[ Disclaimer, Create new user --- Wiki markup help, Install P99 ]
Difference between revisions of "Project ZEM"
(→Step #3: Normalization) |
(→Zone Zero) |
||
| Line 75: | Line 75: | ||
Priestess Jahnda, Runethar Hamest, Suuspa Clanim, Togahn Sorast, Tonmerk Plorsin | Priestess Jahnda, Runethar Hamest, Suuspa Clanim, Togahn Sorast, Tonmerk Plorsin | ||
</pre> | </pre> | ||
| + | |||
| + | See also: [[Project ZEM: Freeport]]. | ||
Latest revision as of 17:49, 17 July 2025
Contents |
What is Project ZEM?
Project ZEM is an effort to organize the Project 1999 community into killing a variety of fixed-level mobs, in various zones, repeatedly (50 times?), using a character exactly two levels higher ... and recording exactly how much XP they gained from those (50?) kills.
By collecting that data (and doing a little math), we can then add accurate "ZEMs" (Zone Experience Modifiers) to the wiki.
Who's Crazy Idea is This?
Project ZEM was started by long-time wiki-addict Loramin: please direct all questions, rants, flames, and other correspondence to him (preferably through the forum's private message system, although Wiki Talk also works). But of course, this page is part of a wiki, so anyone is welcome to edit and improve it.
The Problem
The Zem page shows the modifiers for every zone in Project 1999 ... but unfortunately those values are inaccurate.
I don't know where the values on the wiki come from, and we do not reference those. We do not publish the specific values, as it's part of the mystery and should be discovered by the players how each zone stacks up.
Project 1999 Isn't (Exactly) Classic
In short, while the wiki's values are based on the original EverQuest values, Project 1999's values aren't, and this was done on purpose to make the server feel more classic.
(Side Question) How do we Know the Original EverQuest Values?
Originally (on live) the EverQuest server sent information about a zone's modifier to the client. Programmers were able to watch the internet traffic on a computer playing EverQuest (with a program called ShowEQ), which let them see the "secret" modifier. At least one player saved all of the modifiers for every zone, and those numbers form the basis of every stock EQ Emulator server.
As mentioned, Project 1999 does not use those numbers. Also, Project 1999 does not send the ZEMs to the client (and they also detect and ban anyone using ShowEQ), which means that Project 1999's ZEMs can't be determined the same way Live's was.
The Solution
Since we can't cheat, the only way for the P99 commnunity to determine the game's ZEMs is to repeatedly (at least 50 times?) kill a mob of level ___ in zone A, and then repeatedly kill a mob of the same level in a different zone, and compare how much experience each zone gave. Only at that point you can definitively say "Zone A gives X% (more/less) XP than Zone B".
However, while knowing that Sebilis has a higher (or lower) ZEM than Sol B is useful, what you really want is an objective number (eg. Zone A has a ZEM of 120% that of an average zone). Calculating that will require killing a lot of mobs, of the exact same level, in two different zones, by characters of the exact same level (2 higher than the mob is probably the most convenient).
Zone Zero
To save ourselves from as much math as possible, we should pick one zone as a basis to compare with other zones (see Zone Zero for that research). Right now it looks like (although more research is still needed) that South Qeynos might be the best Zone Zero. That zone has the following fixed-level mobs (apologies for list format):
1: A Koalindl, A Large Rat, A Mangy Rat, A bat, A gnoll pup, A snake, Tomer Instogle
2: A Fire Beetle, A Klicnik Drone, Prince Klicnik, Princess Klicnik
4: Djerr Darkpaw
5: Guard Aerryn, Guard Erabb, Guard Liben, Guard Simkin, Hurrieta Ironforge, Lashun Novashine
6: Nixx Darkpaw
7: Kraxz Darkpaw, Ziska Ironforge
8: Guard Furithane, Guard Obyn
10: Guard Buce, Guard Deregan, Guard Ginton, Guard Hewet, Guard Rashik, Guard Shorm, Guard Vaskif, Sabnie Blagard
11: Dren Ironforge, Terago Omath, Ulia Yovar
12: Jracol Brestiage, Rolon Banari
13: Aenia Ghenson, Guard Weleth, Moodoro Finharn
14: Leanon Ruksey, Tubal Weaver
15: Flynn Merrington, Guard Drath, Guard Sylus, Nerissa Clothspinner
18: Faldor Hendrys, Svena Ironforge
20: Nax Ghruna
21: Quinon Hulleaf
23: Pelshia Thuxpire
25: Rendallen Ironforge, Whysia Flock
27: Shenro Kazpur
29: Brohan Ironforge
30: Enic Ruklin, Tyokan Mekase
34: Henlom Visrek
35: Guard Elron, Guard Treitan, Guard Wefnin, Segran Rajhar
37: Ghil Starn
38: Berane Constarg
42: Dranom Ghenson, Seta Bakindo
45: (10) Ania Klephia, Balhallia, Enpero Mesog, Gerault Takenmaul, Kragie Rolana, Ladia Qursii, Mellisa Purgor,
Rodgar the Tall, Sneed Galliway, Tanlyn Galliway
47: Crow
50: Corporal Lancot, Hanns Krieghor, Koalindl, Lieutenant Dagarok, Priest of Discord, Umvera Dekash
60: Kane Bayle (NPC), Phin Esrinap
61: (11) Astaed Wemor, Camlend Serbold, Jaylia the Faithful, Knargon Lanenda, Nomsoe Jusagta, Priestess Caulria,
Priestess Jahnda, Runethar Hamest, Suuspa Clanim, Togahn Sorast, Tonmerk Plorsin
See also: Project ZEM: Freeport.
This means that if South Qeynos is Zone Zero, we can calculate a ZEM relative to it for any a zone that has a mob with a fixed level of 1, 2, 4, 5, ... 42, 47, or 50.
However, since players are curious about the ZEM of high-level dungeons, South Qeynos (despite having a great variety of lower level mobs) might not be the best zone zero. Also, we need to consider mob respawn time and how viable targets are for soloing (eg. [Ulump Pujluk]] in Swamp of No Hope would seem to give that zone a 55 ... but because he is an epic quest mob Ulump is not soloable). This is why further research is needed.
What About Zones That Don't Overlap Zone Zero?)
There probably isn't a perfect Zone Zero that overlaps every other zone in the game. However, if Zone A overlaps Zone Zero, and Zone B has a static mob that overlaps with a static mob in Zone A, we can still determine Zone B's ZEM, transitively.
For instance, if Zone A has a ZEM of 120% of Zone Zero, and Zone B has a ZEM that's 120% of Zone A, we would know (through math) that Zone B has a ZEM that's 144% of Zone Zero.
In theory, we can even calculate things across several zones using this approach ... but it requires more work/math (and thus has a great chance of a mistake), which is why picking the best Zone Zero is so important.
What Comes After Zone Zero
Once we find a Zone Zero, we can begin the hard work of collecting ZEM data. Here's how I'd imagine doing that.
Step #1 Matrix of 1-60
If we imagine Zone Zero has a level 1 and a level 3, but not a level 2 or 4, we'll start with a matrix that looks something like:
1: Zone Zero 2: ? 3: Zone Zero 4: ?
Our next goal will be to "fill in the blanks" using as few zones as possible. Ideally we could fill in 2, 4, and every other missing level with a single zone, but that probably won't be the case, and we'll have to use a few zones.
The key (at this stage) will be to try and find zones (likely failed zone zero contenders, but also probably a zone or two to cover the 40+ range), which both overlap Zone Zero with at least one level of mob.
Once we find zones to cover every level in the game, we can then start by calculating the ZEMs of those zones, relative to Zone Zero ... and once we have that we'll be able to calculate the ZEM of any zone in the game by killing just one mob in it, because every zone will have at least one "overlap" with our matrix.
Step #2: The Popular Zones
It'd be great to know the ZEM of Surefall Glade, but realistically not a lot of people earn XP there. However, lots of people kill in a zone like Karnor's Castle, so knowing it's ZEM is a lot more useful.
Since it makes sense to start by calculating the popular zones, Step #1 would only focus on the top (10? 20? 25?) zones.
We'd start by finding out how many of them "overlap" with Zone Zero, and then ask volunteers to kill mobs of that level in both zones.
For instance, Karnor's Castle has a level 50 Hangnail, while South Qeynos (if it were Zone Zero) has several level 50's (Corporal Lancot, Hanns Krieghor, Koalindl, etc.). We would ask two volunteers to take level 52 characters to Qeynos/Karnor's and kill Hangnail/Corporal Lancot 50 times, and report back how much experience each gave.
Once we do the math on those numbers we could definitively say what the ZEM of Karnor's Castle is! For the first time ever, the wiki would have an accurate "ZEM" for that zone.
We could then repeat the process for every popular zone that is "one degree of separation" away. For instance, a zone might not have a fixed level 50 mob (or any others that overlap with South Qeynos) ... but it might have a fixed level 44 ... and Karnor's (which we already calculated) does have a level 44 (Skeletal warlord). If we could find two volunteers to kill the 44's in those zones, we could calculate the second zone's ZEM, even though it doesn't overlap Zone Zero.
Race/Class Note
Although racial XP penalties have been removed at this point in the timeline., class penalties haven't.
This means that when volunteers report their XP results, we'll also need to collect their class, and then do a bit of math so that we're not comparing apples to oranges (ie. Warriors to Paladins).
Step #3: The Less Popular Zones
At this point we should have 50+ kills worth of data from a whole lot of mobs in Zone Zero, and also a bunch of mobs in other popular zones. Now it's time to figure out the "less popular" zones ... but many of these will only require killing a single mob (in the new zone), because we'll already have data for a mob of that level in a zone whose ZEM we know.
This means that even though we'll have a lot more zones to do in step #2, it should actually be easier in many ways.
Step #4: Normalization
If we stopped at step #2 we'd be in great shape, because we'd have an objective number to measure every zone in the game by. However, that number won't be perfect: it will be "relative to South Qeynos" (or whatever Zone Zero is) ... but what players really want is a number relative to the average zone.
Once we have every zone (or even just most of them) it will be easy enough to average their ZEMs, to find the average P99 ZEM. We can then do some math to convert "Zone Zero-relative" ZEMs into true (average-based) ZEMs.
After all that killing (and recording of XP), we can finally pat ourselves on the back, and relish the fact that the wiki now has accurate ZEMs. That is, until ....
What if Rogean and Nilbog Change the ZEM's?
There is a chance that if this project's success might cause Rogean and Nilbog to shake things up and change the game's ZEMs. This would not be a failure!
P99's maintainers have been extremely hesitant to change the ZEMs ... but (judging by Reddit and the Project 1999 Forums), the player base is overwhelmingly in favor of more frequent changes. After all, for us players a ZEM change means getting to adventure in fun new places, instead of being stuck in the same leveling treadmill (of a few popular zones).
But there's also more good news: even a ZEM change wouldn't ruin all of this project's work. Past ZEM changes have only impacted < 20 zones, and the staff even published a list of the zones affected. If a ZEM change happens, we wouldn't throw all our work out ... we'd just have to repeat 21 * 50 kills (ie. one mob in each of the twenty changed zones, plus one in Zone Zero, to figure out if it changed).
After that we'd be back to accurate wiki ZEM's, and the staff would be very unlikely to change the ZEMs again (or at least, not for awhile).