Map seeds effect on drops

[Shagsbeard]

Two things have exacerbated (sp?) a personal myth that I've had for some time and I want the myth either supported or refuted.

1) I have played a lot of SP and have had Shako's drop three times. Two in the last week in exactly the same spot, using the same map. I don't have -seed flagged, and the shako's have different defence and fingerprint. But they were from the same monster type in the same spot. So 2/3 of my shako's have come this way. Way too wierd for me.

2) People claim to have played for ever and never seen Duriel's shell. I've played a whole bunch and have some relatively common uniques missing from my acquisitions. Could this be a function of map seed? If a player was to play on the same seed for years and years, never changing it, would there be items that would be impossible to obtain that would be relatively easy on a different seed?

[pncwd]

Two things have exacerbated (sp?) a personal myth that I've had for some time and I want the myth either supported or refuted.

1) I have played a lot of SP and have had Shako's drop three times. Two in the last week in exactly the same spot, using the same map. I don't have -seed flagged, and the shako's have different defence and fingerprint. But they were from the same monster type in the same spot. So 2/3 of my shako's have come this way. Way too wierd for me.

2) People claim to have played for ever and never seen Duriel's shell. I've played a whole bunch and have some relatively common uniques missing from my acquisitions. Could this be a function of map seed? If a player was to play on the same seed for years and years, never changing it, would there be items that would be impossible to obtain that would be relatively easy on a different seed?

I would like to expand that question to the realms as well, because I did something very similar no to far back, it sticks in my mind because it was really wierd.
I was running the countress and just as I killed her the lag kicked in, I could not pick up any items but did see what was on the ground, Not much btw, but there was an amn, tal, and el rune. So anyway once the CI took place I tried to rejoin the game, Game does not exist. Ok so using my naming convention that I use to make games I remade the game with the same name, got the same map, and then at the countress got the same drop. So although mine might be a little different in the sense that the map on the realm is not set as it can be in single player, what could have caused that to happen.

[Shagsbeard]

One of the reasons I posted was the post in SP forum from a guy who says that he's played the same character in hell for years and has accumulated enough runes to cube up to a Cham, but yet has never seen Duriel's Shell drop. I discount the guy as a liar or a cheat, but wonder if it's just a factor of his extreem play style. Maybe his map is "pro-rune" and "anti-breastplate". I doubt it, but it's got that corner of my brain itching.

[krischan]

The influence of the map seed on drops is like the influence of a butterfly on the weather. It is one of the factors which will decide if a storm will come up or not, but there are millions of other factors as well. Regarding drops, other factors are the interaction between your actions, the monsters' reactions and your counter-actions etc. All theses factors together decide about what will drop and there is no way to know what exactly will happen. So the answer is: To your own knowledge, the influence of the map seed is irrelevant. You can't make predictions about how its influence will affect drops.

[Shagsbeard]

You can't in the normal course of play, because your map seed changes so often. Each new character, each new difficulty, or on Bnet, each new game gives you a different seed.

But if you were to restrict yourself to a single seed, never to change, would that have a noticable long term effect on the availablity of certain drops?

[TheJarulf]

You can't in the normal course of play, because your map seed changes so often. Each new character, each new difficulty, or on Bnet, each new game gives you a different seed.

But if you were to restrict yourself to a single seed, never to change, would that have a noticable long term effect on the availablity of certain drops?


If you use the same initial seed for the game, commonly known as "map seed" although it do affect other things too, you will get a similar game in many aspects. Since the randomness of the game varies in part based on the behaviour of the player, many things in the game will still be different although some will not. This has been explained in detail in various old threads. I sugest making a search for "map seed" or something similar.

As for similar drops (or rather the same item) from same monster or chest, yes that can happen if you start with the same seed. However, you can also get it from two different start seeds. You mentioned that the "fingerprint" of the two identical items were not the same. Then they are, from the games point of view, two completely different items (although with sam stats). It was then NOT the effect of the dentical initial seed of the game. For that, the fingerprint of the items would also have to be the same. The "fingerprint" is in fact nothing more than the current seed at tht moment the item is to be created. When you see an initial seed creating the same map, the same monsters and the same drops, the seed and hence the fingerprints are the same. So in your case, it was just the randomness that managed to create the same item fom the same place and monster by chance.

As for what uniques one see and don't see, it is all about randomness. Out of all poeple that play the game, most will have som euniques they have yet to come across, while others they have seen a lot. So nothing strange with that. That is the result of randomness.

As for the realm creating two identical maps, no idea. The realm should as you say not use saved seeds to start new games. BY the countess, do you mean the items that pop out of the chest? Or from herself? Drops from monsters tend to not be identical even if you use the same start seed for the game/map for various reasons unles you managed to behave EXACTLY the same in both games. And I mean EXACTLY asin how you move over the map, cast spells and so on. Chests are slightly different but I don't recall details by heart, better to search for those previous threads. You might also want to include other sites in that search since I recall this also being discussed in depth on for example the Amazon Bassin and Lurker Lounge forums, not sure what exactly has been written here and in other forums although I know I have participated in all sites :)

[AshenPlanet]

.

I had a map in single player 1.11 where I could get a sacred targe from the armor rack in mephisto's room.
As long as I went to that rack using the same path, I would always get a sacred targe.
It didn't matter how many monsters I killed along the way or how many or what spells I cast.
If I didn't deviate from the path, I got a sacred targe, period.

I was shooting for a +45 resist all, but I had to settle for a +43 resist all after about 50 runs or so since my inventory and stash were completely full with good stuff.
I would kill mephisto after I checked the rack every time, and some of the other monsters sometimes.

Mephisto, and all the other monsters always dropped different random stuff, so their seeds were generated on the spot.
Their seeds may have been based on the original map seed, but following the same path, or timing or casting, did not generate the same items.
The rack and chest seeds however, were definitely generated from only the original map seed and the room seeds that were along the path.

The items that dropped from the racks and chests had different item seeds though.
I'd get all different kinds of auto-mods on the sacred targes that dropped, and different qualities of the same weapon from the weapon rack which even dropped an artifact once.
That's why I kept running hoping for the +45 resist all, although I settled for a +43 resist all.

.

[TheJarulf]

I had a map in single player 1.11 where I could get a sacred targe from the armor rack in mephisto's room.
As long as I went to that rack using the same path, I would always get a sacred targe.
It didn't matter how many monsters I killed along the way or how many or what spells I cast.
If I didn't deviate from the path, I got a sacred targe, period.


Yes, because the item from the rack would use the racks seed which is set once it is created, that is, once you get close enough to it for the game to generate the map area arround it.

A monster will typically have its seed changed several times (due to random rolls in AI, firghting and such) done before you kill it.

Blizzard has changed the way the initial seed for an item is fetched many times over the time of the game. I think that currently the game actually doesn't use the seed from the object like I might have implied but a general seed stream for the map. That means that as long as you don't spawn any more or any less "units" (which includes items, monsters, spell effects, even invisible ones and so on), you will get the same item. For monsters it is still slightly different I think though but check threads searched as I said in previous reply.


So in conclusion, yes, that is the behaviour you would expect and it is especially common for a rack.



Mephisto, and all the other monsters always dropped different random stuff, so their seeds were generated on the spot.


See above, the reason is not the "on the spot" though.



Their seeds may have been based on the original map seed, but following the same path, or timing or casting, did not generate the same items.


Because many things, especially involved in fighting, even more so when there are spells, arrows and such involved, will affect the seed.



The items that dropped from the racks and chests had different item seeds though.


I don't recall the details, but the fingerprint seed is not the same as the initial one that pick an item.

There are, or have been in many versions, two different seeds used and they can come from slightly different places. There is one seed used for picking what item to drop, its quality and so on and another seed used for the magical protperties. The fingerprint seed saved, should be the one used for magical properties.

For a rack, the game does not use the TC drop system and uses a small, custom routine for the rack.


I'd get all different kinds of auto-mods on the sacred targes that dropped, and different qualities of the same weapon from the weapon rack which even dropped an artifact once.
That's why I kept running hoping for the +45 resist all, although I settled for a +43 resist all.


Yes, because the seed for the magical properties was still affected enough by the games happenings up to that point while the seed that picked the base item from the rack was the one it got seeded with when created, which could change if you vary your doings in the game more prior to going there.

Please note that I am wirting this from memory and it has changed in details many times in various patches, but basically that is it. It IS explained in detail in vaiour other older discussions though. Check them out.

[AshenPlanet]

.

We are getting close to the answer.

I think that currently the game actually doesn't use the seed from the object like I might have implied but a general seed stream for the map. That means that as long as you don't spawn any more or any less "units" (which includes items, monsters, spell effects, even invisible ones and so on), you will get the same item. For monsters it is still slightly different I think though but check threads searched as I said in previous reply.

It's impossible, I think, to not spawn any more or less "units" on the way to mephisto.
Even with the same map seed, sometimes you get bone dolls, sometimes zombies, sometimes urdars, sometimes vampires on the way there.
There's gonna be 3 out of the 4 possible mobs on that level, and they will spawn in different places, sometimes you get a boss pack on the way, etc.
There's lots of possibilities for these monsters which, I believe, are not seeded solely from the map seed, but also other factors like the internal clock, etc.
Sometimes I kill a few, sometimes many, sometimes none, I may stop to pick up an artifact, I might not.
All those possibilities aside, when I got to the armor rack, it gave me a sacred targe every time without fail as long as I stayed on the same path.
That's why I believe that it was generated solely from the map seed and room seeds along the way, and not affected by monster or items generated along the way.
The items themselves as well as the monsters and their drops were always different, and so were probably useing some constantly changing seed instead of or in addition to the map/room seed.
These seeds may not be generated "on the spot" but are constantly changing, and don't give the same result during a run from the same map seed.

Note that there are some chests that exhibit that rack behaviour as well.
Certain chests wouldn't always be there, but when they were there, they'd drop the same item (these chests always dropped only 1 item).
Other chests dropped multiple items potentialy, but were different items each time.
It may be that the sometimes appearing chests used rack code instead of chest code or something, I don't know, it's peculiar though, and reproducable.

What this behaviour seems to indicate is that the game uses at least 2 different evolving seeds.

1 based on map and location seeds which is not affected by monsters/items/spells/attacks/time/etc.
1 which may be based on map seed, but is affected by any number of variables like time/spells/monsters etc.

I remember in a previous version, the item drops from monsters were also based solely on the location of where they were killed.
This was exposed and reproducable with the necro bone wall spell which could drop items back then.
That spell was fixed and monster drops changed, but certain things like racks and certain chests may still use that location system to determine item type although fingerprint seed is definitely from a constantly changing seed.

.

[TheJarulf]

>There's lots of possibilities for these monsters which, I believe, are not
>seeded solely from the map seed, but also other factors like the internal
>clock, etc.

I would, again, advice people to search for previous threads on this topic, both here and on other forums.

I don't recall how the game handles different maps (like durance level 2 and durance level 3) they might not affect each other for example.

The game actually handles a whole bunch of different seed streams (if you want to call them that). What you call the "map seed" is in fact the initial game seed, from which all other seeds in one way or another are derived. The clock is only ever used for the initial seeding unless you set a specific map seed instead.

Note that spells are also "units".


>That's why I believe that it was generated solely from the map seed and
>room seeds along the way, and not affected by monster or items generated
>along the way.

Not sure what you refer to by "room seed". However, for item generation, there are basically only a few sources for the seed. The game seed sthread, which is basically only used when you initialize a new unit (and possibly when initializing a new map, I don't recall any longer). Or the seed of the unit that drops the item (for a rackm this would be the same as when it was initially generated and based on the game seed at that time.

If I recall correctly, and don't mix it up with D1, the game saves the initial map seed and uses it as a seed stream specifically for use by objects as well. That is in part why objects, such as chests and reacks are much less randomized than monsters. Again, I don't recall the details though. Sorry. Seems I am mostly telling I don't recall much at all :)


>The items themselves as well as the monsters and their drops were always
>different, and so were probably useing some constantly changing seed
>instead of or in addition to the map/room seed.

As I mentioned above, there is a seed on the monster that changes for anything it do, I think that is the one used for magical properties and such. The global game seed would change for any spell cast for example, Mephisto for example cast spells if I am not mistaken as do many other monsters on level 3 of Durance. The seed for the "room" is not used for anything related to items or monsters.


>What this behaviour seems to indicate is that the game uses at least 2
>different evolving seeds.

It has a whole bunch, one for each active unit for example :)


>I remember in a previous version, the item drops from monsters were also
>based solely on the location of where they were killed.

That has never been the case.

>This was exposed and reproducable with the necro bone wall spell which
>could drop items back then.

No, it was a wrong conclusion based on an observation. it had to do with uninitialized seeds and how the game regenerated monsters (the bone wall is a monster) when you go back to a former unactive room. The game did not regenerate the bone walls correctly leading to them droping items while not being initialized for that, hence giving strange item creation. Thinnk of it as using an uninitialized variable or one initialized for a completely different purpose. Another thing not geting proper recreated was the timer of the walls which were not saved, and hence all walls in a room would collapse at the same time even if they had not been cast at the same moment.

So no, it had nothing to do with the location of the death, it had to do with non correctly recreated monsters that did not have its data correctly initialized.


>That spell was fixed and monster drops changed,

The spell worked correctly, it was the regeneration of monsters when you left an area and came back that was vbuggy and got fixed (making sure the bone walls got the correct flag set on them so that they don't spawn items upon death). Item generation worked perfectly and didn't needed any change or fix, neither did the spell.

> but certain things like
>racks and certain chests may still use that location system to determine
>item type although fingerprint seed is definitely from a constantly changing
>seed.

No, they don't and have never done.

[AshenPlanet]

.

Very Insightful.

I don't recall how the game handles different maps (like durance level 2 and durance level 3) they might not affect each other for example.

There can be racks and certain chests which spawn on dur2 that exhibit the same property of dur3 racks of not being affected in any way by "units" (inlcuding spells/monsters/items/etc.).

Not sure what you refer to by "room seed".

In "room seed", I am referring to location.
As stated before, spells you cast, monsters you kill, items that are generated do not affect the "seed stream" for racks and certain chests (most chests are affected by "items" and will drop differently every game).
What does affect these drops is location - what I referred to as "room seed".
In room seed, I am not referring to a whole room, but whatever part of the room that d2 considers a section group (or extended tile).
eg. following a "path" may involve going down the center of some hallways, the right side of others, and the left side of some other rooms.
Deviating slightly (which any human will do) is allowable and will not break the path giving the same room seed stream which produces the sacred targe at the end.
Passing down the left side of a hallway instead of the right, for example, would break that room seed stream resulting in some other item propogating from the ending rack.

Thinnk of it as using an uninitialized variable or one initialized for a completely different purpose... No, they don't and have never done.

This is enlightening, but it is a reproducable effect of being able to get the same item from the same bone wall multiple times if the bone wall was in the same "location", and other locations would drop other items, but would again always drop the same item at the same locations, iirc.
It has been a long time since the bone wall bug, and I might be remembering aspects of it differently, but that's to my recollection.
The bug you explain above may cause d2 to revert to location as the only determinate to item type.
This bug you explained may still be in use in 1.11 when determining item types for racks and certain chests which exhibit the behaviour listed above.
Monsters and fingerprint seeds do not exhibit that property, and so it is definite that they are using a different system than used for rack/cchest seeds.
Perhaps rack/certchests are missing/skipping a portion of a multi-part sequence for generating aforementioned seeds.

.

[jrlafrance]

I had a map in single player 1.11 where I could get a sacred targe from the armor rack in mephisto's room.
As long as I went to that rack using the same path, I would always get a sacred targe.
I did several Countess runs with my Frenzy barb. Obviously using the same map (the Tower was really close to waypoint). I found a rack that I would get some form of barbarian helm very often ( I don't recall right now if it was a Jawbone visor, or what), and lots of times it was ethereal too. I underlined "very often" because I didn't get a helm ALL the time. I bolded your "always" thinking that it would be strange to get the same item type EVERY time. Has this happened with others too, getting the same item type EVERY time?

[AshenPlanet]

.

This is a reproducable effect that you can duplicate with an appropriate map and careful methods.
The barbarian helm you got very often means you followed the same path getting there very often.
If you could follow the exact same path every time, you can get the same helm every time.

Note that to do this, it's far easier to use a sorc who can teleport the entire way there without walking/running.
Also, it's not neccessary, but can be helpful to hug walls as you go.
You can vary slightly, but as I mentioned above, the "room seed" as I called it does not encompass the entire room, but an area that you move through.
If you trigger these areas in the same sequence, the ending trigger (the rack or certchest) should be the same.

Note also that some racks and the aforementioned chests may not always be generated in your map on any particular run.
When they are there however, they'll give you the same drop.
The racks in mephisto's room are the easiest to show this effect because they are always there.
Although, depending on how far you have to go through dur2, it may be difficult to find a consistent path.
If not successful, you can generate a new map to aim for a shorter run that would be easier to duplicate on a regular basis.

I would suspect that someone who wrote a bot could follow the path very easily regardless of the maps generated.

.

[TheJarulf]

AshenPlanet wrote:


>In "room seed", I am referring to location.

Still don't get it. What location? I would think you mean x-y coordinates by location. Alternatively a tile or if we go up in data structres, what the game indeed call a "room" which is basically a map segment a few screens in size. That is basically the only relevant locations that exists (we can go further up to complerte maps and acts of course). However, there is no relevant seed for any of those. The "room" used to have a seed and it might be used long ago in early versions but I think that back then but I don't recall the exact details of every version.

Also note that by map seed, I would typically think of the original game seed with which the game is seeded or started (which you can set as well if you like and is preserved if you play single player, right?), this seed actually will then affect everything, not just tthe map. It will for example be used by the map generation. It will however also, currently, be used when creating new units (and eveything in the game is really a unit, players, monsters, items, missiles and so on). This does not grab a seed from the "map" but from the global game data structure.

I think it might be in place to slight touch on how the game really keep data and such, and were it keep various seeds. First of all, there is the global stareting seed of the game, as mentioned. This is stored away. It is also used to initialize the stream of seeds (which can each be used for either one random number generation, or to initialize another such seed stream (two such streams will never be idential, even if based on each other, due to how the rng worls in D2). When the game creates a monsters, item or whatever that need random numbers in some way, it currently (at least in 1.08, 1.09 and 1.10 if I recall correct), initilized the seed of the unit from this game global seed stream.

For items, before there even is an item that needs initializtion, the game need still to determine WHAT item it should be, that is, what base item and what quality and so on. this is done by the seed stream of the unit generating the item. From a monster, it it s the same seed stream used for everything else random the monster do, and thus can change, just by the monster walking arround. from a chest, there is no such change since chests typically do nothing random. Thus, the seed the chest got once created, it still has when it is time to drop items, same with racks.

Monsters, chests and such are generated when a player get close enough to the "room" (a part of the map) they are part of, then the game create that part of the map, including populating it with monsters, chests and such. Now, if I recall correctly from the threads were I detailed on how the mapseed command works, chests and other containers are slightly special in that when they are created, the game actually get their initializationn seed frmo a special stream of seeds (based on the initial one, but seperate from the game global) which is stored away (don't recall if it was still on a global bassis for all containers in the game or on a per map/act/whatever). hence they are more predictable than monsters, including items from them.

To my knowledge, there are no other seeds inviolved than those (at least not of improtance).


>As stated before, spells you cast, monsters you kill, items that are
>generated do not affect the "seed stream" for racks and certain chests
>(most chests are affected by "items" and will drop differently every game).

Almost correct see above. Magical properties should still be affected since they get them from their own seed which comes from a common seed stream (unless they changed it again in 1.11).

>What does affect these drops is location - what I referred to as "room
>seed".

And what seed stream is that of the above? Were is it kept and what is it used for? As I mentioned, containers (and I think racks count), have their own special on.

>In room seed, I am not referring to a whole room, but whatever part of the
>room that d2 considers a section group (or extended tile).

See above, whoever, there is no seed used for such a room. it is more global, for the whole game really as far as I recall.


>eg. following a "path" may involve going down the center of some hallways,
>the right side of others, and the left side of some other rooms.
>>Deviating slightly (which any human will do) is allowable and will not break
>the path giving the same room seed stream which produces the sacred
>targe at the end.

No, it would not in any case since the game works with maps in much larger chunks, they are the size of a few screens wide.


>Passing down the left side of a hallway instead of the right, for example,
>would break that room seed stream resulting in some other item propogating
>from the ending rack.

That I can't see at all, unless you happen to be in some exact division of rooms. On the other hand, this scentence seems in direct contradiction to the previous one, so perhaps there is some error.



>This is enlightening, but it is a reproducable effect of being able to get the
>same item from the same bone wall multiple times if the bone wall was in the
>same "location", and other locations would drop other items, but would
>again always drop the same item at the same locations, iirc.

Now you bring in bone walls which are something completely different. For a started they are monsters, and second they were wrongly initialized (or buggy if you want). THAT is what caused their strange drops, not item generation in general. The drops had nothing to do with their location, you are drawing the wrong conclusion. besides, it would be dead easy to verify the falsness of that statement, since killing a monster in the same spot would yield a completely different item, even if you stand in the exact same spot and kill it in the exact same spot.

The reason they all droped the same, and also all of them collapsed, had to do with the error when they were regenerated. First of all, you had to spawn them, and recall, they are monsters from the game's point of view. Then you needed to go so far away so that the room in which they were present (room, refering to the map segment as I detialed above) was turned inactive by the game. The game does only keep those rooms there is a player in or near active. Others are stored away in a compressed form and not updated at all. When you got back, the game regenerated it all, including puting the bonewalls back. However, several things did not work out well. The timer for the walls did not for example since it was not stored away (or at least not restored) and thus all walls within such a room would collapse at the exact same time regadless of when they were cast or how long they had been up. Same with a flag indicating the monster should not drop item. It was not set. Hence when the bonewalls collapsed, the game tried to create an item for the, However, there were no data set up for them creating an item since none of the proper variables were set. In such a case, anything goes, the game could have crashed (althought that would have been cought in play testing of course), or the game could create the most strange and bizzare items. As it happened, it tended to create similar items all over. However, that was NOT due to some location being a factor in the item ceation. Further, since we talk about seeds, it would require the exact same SEED, not SEEDSTREAM, since that would be updated for each item/wall/whatever. Right? So if your theory needs to fit, you need to explain that. Of course, the easiest is to look at the code and how the bug occured and what was happening. Conclusion, coordinates does not matter for item generation, although people put out that theory as they did not understand how the game works (they put out all sort of other even more bizzare ideas though).

In addition, note that the items from bonewalls was only of same type, the magical properties would be different. This fits perfectly with what else I have told you, that is, the magical properties comes from a global seed stream, while type comes from the, in this case, monster itself. However, a bonewall, did not do anyting and hence had no need for a seed and did not get one back then, or at least did not get one properly, hence, it was wrong, could all be the same and so on.


>It has been a long time since the bone wall bug, and I might be remembering
>aspects of it differently, but that's to my recollection.

You recall correctly, but the conclusions and reasons for it is not correct. And indeed, hard to understand if you have not looked at how the game works internally.


>The bug you explain above may cause d2 to revert to location as the only
>determinate to item type.

No, it did not revert ot anything, it tried to create items based on a monsters that had not had its variables set up for item creation. Try writing a program an not initializing your variables. It will most likely work, but you are likely to get the most strange (and even erepeatable at times) results, typically making you think of something completely else being wrong. That is why they can be so treacherous.


>This bug you explained may still be in use in 1.11 when determining item
>types for racks and certain chests which exhibit the behaviour listed above.

The bug ONLY existed for th recreation of inactive monsters. There was no bug in item creation. And no, it no longer exists. Actually, there was even a special assertion included so even if bonewalls for some specific reasons were to be the source for an item drop, the item drop routine will abort just since it specifically check for bonewalls now. It never gets there though since bonewalls are correctly recreated.


>Monsters and fingerprint seeds do not exhibit that property, and so it is
>definite that they are using a different system than used for rack/cchest
>seeds.

The difference is as I explained initially. Still, no "location" involved though.

>Perhaps rack/certchests are missing/skipping a portion of a multi-part
>sequence for generating aforementioned seeds.

The difference is as I explained above.

[AshenPlanet]

.

Hmm, clarifications needed...

Almost correct see above. Magical properties should still be affected...

I stated many times in multiple posts above that the item type from racks and certchests were not affected, but that item fingerprint and properties were affected. Hence you can keep hitting the rack for a sacred targe and hoping for a perfect one or hitting a laquered plate rack and hoping for the set item.

containers (and I think racks count), have their own special on.

Most chests are affected by units, can drop multiple items, and will give very random item types like monster drops. Some chests act like racks.

No, it would not in any case since the game works with maps in much larger chunks, they are the size of a few screens wide... That I can't see at all, unless you happen to be in some exact division of rooms. On the other hand, this scentence seems in direct contradiction to the previous one, so perhaps there is some error.

It may be error, or it may be by design, but there is no question that the order in which you pass through the tiles does affect the ending drop. It does not seem to be using the bigger room size, but rather a smaller tile size. How this works, I don't know, but it can be shown through empirical testing.

The drops had nothing to do with their location... since killing a monster in the same spot would yield a completely different item,

Yes, killing a monster produces random items as I've stated many times, but the bone wall did not produce random items. The game may have treated walls as monsters, but not when they dropped. I can accept what you are saying is the reason for this, I do not want to argue that. It's the behaviour of the drops that I was pointing out as being somewhat similar to rack/certchest drops. It would be difficult to load an old version to do tests on bone wall, and I don't even remember which version I saw this effect with.

The rack effect is reproducable now though and shows behavior that would indicate use of a different seed stream than monsters, item fingerprints, and most chests. By testing, one can see that how you get to a rack/certchest affects what drops, but not it's properties, and this effect does not take place in monster drops.

How this happens, and whether it's bugged code or not, I don't know, but it's behaviour that can be verified.

.

[skoolbus]

The reason why you might never see some seemingly common uniques is the TC system. I don't remember how exactly it works, but there are a few mid level TC classes that are extremely rare, which is why HoZ's, skullders, war travs, etc, are fairly hard to find. I don't think this has anything to do with map seeds, but im no expert.

[TheJarulf]

>I stated many times in multiple posts above that the item type from
>racks and certchests were not affected, but that item fingerprint and
>properties were affected. Hence you can keep hitting the rack for a sacred
>targe and hoping for a perfect one or hitting a laquered plate rack and
>hoping for the set item.

Don't understand what you say. Seems we agree or?


>It may be error, or it may be by design, but there is no question that the
>order in which you pass through the tiles does affect the ending drop.

No, tiles you walk on has no such effect on itself. What can happen is if by walking a slightly different path, for example wake up monsters or generate a new map area to be created and so on. The tiles themselves have no importance. If you see effects it is due to something else like the examples I just gave.

> It
>does not seem to be using the bigger room size, but rather a smaller tile
>size. How this works, I don't know, but it can be shown through empirical
>testing.

As I said, the game DO use the "rooms" which are a few screens. A room is the fundamental area by which the game operates. Tiles are just graphical elements within a room, they have no other effects. That is how it is, even if it seems different to you.


>Yes, killing a monster produces random items as I've stated many times, but
>the bone wall did not produce random items. The game may have treated
>walls as monsters, but not when they dropped.

Sorry, a monster is a monster, including bonewalls, The game can't "treat" it different. Lets see if I can tell this a third time. The bone walls were NOT properly initilized, hence would not create random items. It has nothing to do with being treated differently.


> I can accept what you are
>saying is the reason for this, I do not want to argue that. It's the
>behaviour of the drops that I was pointing out as being somewhat similar to
>rack/certchest drops. It would be difficult to load an old version to do tests
>on bone wall, and I don't even remember which version I saw this effect
>with.

I think it was patched in 1.02 or something like that. I can assure you that bonewalls have NOTHING to do with racks, no matter what version you look at though.


>The rack effect is reproducable now though and shows behavior that would
>indicate use of a different seed stream than monsters,

Yes, I have told a multitude of times that. In addition there are two different seeds, the one used for base item selection and the one used for magical properties (which would end up as the "fingerprint". As I said, the objects, if I recall correctly, has a special stream of seeds.


>How this happens, and whether it's bugged code or not, I don't know, but
>it's behaviour that can be verified.

There is no bug really, it is the way it is designed. ONe may call it bad design or one may call it good design. I have tried to expalin several time in general how the seeds get assigned. For a complete detailed explaination, you have to go out and seek my (and others) detailed explanation in threads regarding map seeds. It seems these forums only have threads from the lats your or so and thus no longer hold those threads, so your best bet is Lurker Lounge or Amazone Bassin which both had such threads as well if I recall correctly.

[AshenPlanet]

.

Don't understand what you say. Seems we agree or?

I think we've always agreed...

No, tiles you walk on has no such effect on itself. What can happen is if by walking a slightly different path, for example wake up monsters or generate a new map area to be created and so on.

The order in which new map areas are created is the key here.
This is why you can walk down the left side of a room instead of the right side and screw up the pattern.

The bone walls were NOT properly initilized

I won't argue about bone walls. I never disputed anything you said about bone walls except that I recalled that it did matter where (x,y) the bone walls were as far as item types generated. It would be too much trouble to install 1.01 to test this.

There is no bug really, it is the way it is designed.

I don't consider it a bug, I have used this "feature" multiple times to farm specific items, and gotten even better things than the +43 resist sacred targe.

In summary, rack and cchests can be manipulated to drop specific item types simply by generating new areas in a specific pattern.
The item fingerprint/stats of those items, monster drops, and most chest drops can not be manipulated as easily since they are affected by a multitude of things in the game (units, etc.).

.

[Cannon Fodder]

I realize that this is a bit different than what you've been discussing, but it brings up an interesting thing I had happen to me some time ago and have only now remembered to ask about. Basically, I was doing trist runs for some friends. I did at least 30 consecutive runs, killing Griswold each time. Every single run he dropped a jawbone visor. As you have discussed, the magical properties on it varied- sometimes it was white, magic, ethereal, etc. Playing this game has shown me many odd strings of luck before, but I just can't understand Gris dropping in such a patterned way. This, by the way, was on the realms~ east scnl. Obviously I wouldn't have been following the same path or casting the same spells, and there were always other players in the game whom were also doing different things each time. Since, I have noticed other patterned drops from him while leveling characters- just today he dropped a shrunken head 5 times in a row or so. Is there any explanation for this or is it just 'one of those d2 occurrences?'