Stat Breeding: Birth Stats Formula Update March 12

Terms and definition

These are definition I personally use to avoid confusion.

Individual stat: the singular stat of a wolf’s five types of stats. In this case, a wolf have five individual stats (strength, speed, agility, wisdom, and smarts).

Parents’ total individual stat: the combined same-type individual stat of two parents. For example, if father has 103 strength and mother has 64 strength, then the parents’ total individual stat for strength is 167.

Grand total stat: the total of all five individual stats of a single wolf. For example, if a wolf has 103 strength, 196 speed, 136 agility, 103 wisdom, and 185 smarts, that wolf’s grand total stats is 723.

Parents’ grand total stat: the total of all five individual stats for a pair of breeding wolves.

Updates

Statistics Update 1: November 21, 2020
Statistics Update 2: November 28, 2020
December 5, 2020: Iron #7030 released a calculator here.
March 12, 2021: It appears that for parents' total individual stats greater than 450, there have been patterns where the range is below 21 (I personally have seen data for add-ons to be as low as only 12). At this time the formula can continued to be used for parents' total individual stats lower than 450. We are in the process of identifying the patter for high-level breeding.

The Formula

By extension, the formula to predict grand total birth stat is as follows:
grand total birth stat = parents' grand total stat x 20% + (range of 105 - 155).

This formula excludes runts and largest puppies.

Example

For example, I had a litter between these two:

Father: 102, 196, 133, 103, 184 (str, spe, agi, wis, smr)
Mother: 68, 114, 117, 69, 72

Based on my formula I am predicting that the litter will fall within the following range:
Str: 55 - 65
Spe: 83 - 93
Agi: 71 - 81
Wis: 55 - 65
Sma: 72 - 82

Grand total birth stat is predicted to be 337 - 387.

Actual puppies were as follows, all of which fell within the predicted range:
Puppy 1: 59, 88, 76, 62, 77; total 362
Puppy 2: 59, 89, 77, 59, 78; total 362
Puppy 3: 61, 88, 78, 61, 76; total 364

I only have stats for one litter, but your formula works quite well. Need more data to confirm.

Father: 55, 85, 67, 97, 64

Mother: 54, 46, 55, 89, 79

Predicted results: 42.8-51.8, 46.2-56.2, 44.4-54.4, 57.2-67.2, 48.6-58.6

Pup 1: 46, 50, 47, 63, 54

Pup 2: 44, 49, 50, 63, 54

Pup 3: 44, 49, 47, 61, 54

Pup 4: 44, 49, 47, 64, 51

Pup 5: 44, 51, 48, 66, 53

Hey Sol, just to let you know I've not gotten a single pup that was less than 30% of parent's overall stats. The randomization seems close to around a variation of 5 than 20-30.

As far as individual stats go the results are very consistent in pups suggesting that randomization is not at play for the overall bulk individual stats.

Hi Magpie, may I ask what stats are the parents of your litters?

I've been keep track of a variety of different total stats ranging from 800 - 1500s, and you will definitely see a drastic decline when you hit the 1300 - 1500 range. Puppies in the ~500 birth stat range have parents' combined stats of ~2000, which would fall in line with my formula of 20% x 2000 + (5 stats x 20-30) = 500 - 600 range.

My line of best fit was actually y = 0.19x + 28 (removed excess decimals), which is what I based my loose formula on.

I agree that randomization may not be as large as 20 - 30, but I've kept it large in my suggestion to just try to capture all possible ranges. With more data we might be able to determine the actual range.

In terms of the formula though, does it accurately predict the individual stats of your litter?

That's interesting I wasn't seeing the large stat wolves when I first collected for my litters. Which the huge gap in randomization it would but at lower levels the randomization does not fit. The stats system seems to fluctuate based on the number of stats the wolf has so I'm not sure if there are multiple formulas or how it changes with the difference in totals.

At a lower level litter project I got with 5 different litters I only got a 4 stat variation across 20 pups in a specific stat and a 7 stat variation in total stats. Your formula covers what seems to be a planned variation in individual stats. i.e. two wise wolves will produce wise offspring.

The formula calculates individual stat, not total stat if that's not clear. The range is not meant as an indicator of the spread of your pups, but as a mean to catch all possible outcomes. 

For the strength example in my OP, I had predicted the strength of my pups to be within 54 - 64. If any pup is born within that range I would say my formula worked. If I had a pup that was born say 50 or 68 strength, then I would say the formula did not correctly project the breeding.

Can you provide an example of how my formula does not fit lower stat breedings (with parents' individua stats at breeding, and pups' individual stats at birth)?

At the higher stat breeding does the percentage of the parents stats ever go below 30%? Just curious since in my breedings they never go below 30% of parent's stats.

Dad wisdom 81

Mom wisdom 47

Total = 128

20% = 25.6 (+20-30) = 45.6-55.6

Stats in wisdom varied from 48-51 throughout 20 puppies. So it did predict it the range it was in. However out of 20 puppies from 5 different mothers that was the range in wisdom. Excluding runts and largest puppy.

In smarts out of 15 puppies it ranged from 59-62. Different total stats(171), but the point is more so how narrow of a range there was.

Yes the range is high right now, but that's so I want a good starting point. I'm currently gather more data so that hopefully we can reduce the range down to close to what it actually is.

Parents' total stats at around 1500 would produce pups at around 28%, and 2000 would produce pups at around 25%. 

Statistics Update 1: November 21, 2020

The graph is plotted from 245 individual birth stats (49 puppies). The lowest parents’ total individual stats is 150 and the highest is 463, with a mean of 232 and a median of 230. The range of parents’ grand total stats ranged from 956 to 1,501, with a mean of 1,176 and a median of 1,120.

The line of best fit for the data is y = 0.2046x + 25.721. This means that the average pup’s individual birth stat is equal to 20.46% of its parents’ total individual stat + 25.721.

I have received feedback on the top-up range (20-30) for my formula, so I also calculated the difference from the actual individual birth stat to 20% of parents’ total individual stat.

Although most top-up falls in the range of 24-30, there are still a few that falls in the 21-23 top-up range. As such, I feel that a wide range is still necessary to capture all the possibilities.

In addition, I’ve noted that one of the individual birth stat actually fell on 31 and none fell on 20. This could be the way the rounding is coded into the system, and a better range may actually be 21-31. I will wait for another of stats before officially amending the formula.

I will post an update every time I have about 50 new puppies. Once again, feed back and your observations are welcomed.