How can we improve the existing moving creatures of Minecraft? Show us your coolest fantasy creatures. We are very interested in ideas that are unique and not just lists of mobs - spamming lists of animal names is spam and violates our posting guidelines. Be sure to check the Previously Considered Suggestion page (no sharks, MINECON Vote rejected monsters) before you post!

18

horse breeding change - heritability of the stats

3 Comments

Please sign in to leave a comment.

Sorted by oldest
  • 0
    Registered User commented
    Comment actions Permalink

    This or any system that promotes breeding for better stats past the average would be lovely. Would make things like selling horses on servers a new occupation and breeding horses worthwhile.

    Another good change on this might be making horse stats more visible, which is the part that makes Llama breeding less frustrating despite being similarly random with tendency to an average.

  • 1
    Registered User commented
    Comment actions Permalink

    I suspect I know what they were going for, but I also strongly suspect the person who did this didn't understand the math they were creating.

    In real life, domestication and breeding for desired traits beats hunting the wild for desired samples all day. In the implementation they create an asymptotic limit to how good a horse you can breed that means the better the more likely a much worse offspring. So you are better off just finding random horses than you are trying to breed good ones.

    Even worse, the Minecraft implementation isn't fun. It's tedious and frustrating. This isn't good gaming math.

    I don't think it needs to be as complicated as proposed. Simply having a chance for each stat to always improve or degrade would do it. I think what the intention was, was to have children generally inherit from the parent with a chance to get better or worse but they mangled it.

    Say you have a stat that is a float between 0.0-1.0 where 1.0 is the best stat and zero is the worst.
    p1 is this stat for parent 1. p2 is the stat for parent 2. p3 would be the child stat. R is a pseudorandom value between 0.0 and 1.0.

    p3 = min( max( absolute( p1-p2)/2 + 1.1*(R-0.5)*(p1-p2), 0.0), 0.0, 1.0);

    What you end up with (if I haven't mangled it) is an even probability centered on creating offspring that is the average of it's parents. i.e. R will have an average value of .5 which will turn the second part of the equation to zero on average. But there is also a chance for a plus or minus 10%.

  • 0
    Registered User commented
    Comment actions Permalink

    Yeah, It is way to complicated. It is annoying how long it takes to breed a super horse.