New revision of Envenom spell
It seems you have updated the data on the page for the Envenom spell.
But I think you are mislead. Both Level and Magic Level are taken into account.
You would've seen this if you'd tested it thoroughly, and by thoroughly I mean that you should've taken different characters with same Magic Level and different Level ranges, and not just one character (?) (as mentioned in the page).
And you can only do such precise data collection in Testservers - you can only get precise data by testing in Testservers. In which you would be able to test with different level disparities (200, 100, 50, 300, etc.), and more importantly, with Magic Level 0* (1, 2, etc.) which would make you realize the effect of the character's level in the spell damage.
I've had resources and time to confirm in this in this last Testserver in which I was invited to.
Right now I don't have the data log anymore as I have concluded it at the time I had edited that page and others from utori-like spells, but I can assure you that it does considers the player level. —The preceding unsigned comment was added by RaymondTFR (talk • contribs). Remember to sign your comments!
I changed the description for the spell, because I thought that the provided information was a bit misleading. It looked like the spell would do only minor damage and is not worth to learn or use.
About the methodology of my tests:
I used only one char but at different levels and magic levels. So I started at level 336 and stopped at 376. The magic level ranged from 97 to 112 and I used different equipment to get the same magic level at different levels and vice versa.
Then I casted the spell to Hellfire Fighters, as they dont heal and are neutral to earth. I recorded the damage taken from the start until it vanished and I was able to fit a logarithmic function (87.4 e^-0.05t). That means, the total damage and the time it lasts do only depend on the start value, its not a fixed number of turns.
To determine the start value, I casted the spell about 40 times for each level/magic level combination to get resonable statistics. At one level/magic level combination I used a much higher statistics and found, that the distribution in a range of about 16 points is flat (a gaussian distribution would be much more "realistic"). The mean damage value is given below:
There is a clear dependency for the magic level (0.65 dmg/mlevel) and no significant dependency for level (0.027 dmg/mlevel).
The data do not allow to rule out a dependency for level as the tested range is limited, so using a druid at much lower level might be helpful to clarify this. Thus I wrote "seems only takes the magic level into account"
Gundli (talk) 07:43, October 7, 2016 (UTC)
Hello Gungli, glad to hear your insight.
Yes, the information was messed up. Beni helped me out a bit to put the description in a clearer way, which he did, but he messed up the part where I mentioned the fact that the damage values from 19 to 1 may occur more than one time (in a fixed amount of times...).
I should've fixed that, but I'd forgotten about it.
Your data is good, although I really don't think the start value (are you referring the maximum damage by that?) or the med dmg aren't of much use in this kind of research, since you need to make sure to confirm the minimum and maximum values possible, with at least two characters and no margin of errors so that only then you can think of a formula.
Anyway, I'd to use the previous description for the spell I did and combine it with yours as a complement.
The information I wrote about it I had confirmed in this last testserver, so you can have sure the level is taken into account (although a very small part of it).
The fact that the damage duration depends only the start damage is good, but not so true. It also depends (considering the target is neutral) on the decrement range from one damage to another (which I've confirmed to be a random value between the ranges [d/20, d/20+1] for all d values greater than 21 and 1 for those less than or equal to 21).
As a result of that, your logarithmic function woud fail since you cannot infer exactly how the damage is going to be decremented and how many occurrences of damages values (from 19 to 1) are going to happen, but you could infer the minimum and maximum of it (minimum occurrences of damages values [19, 1], and minimum decrement range(which depends on the damage value))...
I have confirmed all utori-like spells (including Soulfire Rune, which in this case I haven't edited its page with my research, as it already included useful information, but it is exactly as the Ignite spell) to be influenced by the character Level, in means of damage (curse, inflict wound, envenom) or total occurrences (holy flash, ignite, electrify, soulfire) of it.
Anyway, I suggest you to use the previous description and add yours as a complement, and if you are still willing to take these researches to the next level, then you probably could talk to someone here (admins) in order to have the opportunity/chance of being invited earlier to the upcoming testservers and re-confirm my data.
Unfortunately I don't remember one exact data log for Envenom but I do for spell Inflict Wound (to clarify the Level part).
Example (1): at level 40 (the level required to use it) and a skill of 10 (to be sure, all melee (and the rest) skills were 10):
- the exact minimum value possible you could deal would be 4.
- the exact maximum value possible you could deal would be 6.
Example (2): at level 100 and a skill of 10 (to be sure, all melee (and the rest) skills were 10):
- the exact minimum value possible you could deal would be 5.
- the exact maximum value possible you could deal would be 7.
Hi Raymond, I just realise, that I misunderstood your (or Benis) sentence: "If the target is neutral to earth damage, then the damage ranges from 1 to 19 and may occur more than one time". That is true, but your statement that one can not describe it with a logarithmic function is not. The function (A e^-0.05t) with A=start(maximum) damage and t=time steps=seconds/4 describes the decreas not perfect but very well. Even the decrease below 20 is not random but just the logaritmic decrease rounded to full numbers. You can see that if you compare the calculated with the recorded damage for A=94. You can see that the formular is a bit off but describes the shape even in the lowest tail very nicely.
About the other question, the dependency on the level, I have nothing against quoting "This spell takes into account the level and magic level" as you have better data for a large level range and even in my data I can't exclude that there is a small dependency.
Following your sugesstion I will try to get invited earlier to the next test server, as I tested many things there in the past and had never enough time.
Gundli (talk) 16:22, October 7, 2016 (UTC)
Well, if I am not right about the fact that the occurrences of damages from 19 to 1 are constant independently of the start damages then it could probably mean that it indeed can be described through your function, and you're probably right.
My test ranges ranged from the minimum level to cast the spell to 265, and magic levels ranging from 0 to 70, as I do not have characters above level 170 (npc assistant gives you exp 'til 264~ level).
I saw a video of an ED (level 479, magic level 119) that showed me something new (tibiacast is gone now :s), and it all makes sense now with your function.
I just wish I still had the data log from my tests so I could upload it here for you, but unfortunately I tried a few days ago (because I'd discovered something about the way the damages decrements itself) and I couldn't find it even with recover file programs (as it been a long time now and such sectors of the hdd were already destroyed).
I hope you get invited in the next Testserver and bring us with even more detailed data and descriptions (I chose not to write such infos in the spell pages, since it was a bit of overkill, haha).
If I also happen to be invited in the next upcoming testservers then we could met up there and help each other. Although I wound be going to research new creatures at the extend of their resistances, armor, etc.
I noticed your hard work on testing spells and runes formulae. Would you like to be invited for the next private test server as a TibiaWiki representative so you can make further tests?
Hi Hunter of Dragons,
would be great to get invited! I have many other ideas to test and I am already collecting items on real Tibia for that.
Gundli (talk) 06:50, October 11, 2016 (UTC)
one is to get more statistics of umbral creation (I did 17 masters already), that of course depents on the rosh status on the testserver. Ice sculpturing would also be interesting as CIP claims, that the chance is 10 times higher at the event. It looks more like 5 times higher but that might be due to low statistics without the event bonus. As everybody will wait for the event, it's probably not so interesting anymore.
Gundli (talk) 07:06, October 11, 2016 (UTC)
Gundli, I need to know your character name in order to get you invited. If you don't want to reveal it here, we can find private ways to get in touch with each other.
I'm going to request CipSoft to add gold tokens to Testserver Assistant's trade list, so we can gather info on the missing weapons. I could very well ask the same for dream matter and clusters of solace, would you like that?
Are you Gundli?