest Appearance

« on: December 02, 2009, 02:40:05 AM »

Yes, my set-up was okay. I checked it multiple times. The values match, it's just that the advances were big.

Thought so. I messed around with it some today and it's not a setup thing.

I don't know what's on the other side and I also don't know if random factors from surroudning zones can affect the RNG. I'll just try some other zones and try to finish my experiment.

Btw. what do the areas you mentioned have in common? They all seem to have rare games.

Yeah, that's the real mystery. What do these have in common? I thought perhaps it may have something to do with rare game that spawn based off the clock or time spent, just like you figured out. Ishteen, by the way, has increasing chances as time goes by, which would require frequent values from the RNG. It also spawns in Zeviah and in East-West Bypass (I usually fight it in the room that the Minibug spawns in).

Now, that was just a random guess, because I have heard that enemy spawns aren't based on these RNGs, but I was grasping at straws. The problem, really, is how all of these zones don't directly have rare game based on time, but some are in the same general area. Walk of Mind is the really odd one. It has leaps in my game too, but it does not have the same behavior in either of the bordering zones in the Stilshrine. But there's nothing (that I know of) that needs a steady stream of values in the zone.

I did some tests, not very much and I haven't rechecked, so nothing is confirmed, but the areas that had odd Cure behavior were the Slumbermead and Walk of Mind and also Cloaca Spur No 4, and once on Needlebrake (but not on the next attempt, I could have messed up). Cloaca did show some light skipping, I'm unsure of Needlebrake, and none of the others jumped out, although I think I may have seen the RNG skip one number or two once early in the shift, but it wasn't a constant alteration in behavior. That could even be due to enemies doing something every once in a while, if I did actually see a skip.

Also, no adjacent zones to any of those zones have the same behavior (I didn't test Ozmone), not even the ones that border zones that do act funny.

At this point, I can't figure out why it happens in Walk of Mind at all. My theory now is that it's something that if it even has a purpose, it will be hard to figure out. It could be something the game is doing behind the scenes that doesn't matter much to the player or is related to a different zone or function, or it could be something that isn't necessary anymore but was left in the code. It could also have a real reason related to the zone that I just can't think of.

(Does “NPC” = neutral character?  like Garif warrior?)

Yes, sorry. The Garif is what I meant by NPC. It means "Non-Player Character", and is often differentiated from an enemy.

In the three zones where I found the unexplained RNG advances there are neutral characters.

Hm... you're right. Good eyes. There are guys outside the Stilshrine. Hm, I didn't think they were in-battle objects, but maybe since they're not in a town zone, the game is doing something with them.  I don't know if it's the neutral fish in the Slumbermead, but it seems to skip by big leaps in that zone on my game. Perhaps killing them will have an effect.

In my game there is a domesticated chocobo strutting around in the east half of The Shred

Yeah, so you don't get stuck. It doesn't even take any greens if you don't have any.

I think that's an idea worth pursuing, although I imagine a spawn based on time would only require the clock, not the RNG.

Not if it was randomized. The ones that spawn 100% when the clock is right probably wouldn't but I was just taking a shot in the dark anyway.

EDIT: Mainly @ Gramps: I was also thinking about other ways to determine your starting position and it's not jumping out at me. Shades of Black would be nice (if it uses these RNG patterns) because the damage wouldn't matter (you'd be using the spells to determine placement). The problem is that it would move at least 3 spots, 2 for the spell it cast and at least 1 for determining its spell. That would potentially remove level and magic from the method, and the output is varied enough that you might not have to have a large list, but you would need three lists (at least). The output is easier to handle, not being numbers but names of spells that players will recognize.

The problem is you can't use it on yourself. You can't use it on an ally either. Gambits don't work on yourself or your allies. You'd have to do it on an enemy. It would need to not kill the enemy (which might be a problem if you're strong), and you'd have to neutralize the enemy's actions while taking into consideration any changes that got through. It would need to be a zone that behaves normally with regards to the RNG and doesn't have enemies that constantly use actions on themselves. Problems, problems.

Now that most techs are out, looking over everything, unless I missed something that can be used as a marker, the best idea are Knots of Rust, which would need to be replaced, or something to do with attacking that doesn't use stats and level.

Knots of Rust varies randomly and doesn't factor level or magic directly, but it does use current HP. You'd have to get your character to a specific HP value and then use Knots on your allies, but it might actually work! And it would be funny to have a use for the darn things. Your numbers could potentially be much smaller (and easier to handle), at potential risk of repeating numbers more often.

Also, this would be pretty crazy... but you could actually use chests themselves as markers (items, spawns, and/or gil amounts)... the behavior of the RNG when you zoned in and out would need to be predictable though.

UPDATE: Knots of Rust will not work. They move by one spot, but they just do too much damage. The formula I had is incorrect; it does not go by current HP. Through Infuse and Sap, I got the character's HP down to 1 and that character still does tons of damage. That means that the game likely considers Bubble an increase in max HP for Knots.

UPDATE II: THE REVENGE Knots of Rust will work, heh. I think, at least. Okay, so I was trying a bunch of different ways to determine your place in the RNG. The problems with Cure Shifting are how cumbersome the chart is and how dependant on Magic and Level it is. Everything I tried that addressed some of these problems had more problems of its own. Mostly with techs needing enemies and with attacks not moving straight down or straight across, requiring multiple solutions for a single chest and making the process less applicable to other chests. Knots of Rust have advantages but do too much damage. But since Bubble affects Knot of Rust damage, but current HP doesn't (max HP does), that means the game treats Bubble as a change in max HP. Is there anything that can make the game think your max HP is lower, so as to lower your Knot damage? Disease.

You can Infuse and then walk a few steps. After gaining a few MP, you Infuse again and your HP becomes your MP x 10. If you are then inflicted with Disease, the game thinks you max HP is that low value, and your Knot damage hits rock bottom. As a special bonus, you don't have to reset Disease. It stays until you touch a crystal, so no messing with the RNG during the process is required.

Depending on your HP, your damage output will be pretty varied. At 10, it's all single digits. It moves one space at a time through the RNG, doesn't combo or do anything weird, won't kill anyone, and isn't based on stats. Unless there's some problem I'm not thinking of, the only remaining problem is that it requires a chart, but the numbers will be much easier to read and work with. It's also unfortunate that you won't be able to do this effectively until you can buy Knots of Rust (unless you want to farm them, heh), and it's unfortunate that you have to go restock, but it shouldn't take you very many at all to find your starting range in the chart and use it to find your solution. Then, 2-3 Knots every time you load will tell you what pattern you're in and you can cast Cure and hit yourself to the spawn value, walk in, and follow FF12GM's method from there.

I haven't tested to see if it's as reliable as Cure Shifting, but this might solve a lot of the problems with Cure Shifting.

EDIT (yes... again) - <Expletive deleted>!!! You can't buy Knots until the next to highest Clan Rank. *is angry* Chest farming for Knots maybe?

14

Final Fantasy XII / Re: RNG and ZS Chest Appearance

« on: December 01, 2009, 09:21:38 PM »

I tried to make a post about that odd Cure behavior, but my internet died in the middle of it.

First try: Loaded a save file from Balfonheim. I had to zone into Cerobi to start casting Cure.

...

Second try: I loaded up a file inside Henne Mines so I can start casting Cure immediately. I got the same starting point as billygoten on Page 5.

I believe different zones advance the RNG differently. At least, we know that moving from the save crystal to the next zone in Henne doesn't advance the RNG because there are no chests or random factors in that second room, but moving from Balfonheim to Cerobi is different. It will advance the RNG.

As far as the values not matching, did you have the Serenity license, all magic lore licenses, at max level with 71 magic and no status (Faith) that affects Cure values? Not to act like you don't know what you're doing, just covering the obvious bases first.

I went to The Shred in Ozmone and I see what you mean.  Oversize advances in the Cure values.  I got suspicious of the Wu hanging around -- they were casting buffs on themselves from time to time -- so I eliminated them, but still got funny Cure advances.

Then I noticed a really big advance after I let my character stand idle while I made some notes.

I was going to mention something similar to this, actually. Sitting around and waiting to see if the advance got greater with time.

My focus was on the Wuu and the Garif guy that appears sometimes. I wonder if you killed him if it would stop moving forward. It's even possible that the game could keep using RNG values if he's set to spawn (or did spawn, which is more likely), but it uses these RNG patterns to determine his buffs when you enter that other zone. At the very least, the structured advancement seems to indicate that it might be some other character other than you that's behind it. Either that or for some reason this zone needs to burn RNG values periodically for some reason. Also, doesn't another rare game appear once you kill that croc?

Maybe Cures advance the RNG farther in The Shred.

For what purpose? I doubt it. The game is moving the RNG for something else, not for your Cures, most likely. That's supported by the RNG moving along despite you using Cure. I bet if you started using something else and seeing how much the RNG advanced each time, it wouldn't conform to how much it was supposed to move either.

I'd say it's an NPC or an enemy, but this is all just what I think is going on, just a guess. The game itself could also be pulling numbers for some reason.

The remote half has the entrance to Henne.

And to the other area in Ozmone.

(Edit) Just teleported and zoned around checking Cures and got 1 RNG advance (almost) always.  But I hit two other areas that behaved somewhat like The Shred: (1) The Slumbermead in Nabreus, and (2) the Walk of Mind at the Stilshrine of Miriam.  The latter seemed really strange.

Did it behave the same elsewhere in the Stilshrine though?

Also, I have another theory on what could be causing it (the Walk of Mind is very telling). I wonder if the same thing happens in East-West Bypass in Barheim, No. 4 Cloaca Spur in Garamsythe, Shunia Twinspan in Lhusu, Golmore's Needlebrake or the Feywood White Magic's Embrace.

I still think -- as you suggested -- that the Cures advance 1, but something else is burning through the RNG in these areas.

Yes, I agree.

15

Final Fantasy XII / Re: RNG and ZS Chest Appearance

« on: November 27, 2009, 11:03:26 AM »

yeah I'm in the last group you mention, I HATE Henne Mines leveling and won't do it. Boring beyond belief.  Anyway, I went to level 99 in my very first FF12 game and won't do it again. The game is too easy anywhere past 70

I know. An admittedly foolish and shortsighted assumption on my part.

I had some words regarding some of the other ideas discussed, but I'll just echo FF12GM and say that I don't see any reason to reinvent the wheel. That method is already solved and is simple and accessible; I don't see any advantages of using another weapon or combo rate when unarmed attacks don't hurt you and are perfectly predictable.

That being said, this whole Treaty Blade thing might be a case of me not understanding zero's point or being ignorant in the mechanics he's talking about. Have no doubt that I'm ever interested if you have something to share.

If anything, Cure Shifting needs tweaks. I haven't gotten to work on a low level game because I do not like the idea of making 5 huge Cure lists to "solve" the problem of not being able to use the method at a level besides 99. It's late and I might not be thinking clearly, but the only solutions I can think of is a perfectly reliable formula for what to expect from Cure or skipping Cure for another action that is more accessible.

Also, I apologize for not doing much lately to help, other than type.