A small dive into the statistics of dinosaur taming vs dinosaur breeding.

[LucLon]

Some years ago, I had an argument with a tribemate:

While searching The Center for our next victim of force-fed mutton in hope of beginning a new breeding line in a freshly wiped server, he made a comment on breeding strategy. Specifically, he mentioned that 'most' official tribes do not begin breeding for mutations until they tame a Giga with 50 points in its melee stat. I disagreed, noting that breeding a Giga with 40 points in its melee stat for mutations was likely to produce 5 new mutations faster than finding one with such a high melee stat. Logically, he brushed me off, quoting some ancient wisdom bestowed on him by some former tribemate that might or might not have been in some official tribe at some time. I did not let this go.

At the time, I had to produce a Mathematics investigation on a topic of my choosing. As a result of my petty disposition, I chose to shoot myself in the foot with a topic well beyond the syllabus I was following. Particularly, I explored the compared efficiency of taming or breeding dinosaurs at a given cut-off stage of points in a desired stat. My goal was to find what I dubbed the 'switching point', the number of points in a desired stat at which it becomes more time efficient to breed for mutations than to tame a new dinosaur in hopes of it having more points in that stat.

This is that investigation.

Maths Investigation..pdf

I had to compress it a lot to upload on here because of the 400kb limit, so some parts aren't particularly legible.

There are some assumptions made throughout the investigation to stay on subject. Many of these are barely relevant to its applications, some should be accounted for when using the formula produced:

Particularly, t-tame is not the actual time to time a dinosaur, but rather the average time it takes you, specifically, to find, tranq, and tame a dinosaur (you should take into account offline time, since raising dinosaurs can be done when offline). E-breed does not account for breeding several dinosaurs at once (as is common practice), since, by doing this, you essentially get one useful mutation every time you hatch a set of eggs, you can replace t-breed by the raising time of your dinosaur in the final formula (you can also just track the average number of useful mutations (level increase, not actual mutations) you get in an hour over a couple of days and replace all of E-breed by this number). I will expand on these and other assumptions if there's any interest.

[SunsetErosion]

10 hours ago, LucLon said:

Some years ago, I had an argument with a tribemate:

While searching The Center for our next victim of force-fed mutton in hope of beginning a new breeding line in a freshly wiped server, he made a comment on breeding strategy. Specifically, he mentioned that 'most' official tribes do not begin breeding for mutations until they tame a Giga with 50 points in its melee stat. I disagreed, noting that breeding a Giga with 40 points in its melee stat for mutations was likely to produce 5 new mutations faster than finding one with such a high melee stat. Logically, he brushed me off, quoting some ancient wisdom bestowed on him by some former tribemate that might or might not have been in some official tribe at some time. I did not let this go.

At the time, I had to produce a Mathematics investigation on a topic of my choosing. As a result of my petty disposition, I chose to shoot myself in the foot with a topic well beyond the syllabus I was following. Particularly, I explored the compared efficiency of taming or breeding dinosaurs at a given cut-off stage of points in a desired stat. My goal was to find what I dubbed the 'switching point', the number of points in a desired stat at which it becomes more time efficient to breed for mutations than to tame a new dinosaur in hopes of it having more points in that stat.

This is that investigation:

Maths Investigation (1).pdf 383.91 kB · 1 download

I had to compress it a lot to upload on here because of the 400kb limit, so some parts aren't particularly legible.

There are some assumptions made throughout the investigation to stay on subject. Many of these are barely relevant to its applications, some should be accounted for when using the formula produced:

Particularly, t-tame is not the actual time to time a dinosaur, but rather the average time it takes you, specifically, to find, tranq, and tame a dinosaur (you should take into account offline time, since raising dinosaurs can be done when offline). E-breed does not account for breeding several dinosaurs at once (as is common practice), since, by doing this, you essentially get one useful mutation every time you hatch a set of eggs, you can replace t-breed by the raising time of your dinosaur in the final formula (you can also just track the average number of useful mutations (level increase, not actual mutations) you get in an hour over a couple of days and replace all of E-breed by this number). I will expand on these and other assumptions if there's any interest.

Conclusion?

42?

 

[GrumpyBear]

For what its worth, the tribes who started mutating r-gigas on our pve official found 46 points melee and mutagened it to 51.   

I'm not sure if that was the highest found at the time.  I would assume that the use of time saving on genesis 2 would offset any drive to wait for a 50 point melee.   IF we are talking a serious run at 253-254, then with gigas its all about how long it takes to get mutations on male studs.   Any time saved waiting for the right one could be washed out by the competitive process.

The race would be won by whoever was the most organized, had the most effective group working on building multiple projects at once to make it happen.  It wouldn't matter who started with 40 point or 50 points if the people who started with 40 were dedicated, and the person who had the 50 had a life and wasn't in a rush.

It would boil down to managing; 

 

• initial set of low level breeders
• farming mutagen
• potentially building a level 1 female
• cloning up to 500 breeders , potentially 2 separate times
• starting a counter roll project , approximately a 14 month project with gigas
• breeding, the simple part....
• and need competition, Your group and some other group have to both really want to be the first to do it.

 

It would come down to how well a group of players managed these priorities.  One person could do it on their own, but they need to be single or ready to get a divorce.

 

[Pipinghot]

3 hours ago, GrumpyBear said:

but they need to be single or ready to get a divorce.

Welcome to ARK.

[GrumpyBear]

@LucLon  I guess I got 1 possible critique on primary assumptions.

• "; however, the taming and breeding processes described are equal for all dinosaurs and servers, differing only in time expenditure." 

 That time expenditure can be quite significant.  I'm not sure you can not include that into the equation.  There's a limited # of gigas on any server at any given time.  This puts an additional limit on how often one could find a giga that could even produce a 40 point or 50 point melee stat.   

Most other types of animals, I can find 10-20 candidates worth taming in one taming session and could knock them all out and all their taming processes could run concurrently.  The giga would be limited to however many can spawn on a given server at any given time.

 

I'd say there's a variable that would, for most occasions never change. But when dealing w/ an animal that only spawns 2 or 1 at any given time, that should change for those occasions.   What you've done on the paper is pretty darn sweet, should be worth an A in a capstone class for that degree.

 

* Assumption we are using Official 1x rates; I still stand on the, 10 point difference is a 40+ day head start, but the race will take well over 800 days ignoring event rates.   Its a marathon, not a sprint.

[wizard03]

Sadly in SP mode, and without the aid of "SUPER SPYGLASS" 

I actually do feel this pain. Its different, but in the end, the similarities to this mathematical equation of foregoing a perfect tame and just breeding for mutations has become a hard truth that's tough to swallow.
In the end, when preparing for boss fights, I typically have to give up on finding perfect or near perfect parents and usually go with whatever I find. And honestly, in the few times I do give up and find whatever, sometimes those stats are actually better than the perfects or "acceptables" that I thought was choice dino's to breed out.
For example, I found a typical run of the mill lvl 145 rex....6800 health, 300%ish melee. Wasn't bad, but wasn't great. I seen a tek rex later on, 7200 health, 300%ish melee......it was only a level 108!!!!!
Sadly without super spyglass involved, the only way to see where the actual levels are at, you need to knock them out first, and even then, you still technically need to tame them to see where the final levels are distributed, and as such Iv found it best to only tame 5-7 creatures over level 100, sort through them to see if you got any true stand outs, breed the stats out of the ring leaders, and then after you have nursed those numbers out into clean offspring, start working on mutations. 

Now while this is on single player, sadly I can see this being an absolute pain in the neck online. Your going to spend years of your life to reach caps, just for a single creature. Considering there is hundreds of different creatures to tame that is breed able, yeah. Thats going to be a life sucker, hence why many people would just pay for eggs with poly.

[LucLon]

On 12/13/2022 at 11:48 AM, SunsetErosion said:

Conclusion?

42?

 

The specific number conclusions reached at the end of the investigation are only for a specific set of server settings (If I remember correctly, official servers before the second round of doubled taming and raising rates). However, part 4 of the investigation produces a general formula that allows you to find the switching point for your specific settings and situation, there's a table that corresponds different switching points (just number of points in the desired stat) to your result from the E-breed formula.

As mentioned in my original post, the E-breed formula should be adjusted for players who hatch several eggs at once (as most do) and thus effectively get one useful male mutation for every round of breeding. If this is the case for you, replace the t-breed variable by the raising time of your dinosaur in the E-breed formula (Simplifies to 3/(140 * t-breed) ).

In other cases (say, you don't breed 24/7, or don't have a continuous stream of eggs to hatch), you'll get the most accurate calculation of E-breed by counting your point increase through mutations over some days, then dividing that by the time tracked (in hours).

In the future, I'll look into generalizing the E-breed formula for the use of several breeders at the same time. If you want to do this yourself, you can replace P(no mutations) by P(no mutations) raised to the number of breeders you have. Then, P(mutations) and the subsequent infinite series can be calculated through the same equations used in part 4 (page 14, P(mutations) is the number 0.0346, P(no mutations) is the number 0.965, you'll have to replace both).

Once you've calculated E-breed by any means, you multiply it by t-tame (simply the average time in hours it takes you to tame a dinosaur, including offline time), then refer to the table in pages 16 and 17 for the switching point.

[LucLon]

19 hours ago, GrumpyBear said:

@LucLon  I guess I got 1 possible critique on primary assumptions.

• "; however, the taming and breeding processes described are equal for all dinosaurs and servers, differing only in time expenditure." 

 That time expenditure can be quite significant.  I'm not sure you can not include that into the equation.  There's a limited # of gigas on any server at any given time.  This puts an additional limit on how often one could find a giga that could even produce a 40 point or 50 point melee stat.   

Most other types of animals, I can find 10-20 candidates worth taming in one taming session and could knock them all out and all their taming processes could run concurrently.  The giga would be limited to however many can spawn on a given server at any given time.

 

I'd say there's a variable that would, for most occasions never change. But when dealing w/ an animal that only spawns 2 or 1 at any given time, that should change for those occasions.   What you've done on the paper is pretty darn sweet, should be worth an A in a capstone class for that degree.

 

* Assumption we are using Official 1x rates; I still stand on the, 10 point difference is a 40+ day head start, but the race will take well over 800 days ignoring event rates.   Its a marathon, not a sprint.

I agree, before beginning my investigation, I held that the hassle of taming most dinosaurs worth fighting for would make the process much less efficient than simply taming one decent dino and just breeding from then on. This investigation has mostly reinforced that belief. Still, as you mentioned, when it comes to more abundant dinosaurs (say, rhinos), it becomes significantly more efficient to mass tame them than to farm mutations, especially since noone wants to keep 200 breeder rhinos chilling in their base, let alone have to deal with 200 baby rhinos at once.

Of course, this investigation is most applicable to servers that wipe somewhat regularly. In these, the fresh-wipe taming and breeding rush are large determinants of which tribes outgrow and out-fight their competitors, so the efficiency gained from selecting the right switching point is not dwarfed by the long marathon of sustained mutation farming.

As to the use of mutagen. It has no effect on the switching point, since it only provides a flat increase in points that is equal for any switching point.