what the max single stat of a dino could be?

[FateRover]

A wild 150 levels rex as example. 100% tame effective ,then it will be  225 level,and then what the max attribute of the damage?

I ever have heard that the single attribute can not be higher than 23% of the whole attributes(so the single max attribute is 225*0.23=51.75≈52,and according to ark smart breeder, the damage can not be higher than 431.6)

except flyer for no wasted point

Is that true ?

[mackyish3r3]

it is possible for all 224 of the points to go into damage, but it is very unlikely. That would be 1443%.

[Algchaotic]

That’s not how that works it’s completely impossible to find a stat distribution like that not just unlikely. When taming something the highest you can see is around 32 points in a stat depending on if it’s a land/water tame or a flyer because flyers don’t get wasted points. If that was the case you’d see people with 100k health golem clones because of a stat anomaly.

[RenanLRME]

well statistically you can have but very unlikely

[GrumpyBear]

I don't think that's true. Most of the animals melee scales along that % range, but I have seen clean melee over 425% caught wild on 150 tames.  There was someone with 431.6% ab spinos along while back.  425% was just so prevelent by that point that it sort of disappeared once melee went past 500%.  I have a tribemate who caught 443% polar bears on valguero, I'm positive it's possible to go higher, it's just the odds of it are really small.

OH yea, I've had 54 points on stamina no mutations on an aberrant megalosaurus for years now on official pve.  That beats the 23% threshold at 24% right there.

[Luewen]

1 hour ago, Algchaotic said:

That’s not how that works it’s completely impossible to find a stat distribution like that not just unlikely. When taming something the highest you can see is around 32 points in a stat depending on if it’s a land/water tame or a flyer because flyers don’t get wasted points. If that was the case you’d see people with 100k health golem clones because of a stat anomaly.

Not quite true. I have seen 40 points in a stat multiple times.

[mackyish3r3]

2 hours ago, Algchaotic said:

That’s not how that works it’s completely impossible to find a stat distribution like that not just unlikely. When taming something the highest you can see is around 32 points in a stat depending on if it’s a land/water tame or a flyer because flyers don’t get wasted points. If that was the case you’d see people with 100k health golem clones because of a stat anomaly.

Not completely impossible. the chances are extremely low, but not zero.

[deedoubleu]

In my experience 50 is almost hard cap on any single stat. Only saw couple 51-53 after taming on special-case creatures(different stats or different point allocation) such as flyers and lightpets.

But I never saw any concrete info on how points are actually distributed. Like parts of the code from devkit or something like that, instead it's always just speculation based on personal experience.

[Slvr]

I once had a baryonyx with clean 51 points in melee.

[DomiDarko]

6 hours ago, Luewen said:

Not quite true. I have seen 40 points in a stat multiple times.

Right! For my breeding lines I don't use stats below 40 points! Highest I've had was 52 in melee on a Megalodon and I get stats between 45 and 50 quite often, especially on flyers.

[Luewen]

19 hours ago, deedoubleu said:

In my experience 50 is almost hard cap on any single stat. Only saw couple 51-53 after taming on special-case creatures(different stats or different point allocation) such as flyers and lightpets.

But I never saw any concrete info on how points are actually distributed. Like parts of the code from devkit or something like that, instead it's always just speculation based on personal experience.

Stat allocation on wild dinos and after tame points are random. Well as random as you can code wise make it. But for your normal user its random. There is no emphasis on specific stat when allocating points. In theory you could build a statistics on it but you wpuld need quite a huge sample before you can present it as a evidence. Like few hundred thousand tames.

[deedoubleu]

3 hours ago, Luewen said:

Stat allocation on wild dinos and after tame points are random. Well as random as you can code wise make it.

How do you know this?

As I see it - if it was honest random, then you would once in a while see say 0-5 points in one stat on wild 150 creature. Or 50 points in one stat on lvl 100. Don't know about you, but I've yet to encounter one.

That's indirect confirmation of points allocation being weighted in one way or the other.

[sjskdjkfa]

1 hour ago, deedoubleu said:

How do you know this?

As I see it - if it was honest random, then you would once in a while see say 0-5 points in one stat on wild 150 creature. Or 50 points in one stat on lvl 100. Don't know about you, but I've yet to encounter one.

That's indirect confirmation of points allocation being weighted in one way or the other.

Don't forget that randomness from computers is actually psuedo random. This has the effect of making unlikely events even less likely because computerized randomness is what a layman thinks of random. True random e.g. of shuffled cards you look at any random deck of cards and you are actually likely to see the unlikely sequence of cards, but a layman would want the randomness of no sequences, and if shuffled in code you are less likely to see it.

[yekrucifixion187]

Tamed 1560 Argy Stam (58), 1005 Argy Oxygen (57), 13,000 Argy Food (55), 1648 Quetzal Weight (53), and 8160 Quetzal Food (58) . Maybe because flyer's don't get wasted movement speed distributions but 50+ is if not common easier on flyers and should be sought to start breeding. 

[deedoubleu]

38 minutes ago, sjskdjkfa said:

Don't forget that randomness from computers is actually psuedo random. This has the effect of making unlikely events even less likely because computerized randomness is what a layman thinks of random. True random e.g. of shuffled cards you look at any random deck of cards and you are actually likely to see the unlikely sequence of cards, but a layman would want the randomness of no sequences, and if shuffled in code you are less likely to see it.

Indeed. What player perceives as most random outcome is usually not random at all. There would often be some correction to ensure outcome doesn't repeat more than X in a row, is at least Y, but never more than Z etc etc etc.

That's kinda my point. I don't believe point are being distributed randomly, as in "there is no additional corrections to ensure more random appearance or rules to avoid/guarantee specific result" .

[deedoubleu]

12 hours ago, Luewen said:

Stat allocation on wild dinos and after tame points are random. Well as random as you can code wise make it. But for your normal user its random. There is no emphasis on specific stat when allocating points. In theory you could build a statistics on it but you wpuld need quite a huge sample before you can present it as a evidence. Like few hundred thousand tames.

So I ran some tests. Wild lvl150 for example.

 

First I tried to distribute points as random as I could. Abnormal deviations for this method was <10 and >60 (I think it's fair to say "this poop never happens" for 60 points in single stat on wild dino) marked by < and > respectively. In 100 iterations there wasn't a single result without multiple abnormal values. 1 case with only 1 such stat.

Here is 10 iteration example. I assure you it doesn't change on 100000 iterations.

"H_25 	 | 	S_2 	 | 	O_61 	 | 	F_9 	 | 	W_6 	 | 	D_0 	 | 	M_47 	 | 	<><<<"
"H_0 	 | 	S_0 	 | 	O_2 	 | 	F_13 	 | 	W_135 	 | 	D_0 	 | 	M_0 	 | 	<<<><<"
"H_14 	 | 	S_2 	 | 	O_114 	 | 	F_19 	 | 	W_0 	 | 	D_0 	 | 	M_0 	 | 	<><<<"
"H_0 	 | 	S_105 	 | 	O_20 	 | 	F_1 	 | 	W_0 	 | 	D_0 	 | 	M_24 	 | 	<><<<"
"H_121 	 | 	S_1 	 | 	O_21 	 | 	F_0 	 | 	W_2 	 | 	D_2 	 | 	M_3 	 | 	><<<<<"
"H_30 	 | 	S_53 	 | 	O_25 	 | 	F_1 	 | 	W_15 	 | 	D_4 	 | 	M_6 	 | 	<<<"
"H_8 	 | 	S_1 	 | 	O_0 	 | 	F_129 	 | 	W_4 	 | 	D_0 	 | 	M_8 	 | 	<<<><<<"
"H_71 	 | 	S_15 	 | 	O_10 	 | 	F_3 	 | 	W_12 	 | 	D_24 	 | 	M_4 	 | 	><<<"
"H_0 	 | 	S_0 	 | 	O_1 	 | 	F_59 	 | 	W_12 	 | 	D_78 	 | 	M_0 	 | 	<<<><"
"H_1 	 | 	S_0 	 | 	O_118 	 | 	F_0 	 | 	W_14 	 | 	D_8 	 | 	M_9 	 | 	<<><<<"

 

Then I modified it to first distribute points in range 10 - 20 and then continue distributing remains in range 0 - 10 while there are any free points left.

Changed anomaly marker values to <10 and >35 and still only got 6 of those cases in 100 iterations. Now that looks much more like what we can observe in ark.

"H_17 	 | 	S_22 	 | 	O_23 	 | 	F_25 	 | 	W_22 	 | 	D_22 	 | 	M_19 	 | 	"
"H_20 	 | 	S_21 	 | 	O_14 	 | 	F_23 	 | 	W_23 	 | 	D_19 	 | 	M_30 	 | 	"
"H_21 	 | 	S_26 	 | 	O_20 	 | 	F_16 	 | 	W_24 	 | 	D_24 	 | 	M_19 	 | 	"
"H_26 	 | 	S_21 	 | 	O_22 	 | 	F_16 	 | 	W_26 	 | 	D_23 	 | 	M_16 	 | 	"
"H_31 	 | 	S_16 	 | 	O_28 	 | 	F_12 	 | 	W_22 	 | 	D_22 	 | 	M_19 	 | 	"
"H_20 	 | 	S_23 	 | 	O_20 	 | 	F_25 	 | 	W_17 	 | 	D_18 	 | 	M_27 	 | 	"
"H_16 	 | 	S_15 	 | 	O_24 	 | 	F_24 	 | 	W_27 	 | 	D_23 	 | 	M_21 	 | 	"
"H_34 	 | 	S_20 	 | 	O_18 	 | 	F_18 	 | 	W_27 	 | 	D_14 	 | 	M_19 	 | 	"
"H_24 	 | 	S_28 	 | 	O_23 	 | 	F_24 	 | 	W_21 	 | 	D_19 	 | 	M_11 	 | 	"
"H_18 	 | 	S_26 	 | 	O_15 	 | 	F_25 	 | 	W_26 	 | 	D_24 	 | 	M_16 	 | 	"

 

I'm not saying they do it exactly as I described, but they are doing something. For the matter of fact they are doing a lot. For example, my approach still technically allows  90 points in same stat, but makes it highly unlikely.

If there wasn't checks for min/max values we would not discuss 50-60 as something special, instead people would brag about getting 100+ points in stam.

[FateRover]

在2020/7/23 在 PM5点05分, Slvr 说：

I once had a baryonyx with clean 51 points in melee.

any higher ？

[Slvr]

12 hours ago, FateRover said:

any higher ？

I've personally never tamed anything with higher points than that in a single stat. I'm jealous of yall pulling in 52-55 points in things. Meanwhile I'm over here taming a 150 thyla that ended up with 45 points in o x y g e n. smh

[wizard03]

3 hours ago, Slvr said:

I've personally never tamed anything with higher points than that in a single stat. I'm jealous of yall pulling in 52-55 points in things. Meanwhile I'm over here taming a 150 thyla that ended up with 45 points in o x y g e n. smh

had something similar.....then did a cave run that had water in it.....was kinda glad during that run o.o

[Thyme]

A 180 perfect tame tek dino would be level 269, so would could have up to 268 points in one stat. The chance of this is:

0.00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000326% (3.263813e-227)

Which would then be deleted on any official server due to having a stat > 255.

[deedoubleu]

10 hours ago, Thyme said:

A 180 perfect tame tek dino would be level 269, so would could have up to 268 points in one stat. The chance of this is:

0.00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000326% (3.263813e-227)

Which would then be deleted on any official server due to having a stat > 255.

I saw this few times and wondering where this info came from, is there any dev post or tweet mentioning it?

[Campi]

Tamed a 150 Spino on sunday with 58 points in stamina. Some times you just have to get lucky on the point-draw

[FateRover]

在2020/7/27 在 AM5点44分, Thyme 说：

A 180 perfect tame tek dino would be level 269, so would could have up to 268 points in one stat. The chance of this is:

0.00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000326% (3.263813e-227)

Which would then be deleted on any official server due to having a stat > 255.

NOT exactly like that ，as the point  r allocated by psuedo random

[Luewen]

On 7/24/2020 at 8:11 PM, deedoubleu said:

So I ran some tests. Wild lvl150 for example.

 

First I tried to distribute points as random as I could. Abnormal deviations for this method was <10 and >60 (I think it's fair to say "this poop never happens" for 60 points in single stat on wild dino) marked by < and > respectively. In 100 iterations there wasn't a single result without multiple abnormal values. 1 case with only 1 such stat.

Here is 10 iteration example. I assure you it doesn't change on 100000 iterations.

"H_25 	 | 	S_2 	 | 	O_61 	 | 	F_9 	 | 	W_6 	 | 	D_0 	 | 	M_47 	 | 	<><<<"
"H_0 	 | 	S_0 	 | 	O_2 	 | 	F_13 	 | 	W_135 	 | 	D_0 	 | 	M_0 	 | 	<<<><<"
"H_14 	 | 	S_2 	 | 	O_114 	 | 	F_19 	 | 	W_0 	 | 	D_0 	 | 	M_0 	 | 	<><<<"
"H_0 	 | 	S_105 	 | 	O_20 	 | 	F_1 	 | 	W_0 	 | 	D_0 	 | 	M_24 	 | 	<><<<"
"H_121 	 | 	S_1 	 | 	O_21 	 | 	F_0 	 | 	W_2 	 | 	D_2 	 | 	M_3 	 | 	><<<<<"
"H_30 	 | 	S_53 	 | 	O_25 	 | 	F_1 	 | 	W_15 	 | 	D_4 	 | 	M_6 	 | 	<<<"
"H_8 	 | 	S_1 	 | 	O_0 	 | 	F_129 	 | 	W_4 	 | 	D_0 	 | 	M_8 	 | 	<<<><<<"
"H_71 	 | 	S_15 	 | 	O_10 	 | 	F_3 	 | 	W_12 	 | 	D_24 	 | 	M_4 	 | 	><<<"
"H_0 	 | 	S_0 	 | 	O_1 	 | 	F_59 	 | 	W_12 	 | 	D_78 	 | 	M_0 	 | 	<<<><"
"H_1 	 | 	S_0 	 | 	O_118 	 | 	F_0 	 | 	W_14 	 | 	D_8 	 | 	M_9 	 | 	<<><<<"

 

Then I modified it to first distribute points in range 10 - 20 and then continue distributing remains in range 0 - 10 while there are any free points left.

Changed anomaly marker values to <10 and >35 and still only got 6 of those cases in 100 iterations. Now that looks much more like what we can observe in ark.

"H_17 	 | 	S_22 	 | 	O_23 	 | 	F_25 	 | 	W_22 	 | 	D_22 	 | 	M_19 	 | 	"
"H_20 	 | 	S_21 	 | 	O_14 	 | 	F_23 	 | 	W_23 	 | 	D_19 	 | 	M_30 	 | 	"
"H_21 	 | 	S_26 	 | 	O_20 	 | 	F_16 	 | 	W_24 	 | 	D_24 	 | 	M_19 	 | 	"
"H_26 	 | 	S_21 	 | 	O_22 	 | 	F_16 	 | 	W_26 	 | 	D_23 	 | 	M_16 	 | 	"
"H_31 	 | 	S_16 	 | 	O_28 	 | 	F_12 	 | 	W_22 	 | 	D_22 	 | 	M_19 	 | 	"
"H_20 	 | 	S_23 	 | 	O_20 	 | 	F_25 	 | 	W_17 	 | 	D_18 	 | 	M_27 	 | 	"
"H_16 	 | 	S_15 	 | 	O_24 	 | 	F_24 	 | 	W_27 	 | 	D_23 	 | 	M_21 	 | 	"
"H_34 	 | 	S_20 	 | 	O_18 	 | 	F_18 	 | 	W_27 	 | 	D_14 	 | 	M_19 	 | 	"
"H_24 	 | 	S_28 	 | 	O_23 	 | 	F_24 	 | 	W_21 	 | 	D_19 	 | 	M_11 	 | 	"
"H_18 	 | 	S_26 	 | 	O_15 	 | 	F_25 	 | 	W_26 	 | 	D_24 	 | 	M_16 	 | 	"

 

I'm not saying they do it exactly as I described, but they are doing something. For the matter of fact they are doing a lot. For example, my approach still technically allows  90 points in same stat, but makes it highly unlikely.

If there wasn't checks for min/max values we would not discuss 50-60 as something special, instead people would brag about getting 100+ points in stam.

Codewise there is no limit how the points can be allocated. But your iterations look very abnormal with getting multiple 0 points on rolls. That is statistically veeeeeeeeery rare on just 10. So something is off there. 

[deedoubleu]

1 hour ago, Luewen said:

Codewise there is no limit how the points can be allocated. But your iterations look very abnormal with getting multiple 0 points on rolls. That is statistically veeeeeeeeery rare on just 10. So something is off there. 

I'm not claiming to be a good programmer, I'm not a programmer at all.
I don't know of an algorithm that can randomly distribute a value between several "boxes" at the same time and had to invent my own.

Though, what should be "normal" distribution for it? More even numbers? Why? If we talking about unbiased random then there is a significant probability of any single stat to "grab" a high value of 100 and above, and since amount of points in stats are not independent, that would mean that there will be very few in others. That correlates with my results - if few stats grabbed a lot, there is nothing to choose from for other stats.

Thus, in my opinion, it's equal stat distribution that looks artificial and unnatural for as-close-as-possible-to-random distribution.