Menu
What's New
By:
Filters

Pet Math - Level vs. Experience

Expensive is fine, if it adds value, for example if I can get my pet daspletor stalker to level 15 and it can do some extra special stuff I will pay up to do it. If just for vanity sake, clearly waste of time and resources.
 
my bristle hit lvl 9 with 1776 xp
bristle-lvl9.jpg
 
Nice, cute little critter:)

What would be cool is a taming HOF board for when all mobs are tamable, maybe taming missions as well.
 
Kombinator said:
For daiki experience level will be much higher, mine is just about level 7 and is at about 6K+ experience.
Click to expand...
Would you mind starting a thread to track it? I'd hate to see good data vanish, and having a separate thread to specifically track Daikiba (maybe separate for <whatever>, Strong?) could become quite useful. F.ex. it's an absolute requirement to be able to define the same formula for pets other than the Bristlehog. Maybe it's as simple as just a scaler, or maybe it's something more elaborate.

Things that would be required would be mob (obviously, e.g. Daikiba Strong), and then for each sample:
  • Level
  • XP
  • Experience
(Level and XP could be combined like "2.04", "3.11", "6.1" - just like Windows versions over time :) ).

Additionally, decay per tick could be good to see checked for different pets (where tick is the 30-second period mentioned previously), and also their Experience increase over regen-time. I think 12 samples for Daik regen could suffice (just to verify no tick has been lost) over a 6-minute (spawned) period, for starters.

Also, max tt of different mobs could be of interest to see.
 
tamlin said:
Would you mind starting a thread to track it? I'd hate to see good data vanish, and having a separate thread to specifically track Daikiba (maybe separate for <whatever>, Strong?) could become quite useful. F.ex. it's an absolute requirement to be able to define the same formula for pets other than the Bristlehog. Maybe it's as simple as just a scaler, or maybe it's something more elaborate.

Things that would be required would be mob (obviously, e.g. Daikiba Strong), and then for each sample:
  • Level
  • XP
  • Experience
(Level and XP could be combined like "2.04", "3.11", "6.1" - just like Windows versions over time :) ).

Additionally, decay per tick could be good to see checked for different pets (where tick is the 30-second period mentioned previously), and also their Experience increase over regen-time. I think 12 samples for Daik regen could suffice (just to verify no tick has been lost) over a 6-minute (spawned) period, for starters.

Also, max tt of different mobs could be of interest to see.
Click to expand...

I would love to, should have done it initially, I am a big data guy myself:)

At this point it may be a bit late to track early levels, I can start noting it from Level 7 on for the daiki strong, maybe someone else can track from level 1.

Tamlin, would you be up for it, I will happily donate a daiki to you if you wil create the statistics from Level 1.
 
OK, interesting, guess lets see how high or deep this level hole goes:)
 
Kombinator said:
I would love to, should have done it initially, I am a big data guy myself:)

At this point it may be a bit late to track early levels, I can start noting it from Level 7 on for the daiki strong, maybe someone else can track from level 1.

Tamlin, would you be up for it, I will happily donate a daiki to you if you wil create the statistics from Level 1.
Click to expand...

The other thing that we've yet to test is the affect of brushing. affection and mood level affects xp per trick so it could be cheaper to brush the pet to 70% affection first, but neither of us have a brush.

Also if tamlin doesn't take you up on that offer with the daiki, I'd be very willing to, although I imagine tamlin will want to. I've mostly been supplementing his data and confirming rather than providing fresh insights.
 
Sure, if tamlin doesn't you are welcome to. And I have a brush if you want to have the pets brushed to whatever affection level let me know as well.
 
For daiki: level 7.1, experiece 7,145, so perhaps these are linear.

Will try to report on more points later.
 
Kombinator said:
Sure, if tamlin doesn't you are welcome to. And I have a brush if you want to have the pets brushed to whatever affection level let me know as well.
Click to expand...

Cool, have sent you a friend request in game to more easily figure out the last few bits of information :)
 
Will be sure to accept when I am back in game, love to see the beginning of the taming revival and enthusiastic players, the dream is still alive!:)
 
Kombinator said:
Will be sure to accept when I am back in game, love to see the beginning of the taming revival and enthusiastic players, the dream is still alive!:)
Click to expand...

I'm very enthusiastic. I'd just decided I wanted to tame when MA announced the cryengine update and that taming wouldn't be back right away. I've waited a long time to start skilling up in this.
 
Good, perhaps taming will rise again:) I am keen to use my skills to their full potential, want to see some meaningful ways to apply them.
 
Level 7 ancient exarosaur is 3000 exp

5 tricks to level 5 so far
Greet dance bow roll sit.

I would venture to guess that the mob stats hold the key to the formula.
(there were no stat changes from a level 1 to level 2 bristle, i dont think they change but might)

agi int psych str stam
Ancient exo: 19 2 2 39 14
ancient daik: 18 4 2 45 18
bristle: 14 4 5 2 10
ancint snable female strong: 21 15 10 35 31
 
Last edited:
If attributes are key, then this can be solved via a series of simple equation.

We have 5 variables to test, and also 3 different creatures (or techinically 6 if we count strong and regular varieties).

We need results of experience values at 0.1 incrementrs for each crature, ideally multiple creatures per set to verify uniformity.

We got that we can test the effect of attributes, as well as create a formula that is either general across the 3 known pets or mob specific.
 
These are for the hog

12 at 3200
13 at 3800
14 at 4440?
15 at 5100
16 at 5850
 
 
firejuggler said:
no new trick?
Click to expand...
Nothing new..



also in reply to the strong pic, it seems bigger mobs prolly gain more exp per trick and over time.. They also consume more for what im told.
 
Dennis blighter said:
it seems bigger mobs prolly gain more exp per trick and over time.. They also consume more for what im told.
Click to expand...
Correct. I'm currently performing tests on a Daikiba Ancient Strong. Current data suggests the formula for it is:
Experience = int_lvl^2 * 160 + int_lvl * 200

This so far only correlates with the 17 samples I have available - 12 samples at level 1, 5 samples at levels 7, 8 - why it possibly should be taken with a pinch of salt. That said, I personally have strong confidence these factors will end up proving correct. It also suggests the second multiplier (the one for the non-squared int_lvl) could be fixed at 1.25 times the factor used for the squared int_lvl.

Worth noting when trying to find correlation between the scale factors; max_tt of Bristlehog Puny is 4 PED, and max tt of the Strong Daikiba is 16 PED.

Furthermore, data suggests:
  • "pet decay" (tt of nutrio consumed) when idle is exactly max_tt/40 per hour.
  • Experience increase when idle is, as perhaps should be expected, a function of tt decay. While the function currently seems to be linear, samples from pets with max_tt other than 4 and 16 PED is required for verification.

Appropriate first posts in one or more of the "Pet Math" threads will be updated when data is verified.
 
Dennis blighter said:
These are for the hog

12 at 3200
13 at 3800
14 at 4440?
15 at 5100
16 at 5850
Click to expand...
Can you verify these claims? Some of them do not correlate with the otherwise seemingly correct formula discovered.

F.ex. the claim that Level 12 occurred at 3200 Experience? It should have happened no earlier than 3234. 13 @ 3816. 15 @ 5124.

While I appreciate the effort, at this point I believe only any observed deviations from expected Experience at given level, something that could potentially falsify the formula, would be useful.
 
tamlin said:
Can you verify these claims? Some of them do not correlate with the otherwise seemingly correct formula discovered.

F.ex. the claim that Level 12 occurred at 3200 Experience? It should have happened no earlier than 3234. 13 @ 3816. 15 @ 5124.

While I appreciate the effort, at this point I believe only any observed deviations from expected Experience at given level, something that could potentially falsify the formula, would be useful.
Click to expand...


Hmm, these recorded numbers of mine may be wrong as i rounded down the numbers to make sense, as most things are nice and even.(sometimes i was afk and gained a few percent before i could catch the 0%, so i rounded down) Once i gain the next lvl, ill track the exact exp.
 
Dennis blighter said:
Hmm, these recorded numbers of mine may be wrong as i rounded down the numbers to make sense
Click to expand...
OK. I kinda suspected something like that.

Anyway, first post updated with (preliminary) Daikiba Ancient Strong formula.
 
Alright,

lvl 17 pet(hog) at 6624 exp.

this one is exact.
 
tamlin said:
OK. I kinda suspected something like that.

Anyway, first post updated with (preliminary) Daikiba Ancient Strong formula.
Click to expand...

The rations between first and second scalar seem to be exactly 1.25 for both bristle and daiki, may help in further generalization of the fomula.
 
Kombinator said:
The rations between first and second scalar seem to be exactly 1.25 for both bristle and daiki, may help in further generalization of the fomula.
Click to expand...
Yup. Already noted. I think the internal formula is more like:
(int_lvl^2 * factor1 + int_lvl) * factor2,
and I have a suspicion it's only using integer math internally for it; the large variations in both cost and presented data for the smaller pet suggests this could be the case.
I've just keep the formula in its current form to help simplify both visualization and solving it, for the ones so inclined to go from Experience to Level.

What is needed now is data on some other pets; Daikiba non-strong (if they exist), snable(s) and exa. If you, dear reader, have such a pet it would be appreciated if you could provide
  • pet type name (e.g. "Snablesnot <something>")
  • max tt
  • Level, in the form x.yy (x=level, yy=XP to next level), and Experience.

Should you also measure the cost of doing one trick, report if it deviates from the current formula for cost-of-trick.
 
my pets just eat a lot nutrios and no dung coming out ? some must be weird :D
 
trance said:
my pets just eat a lot nutrios and no dung coming out ? some must be weird :D
Click to expand...
Constipation. I have encountered many creatures on Calyso with this defect, making them irritable and agressive. Early research suggested it was due to the time spent designing the genitalia they never got a working rectal area.

Sometimes, if you walk really slow, and the moon is just in the right phase, small amounts of by-products can - as if by magic - transmutate from a creature onto the ground and you can spot it. This transmutation has no correlation to known physics, and may be the result of some strange research performed by early colonist Professor Scatfinger, but data referring to early settlement history is scarce.
 
tamlin said:
Yup. Already noted. I think the internal formula is more like:
(int_lvl^2 * factor1 + int_lvl) * factor2,
and I have a suspicion it's only using integer math internally for it; the large variations in both cost and presented data for the smaller pet suggests this could be the case.
I've just keep the formula in its current form to help simplify both visualization and solving it, for the ones so inclined to go from Experience to Level.

What is needed now is data on some other pets; Daikiba non-strong (if they exist), snable(s) and exa. If you, dear reader, have such a pet it would be appreciated if you could provide
  • pet type name (e.g. "Snablesnot <something>")
  • max tt
  • Level, in the form x.yy (x=level, yy=XP to next level), and Experience.

Should you also measure the cost of doing one trick, report if it deviates from the current formula for cost-of-trick.
Click to expand...

If you want to do the data collection study on any of the above let me know, happy to provide the pets for study.
 
You must log in or register to reply here.
Back
Top