ZEM chart wiki question for Loramin

[Tecmos Deception]

Posting just incase anyone else feels the interest in undertaking something similar.

My wiz is in Najena, and I was annoyed that the ZEM "felt" lower than what is listed on the wiki. So I took note of my yellow bar position, noted the number of kills I had and their levels, used the exp formulas from the wiki, and determined that Najena actually has a ZEM of 100 (give or take), not the 130 listed.

(I was slightly more than 1 yellow from level 25 as a gnome wizard, I called it 22% of my level left. I killed 9 spiders of levels 20-22 (I got hit while sitting to verify levels of several spidies) to ding 25 and entered 25 with no visible yellow bar. Given the exp required to get through level 24 from https://wiki.project1999.com/Experience_table * 1.1 for wizard exp mod * .22 for the amount of 24 I earned from 9 kills / 9 said kills / 21 / 21 for the average-ish level of mobs (see https://wiki.project1999.com/Experience for exp awarded for kills formula) = 100ish ZEM)

I'll probably keep track of my progress in 25 for a bit too to make sure it wasn't a fluke, but yeah, even if my numbers were a bit off for the % I needed based on eyeballing my yellow bar or for the precise levels of the mobs I fought, there's no way the ZEM could be 130.



ANYWAY. Loramin! Do I just go ahead and edit the chart at https://wiki.project1999.com/Recommended_Levels_and_ZEM_List, or should I stick a note in there somewhere about "observed ZEMs on green99 as of _____ date" or something?

[loramin]

Wow, I definitely have to break the gifs out for this: this is exactly how we fix the wiki ZEMS!

https://i.imgur.com/VF1E5hB.gif
https://i.imgur.com/ZAiyHJu.gif
https://i.imgur.com/dhMeAzK.gif

Now, to be fair, I'm not sure if your data is enough to give us statistical certainty (any stats buffs willing to weigh in?) ... but even if your number is a little off, it's bound to be way more accurate than what we have (which was based on old unclassic EQEmu numbers, from what I understand).

In any case, you rock! And yes, please do edit the wiki. I'd just add either your research details, or maybe just a link to this thread (it's shorter), and I don't think you even need to add it to the page itself: if you just add it in your edit message it will be there on the history tab, which should be good enough if anyone ever questions where it came from.

Also, each zone (or at least most) has their ZEM listed on their zone page, so feel free to update the Najena page itself too.

One last note though: I know Dolalain submitted a bug with evidence for more classic ZEMs. It sounds like you didn't go out of your way to get these numbers, but still you might want to hold off on further research until the fruit of his research is patched in (because who knows: Najena or any other zone, might change when that happens?)

[Tecmos Deception]

I won't edit based off 9 kills. But by the time I finish a level here, I'll be confident enough that I know the real ZEM to change it. I really don't think it's actually 130. I don't think it's as low as 100 anymore though.

[loramin]

I won't edit based off 9 kills. But by the time I finish a level here, I'll be confident enough that I know the real ZEM to change it. I really don't think it's actually 130. I don't think it's as low as 100 anymore though.

I never took stats (and laughed at my numerous Psych friends that had to), so while I understand the basic ideas, I don't know the math to tell you "we need ____ kills to be certain". Psych is a popular major though (and other majors require Statistics too of course), so surely we have someone here who can chime in with a number that has the correct "R coefficient" (or whatever)?

But again, let's not let "perfect be the enemy of the good". I think waiting until you have as much data as possible is smart, but also ... if the actual ZEM is 110, and the wiki has 130 ... but you change it to 105 or 115 ... we're still better with your change than without it.

[Glasken]

I have a paladin that has leveled exclusively on the three skeletons and spiders in Najena from 24-27. I also have logs if you would like them.

[Dolalin]

Classic zem for Najena was 80, so this is pretty interesting.

https://www.project1999.com/forums/showpost.php?p=3042352&postcount=12

[BlackBellamy]

Lower Guk zem on that page is definitely wrong.

[cd288]

Could it also be that our data for things like the experience table and how much the mobs should be worth in EXP are off rather than the ZEM?

[Dolalin]

It might be the case that mobs are worth more raw xp here on P99. No way to tell without mind-controlling someone with database access.

Is anyone a telepath?

[cd288]

It might be the case that mobs are worth more raw xp here on P99. No way to tell without mind-controlling someone with database access.

Is anyone a telepath?

Or I guess less in this case since TD is saying the EXP is slower

[Tecmos Deception]

It might be the case that mobs are worth more raw xp here on P99. No way to tell without mind-controlling someone with database access.

Is anyone a telepath?

If this were the case, the exp formulas on the wiki wouldn't ever make sense unless ZEMs were all wrong in precisely the right way to make it look like those formulas are actually right. Which seems improbable or impossible to me.


Regarding stats, Lor? I think this is more an issue of a larger sample having less room for user error. Judging the % of my level I'm at by looking at the exp bar leaves room for error, both on the number of pixels there displaying it and on my eye.

If I count kills from ding to ding, then being 1% off in my estimation of my yellow bar is less of a factor than if I kill mobs for one yellow.

[zaneosak]

Was this experiment on white cons? It's really hard to get it perfect unless it's white cons or you 100% know the level of every mob (or at least very close to every mob) Interesting stuff.

I did calculate on Bloodgills (which are only a range of 3 levels) using their median level an entire level on my necro on blue and found the ZEM to be 100% accurate , within 4 kills actually, which can be accounted for in level variance.

It's always good to see these kinds of calculations.

[Tecmos Deception]

Was this experiment on white cons? It's really hard to get it perfect unless it's white cons or you 100% know the level of every mob (or at least very close to every mob) Interesting stuff.

I did calculate on Bloodgills (which are only a range of 3 levels) using their median level an entire level on my necro on blue and found the ZEM to be 100% accurate , within 4 kills actually, which can be accounted for in level variance.

It's always good to see these kinds of calculations.

Like I said, I verified level range on the spiders by their max hits.

[loramin]

Regarding stats, Lor? I think this is more an issue of a larger sample having less room for user error. Judging the % of my level I'm at by looking at the exp bar leaves room for error, both on the number of pixels there displaying it and on my eye.

If I count kills from ding to ding, then being 1% off in my estimation of my yellow bar is less of a factor than if I kill mobs for one yellow.

Absolutely. I just meant ... there's a mathematical way you can you be reasonably certain that ___ yellow bubbles (or whatever) is enough. You may well have passed that threshold, and I wasn't trying to say that you didn't. I was just saying "I know that a threshold exists, but I don't know what it is."

Was this experiment on white cons? It's really hard to get it perfect unless it's white cons or you 100% know the level of every mob (or at least very close to every mob) Interesting stuff.

I did calculate on Bloodgills (which are only a range of 3 levels) using their median level an entire level on my necro on blue and found the ZEM to be 100% accurate , within 4 kills actually, which can be accounted for in level variance.

It's always good to see these kinds of calculations.

This raises an excellent point: even though you confirmed a wiki number, the fact that you confirmed it is useful info on its own. It makes the LOIO number "better" than most other ZEMs, as it's "confirmed".

It seems to me it'd be handy if people could indicate such confirmation in the wiki somehow. The simplest way I can think of would be to edit that table at the bottom of: http://wiki.project1999.com/Recommended_Levels_and_ZEM_List, to fill in the "P99" column (and then when you do you could include your "proof" in the edit message?)

[tommydgun]

As a stats / math person I can safely say this is not a statistical problem. This is a math problem. If you are certain of the formulas for exp to level and exp per mob then you dont need to worry about sample sizes that's a simple math equation.

[astuce999]

As a stats / math person I can safely say this is not a statistical problem. This is a math problem. If you are certain of the formulas for exp to level and exp per mob then you dont need to worry about sample sizes that's a simple math equation.

It would be a math problem if we knew all that. But I think the whole point is that we don't know the exp per mob as a number, only as an increment of yellow in a bubble. So he's thinking of needing to go to a whole level to hear "ding" to be able to figure out exactly what the exp per mob is as a number, to then be able to figure out the ZEM, and Loramin was wondering if there was a way to determine before that if the sample size was enough to be "good enough".

Let's say he did 3 yellows, and the first yellow took 19 mobs, the second yellow 20 mobs, and the third yellow 21 mobs. How confidently would he be able to claim that the number of mobs needed to level is 100 so that he could then determine the ZEM?

cheers,

Astuce

[ewjax]

Just as FYI, if you want more accuracy in your exp bar percentages - turns out if you look at your exp bar at the trainer, the blue bars still show. 1x yellow bubble = 20%, 1x blue bubble = 4%.

Might help your data precision a bit.

[Tecmos Deception]

Well, I take back what I said earlier. I think Najena really is 130 ZEM.

It took 41 spider kills to go through all of level 26 as a wizard. The spiders range in level from 18 to 22, presumably evenly distributed. 18s were green at 26, but I don't think the 19s were. The math works out to right about ZEM 130 if I use level 20 as the level for all 41 kills in the formulas.

I don't think that some were green and potentially giving reduced exp is enough to throw the numbers off since the greens were the 1/5 (or so) of spiders giving the smallest amount of exp anyway, and the higher levels don't give only proportionally more because the exp formula squares level.