After testing the vanilla behaviour and doing statistical analysis, it does not appear that this fix is actually required. The information on the CS Wiki Leveled Creature page and other sources is incorrect, and the vanilla game correctly calculates each creature spawn independently.
The extra file is also not required.
It's very likely that the 'bug' was perpetuated by the CS Wiki and OOO a long time ago, and no-one ever actually tested it.
How though ... why not compatible with MOO, and how dos one test to see if this mod is working, I tried things seem the same. And that is with an odd combo I'm using with WAC/XCV/TIE with this loaded dead last.Things seemed the same.
You would probably need to deliberately set up a test in the CS to see if this mod is working. If you placed 4 enemies in an interior using a 25% form list and entered it repeatedly, with vanilla you’d get all four spawning 25% of the time, and none of them spawning 75% of the time.
With this mod installed, you would get an average of 1 spawning each time, with anywhere between 0-4 spawning overall in any combination. It’s based on binomial probability, if you’re familiar with that.
MOO will not be compatible because it will inject creatures into leveled lists by script. This will mess up the proportions and so change the spawn chances. The reason why this needs to load last and after the bashed patch is so that it overwrites the chance-based leveled lists with the proportion-based edited ones instead. If a mod adds creatures, they will be added to the 100% chance lists, which this mod doesn’t overwrite, so they will still appear.
Alright, here are the details, and you might be able to see what I'm getting at.
The vanilla LL1Beast25 list is set up like this:
Spoiler:
Show
Chance None: 75 (i.e. 25%) Creatures:
1 :: CreatureRat
1 :: CreatureRat
2 :: CreatureWolf
2 :: CreatureWolf
6 :: CreatureWolfTimber
9 :: CreatureBlackBear
12 :: CreatureMountainLion
16 :: CreatureBrownBear
However, like it says in the description, the 'Chance None' does not work as intended. The workaround used in New Vegas changes the list to look like this:
Spoiler:
Show
Chance None: 0 (i.e. 100%) Creatures:
1 :: LL1Beast100
1 :: [Dummy Spawn]
1 :: [Dummy Spawn]
1 :: [Dummy Spawn]
This means the LL1Beast100 (the 100% chance version of the original list) has a 1-in-4 chance of spawning, i.e. 25%. This simulates the intended behaviour of the original 25% leveled list. The Dummy Spawn is a nested list that has a chance none of 100, so if it is selected, it will never spawn anything. No scripts are used.
You can probably see that if the bashed patch was to insert another creature into the leveled list, the proportion would no longer be 25%, it would be 20% LL1Beast100, 20% the inserted actor, and 60% Dummy Spawns. That's why this mod needs to load at the end and after everything else, so that it definitely overwrites any changes. It doesn't matter if you use a mod that adds new creatures, because it will add them to the LL1Beast100, which is still used in the new list, so the new creatures will spawn as intended.
If you've seen Arthmoor say that this mod uses some form of 'control script', he has misunderstood how it works. The only script is a failsafe script attached to a dummy actor in the Dummy Spawn leveled list. If the list were to somehow spawn the dummy actor (which it never should), this script would just disable the actor so that it is removed from the game. I even wrote that in the script itself.
thanks ... I'm not on anybody's team here and not trying to stir anything up. I think I understand the mechanics being presented, but the main of the last question was whether any one had tested trying to merge this plugin into the bashed patch in a more than generic manner, like with tags. It would seem that if tagged with scripts and loaded late or even right before the bashed patch ... that script would win or be there. Maybe I'm being ignorant. IDK really. My knowledge limited maybe knowing just enough to ask dumb questions or make simple answers complex.Apologies if that is the case.
I was curious, so I actually performed the test you suggested, Push. I placed four enemies in an interior using a 25 % form list and entered it repeatedly, and the outcome wasn't what I expected from your description: vanilla actually spawned any number of creatures from zero to four.
This was in a new game, on a completely fresh Oblivion installation without any mods except for my test plugin with the creature spawn points, not even OBSE nor any OBSE plugins.
I used four instances of the LL1Beast25 list used in your other example.
The most common outcome seemed to be one creature, then zero, then two, then three, and lastly four. I had to perform the test quite a few times to get four creatures spawning.
Having your fix installed didn't make a noticeable difference, from my limited testing.
Neither was it necessary to completely restart the game nor exit to the main menu before reloading. Just reloading normally before entering the interior again was enough to generate new spawn chances. I checked this just in case, because Oblivion famously has other problems with not exiting before reloading.
So this is a bit awkward… Are you absolutely sure that Oblivion even has this bug? It doesn't seem completely broken, at least. Maybe the random number is just a bit flawed?
Let me preface this reply by admitting I'm a functional idiot at times. Like in missing why scripts tag won't work.
OK I think, but I do know enough about statistics to say this much ... those examples seem like a great way to really get lost in the mods and averages to be created. Wouldn't it be better to create outliers or pull from lists where you know that it isn't just a number, because that flumuxed me to Push seems to be saying that the bug will present either 0 or 4 for his example, but with the fix it becomes 0 to 4. Is that correct?
maybe a plugin to test the difference would be helpful?
Hmmm, that’s very interesting. I don’t recall if I ever actually tested it. I was going off information from other authors and the CS Wiki (https://cs.elderscrolls.com/index.php?title=Leveled_Creature). Maybe all run the test and see what the outcome is in different scenarios.
Ok, having run a controlled test 50 times and doing the stats analysis, I'm fairly confident in saying that this 'fix' is not required. The results all fell within what would be expected of a binomial distribution, suggesting that the vanilla behaviour is calculating the chances correctly. It could well be one of those incorrect rumours that used to circulate in the early Oblivion modding scene that has persisted throughout the years and no-one ever tested.
I ran the test without any OBSE plugins, but with OBSE installed, so I'm assuming it is not something that has been fixed by an OBSE plugin, either.
What if I started a game but merged it in to a bashed patch? Could I just change the load order and continue or do I need to start a new game as well? Thanks.
This mod is clean and elegant, and works just as described. Download and install it for livelier wilderness.
I advise waiting to activate it until your character is able to deal with several fights in a row without resting. I expected things to be livelier when I traveled, and yes they are! One trip from Vilverin to Bruma saw me accosted by four bandits, several wolves, two trolls, and a black bear that was my motivation to get all the way to Bruma so the guards could deal with it.
Now that my character is a bit more capable, traveling is certainly more interesting and fun, but before that, I did an awful lot of running. :D
Were any of those spawns above the current level of your character? I did a run in the ayleid ruin kemen east of cheydinhal at level 4 and a skeleton guardian spawned. I''m not sure if this mod is the culprit but i deactivated the mod and it didnt happen the second time i ran the ruin.
This might not be for everyone; depends on your setup. In my case, I added tons of creatures from various mods to the Wabba Summon leveled list, so that when the spell is cast, any one of hundreds of creatures or npcs will randomly be conjured up. Worked fine, until I added this mod at the very end of the load order, after the bashed patch. Then I started getting instant ctds when casting that spell. Not always, just sometimes; it became unreliable. It may be related to this mod or it may not, no way to tell. But since I didn't have that problem before, I'm thinking what this mod does to leveled lists could be affecting it.
50 comments
The extra file is also not required.
It's very likely that the 'bug' was perpetuated by the CS Wiki and OOO a long time ago, and no-one ever actually tested it.
This is why I love you, man! Preservation should be way more appreciated!
10/10
I have yet again posted a comment as a response to someone else. I hope I can stop doing that some day lol
My apologies!
The author mentioned that some overhauls fixed this. Anyone got a list? How about Maskar or TIE?
thanks
Maybe not even Frans.
It won’t be compatible with MOO, though.
How though ... why not compatible with MOO, and how dos one test to see if this mod is working, I tried things seem the same. And that is with an odd combo I'm using with WAC/XCV/TIE with this loaded dead last.Things seemed the same.
With this mod installed, you would get an average of 1 spawning each time, with anywhere between 0-4 spawning overall in any combination. It’s based on binomial probability, if you’re familiar with that.
MOO will not be compatible because it will inject creatures into leveled lists by script. This will mess up the proportions and so change the spawn chances. The reason why this needs to load last and after the bashed patch is so that it overwrites the chance-based leveled lists with the proportion-based edited ones instead. If a mod adds creatures, they will be added to the 100% chance lists, which this mod doesn’t overwrite, so they will still appear.
Don't know if you're pro/con bash. so just curious if that has been tried? Like I could try that then compare this to the bashed patch in edit.
Noted about MOO, it is kind of an outlier though in terms of leveled list type mods. It seems. thx for replies
The vanilla LL1Beast25 list is set up like this:
Chance None: 75 (i.e. 25%)
Creatures:
However, like it says in the description, the 'Chance None' does not work as intended. The workaround used in New Vegas changes the list to look like this:
Chance None: 0 (i.e. 100%)
Creatures:
This means the LL1Beast100 (the 100% chance version of the original list) has a 1-in-4 chance of spawning, i.e. 25%. This simulates the intended behaviour of the original 25% leveled list. The Dummy Spawn is a nested list that has a chance none of 100, so if it is selected, it will never spawn anything. No scripts are used.
You can probably see that if the bashed patch was to insert another creature into the leveled list, the proportion would no longer be 25%, it would be 20% LL1Beast100, 20% the inserted actor, and 60% Dummy Spawns. That's why this mod needs to load at the end and after everything else, so that it definitely overwrites any changes. It doesn't matter if you use a mod that adds new creatures, because it will add them to the LL1Beast100, which is still used in the new list, so the new creatures will spawn as intended.
If you've seen Arthmoor say that this mod uses some form of 'control script', he has misunderstood how it works. The only script is a failsafe script attached to a dummy actor in the Dummy Spawn leveled list. If the list were to somehow spawn the dummy actor (which it never should), this script would just disable the actor so that it is removed from the game. I even wrote that in the script itself.
This was in a new game, on a completely fresh Oblivion installation without any mods except for my test plugin with the creature spawn points, not even OBSE nor any OBSE plugins.
I used four instances of the LL1Beast25 list used in your other example.
The most common outcome seemed to be one creature, then zero, then two, then three, and lastly four. I had to perform the test quite a few times to get four creatures spawning.
Having your fix installed didn't make a noticeable difference, from my limited testing.
Neither was it necessary to completely restart the game nor exit to the main menu before reloading. Just reloading normally before entering the interior again was enough to generate new spawn chances. I checked this just in case, because Oblivion famously has other problems with not exiting before reloading.
So this is a bit awkward… Are you absolutely sure that Oblivion even has this bug? It doesn't seem completely broken, at least. Maybe the random number is just a bit flawed?
Like in missing why scripts tag won't work.
OK I think, but I do know enough about statistics to say this much ... those examples seem like a great way to really get lost in the mods and averages to be created. Wouldn't it be better to create outliers or pull from lists where you know that it isn't just a number, because that flumuxed me to Push seems to be saying that the bug will present either 0 or 4 for his example, but with the fix it becomes 0 to 4. Is that correct?
maybe a plugin to test the difference would be helpful?
I ran the test without any OBSE plugins, but with OBSE installed, so I'm assuming it is not something that has been fixed by an OBSE plugin, either.
My bet is though that few would even have questioned it if it weren't for wanting to be placed in the coveted last position.
I advise waiting to activate it until your character is able to deal with several fights in a row without resting. I expected things to be livelier when I traveled, and yes they are! One trip from Vilverin to Bruma saw me accosted by four bandits, several wolves, two trolls, and a black bear that was my motivation to get all the way to Bruma so the guards could deal with it.
Now that my character is a bit more capable, traveling is certainly more interesting and fun, but before that, I did an awful lot of running. :D
every time i go in to the game i run around for max 20 seconds and oblivion crashes dose anyone have any idea how i could fix this ?