Creature Handler Archive
Thread: For those who like numbers... the relationship between LEVEL and DEFENSE
This analysis is for WILD TAMES only. I did not include BE'ed, converted BE'ed, etc, pets, as there is no indication for sure yet whether they are working right. They do seem to follow the pattern however.
For those who are curious, I decided to use the program MatLab to figure out just exactly what the formula is to compute defense from level. Here is the data I collected, with each LEVEL matched to the DEFENSE for a critter at that level (critters of a given level always being identical).
[LEVEL, DEF]
ans =
65 157
58 143
23 74
10 38
32 84
52 118
66 157
22 71
20 67
I tried first to fit this with a linear regression but it didn't work too well... so then I tried something more sophisticated, a polynomial fit. Using just one coefficient plus a constant, it worke quite nicely:
>> P = polyfit(LEVEL, DEF, 1)
P =
2.0164 23.0343
Which means that defense should equal level * 2.0164 + 23.0343 (probably they did a roundoff like level * 2 + 25).
I checked this by taking the level and plugging it into the fitted polynomial, as follows:
>> EST_DEF = LEVEL * P(1) + P(2)
EST_DEF =
154.0973
139.9828
69.4105
43.1979
87.5576
127.8847
156.1137
67.3941
63.3614
>> [LEVEL, DEF, EST_DEF]
ans =
65.0000 157.0000 154.0973
58.0000 143.0000 139.9828
23.0000 74.0000 69.4105
10.0000 38.0000 43.1979
32.0000 84.0000 87.5576
52.0000 118.0000 127.8847
66.0000 157.0000 156.1137
22.0000 71.0000 67.3941
20.0000 67.0000 63.3614
Checking it, this seems to work very nicely. Level is the first column, actual defense the second, predicted defense the third. As you can see, the predictions are pretty darn close.
So, if you want to know how much defense an animal of ANY level has, use the following formula (approximate):
Defense = Level * 2 + 25
OK, I think I got too much geek on my hands. Gotta go wash it off.
C
Good work. I'm impressed.
(I usually am when I don't have a clue what someone's talking about) ![]()
But nevertheless, your work is appreciated.
Message Edited by Gandolforf on 05-27-2005 08:00 PM
The bottom line is that, roughly, Defense = (2 * level) + 25. That's all you need to get out of all the math.
C
Termis wrote:
Igot bored again, lol.
These were calculated from the tests I did before...
again, these tests are all from the same lair against the same critters, cl59 desert gurreck chargers.
how I figured this.
[(dealtdmg - mindmg) / (maxdmg - mindmg)] x 100 =% of dmg done
cl70kinadnitch = 225 out of 343 (100-390 listed in datapad) = 83.8% of total dmg capable
cl69 mr flyer (testA2)= 223 out of 339 (90-390 in datapad) = 83%
cl50 rancor = 179 out of 272 (80-350 in datapad) = 71.1%
cl23gs panther = 122 out of 186 (50-260 in datapad) = 64.76%
cl22 f narglatch = 148 out of 226 (60-320 in datapad) = 63.85%
One thing that I am theorizing from these tests is that had I a cl59 pet the dmg done would be 75%.
So, the dmg done to an even-whitecon from your pets perspective, is 75% of it's dmg range.
Curious as to what happened with that funky test with the flyer,I usedthe funkytest and it came out...
cl69mrflyer (testA)= 157 out of 240 (90-390 in datapad) = 50%
It's dmg (240 base, 157 is after resists) was exactly it's median, weird, no other tests were like this which shows it was a funky bug. This may be happening to us at times causing our pets to bemuch weaker than they should be. Which makes me wonder...
(mindmg + maxdmg) / 2 = minimum possible dealtdmg before resists?
This showspet clis affecting dmg done within range, but not by very much.
Keep in mind, the cl70 wasnot doing 20% more dmg than the cl22, it was merely doing 20% more of it's own dmg potential within it's own dmg range.
What this also shows is if they did give us cl80 pets they wouldnot be allsome hope for. They would only dmg ~5% more within thier own dmg range, which would be about 10-20dmgxtra for a wild pet or possibly10-40dmg xtrafor a good BE pet, possibly taking that much less dmg IF it works both ways. That would be a nice little improvement but it's not going to allow our pets to tank or be dmg dealers. Stats make a huge difference, cl doesn't.
[*note*] Don't use these percentages as exact to figure out calculations per clbecause in my datapad some pets are listed with min-max (min-max) though they are full grown. I'm unsure the numbers can be trusted, even if off bye only a few dmg points. [*note*]
post CU and pre publish 17 the damage dealt by pets was governed only by to hit, min & max damage as explained in this thread
post publish 17 I don't know what the hells going on, defence might be taken into account now, but it looks like CH is bugged and/or pets & creatures are getting damage multipliers with level difference (including group level)so it is hard to see whats really going on.
As an example my wild gurreck (level 50, 107 to hit, 107 defence, 80-340 damage) ungrouped against an arachne warrior (level 63, 147 to hit, 147 defence, 90-360 damage)
Gurreck: did 115 total damage to the warrior per hit
Arachne Warrior: did 861 total damage to the gurreck per hit!
after watching the gurreck get incapped in seconds I thought maybe I should group with it... then this happened
Gurreck: 340 total damage to the warrior per hit
Arachne Warrior: 114 total damage to the gurreck per hit
so it looks like theres definately some level difference multipliers going on here, and when grouped the pets become your level, but it also seems pet damage is gated so it can never be more than the max damage.
C
Chessack wrote:
I have not checked damage. My character is level 80, but I don't quite see how that affects the numbers I posted, as those are the stats in the datapad for my wild tames. It seems to me my wild level 65 is going to have 157 defense whether I'm level 60 or level 80... isn't it?
C
I think for every level your pet is (or you are if your group) over or under its targets level then get a bonus and minus respectively to its defense. Just like the to hit.