Y35 New Gender Selection Format

[Art K Stables]

Read every note, yet I found only one question mildly interesting.. Do the colts run a bit faster like real life or not.. Unless there is a Zenyatta in the game, but thats about it

[Eric Nalbone]

Yes - colts are generally faster. Unless you're Susie or I and breed horses like Whitney/Spark and Amber/Lying I suppose, but even with exceptional bloodstock those types are once in a blue moon horses. (Edit - I know others have fillies that have successfully competed against colts. I don't pay attention to them, so I don't know who they are. I only know about mine and Susies and usually Laura's. Sorry.)

Just for kicks, I've been revisiting this topic throughout the week. I do think that a quick glance at early returns indicates that the new 20-choice limit isn't having an immediate result on gender distribution. Obviously comparing the results of approximately six days to previous years of accumulated 16-week chunks is an apples-to-oranges comparison, but players are still apparently intentionally skewing "random" to a pretty significant degree. With 454 fillies from 833 foals bred so far in the SIM, the probability of having 454 or more fillies if each foal were actually bred as a 50/50 random probability would be 0.4% ... which is to say from a Statistics 101 standpoint (95% confidence interval) we could be reasonably confident that something is intervening in the process to make this result distinctly "non-random."

Of course, there's my own barn as the point that especially with a small sample size (and relatively speaking, 833 is a small sample) random things that would appear "non-random" do occur with surprising regularity. From 14 foals (13 bred randomly, one forced COLT from Whitney), I have 11 fillies ... in a truly random distribution, I'd expect 11 or more fillies from 14 trials only just less than 3% of the time (which is actually slightly within a 95% confidence interval, so your statistics-textbook analysis would tell you to accept the null hypothesis that the distribution occurs randomly). For kicks, that means it's 50/50 (the forced Whitney colt could have gone either way) whether random would have landed me at a result I expect about 1% of the time (12 or more fillies from 14 foals) or 3% of the time (where I am now, with a forced colt).

For sensitivity analysis and context, understand that adding one more randomly-bred colt (which I have a 50/50 shot of doing assuming the next mare I breed is random) would mean that I'd expect 11 or more fillies from 15 foals slightly less than 6% of the time. Breeding two more colts in a row (25% chance of that) would give me 11/16, and I expect 11 or more fillies from 16 random trials about 11% of the time. So we quickly get back towards the realm of "normal," as more random trials are stacked on a small sample size. Luckily, there are ~33,000 more foals to be bred in the SIM this year, so hopefully multiplying our sample size by 40 will help.

For anyone who wants to have fun with SIM numbers, visit a binomial probability calculator here. I haven't gone through and checked the math on anything myself, which means that now that I've posted it publicly Ronnie will surely test it and prove it isn't calculating properly (would be just my luck!) but it seems to work with small numbers so I assume it works with large ones. I'm not doing anything that involves 833! by hand.

[Anna Leroux]

My personal thoughts on the whole gender thing. One of the things I like about this game is getting to pick the genders of my foals, and the colors. No it isn't realistic but theres enough realism in life off the computer that I like my internet time to be less real and more enjoyable. I bred 5 foals last year just because I didn't feel like messing with it to much but I think that removing the option even for one year cuts out some parts of the "fun" aspect because not everyone wants this to be just like real. If this was more realistic you'd have more injuries/deaths of horses which I'm sure isn't something players want. One of the reasons I originally joined was the ability to pick what gender I wanted my foals to be. That is part of what makes the game great. For instance if I am aiming to win all the filly stewards cup appy races then I'd need to make sure I have fillies from my high quality mares to accomplish that goal in a year. Or if I wanted to win all colt races. Yes these are random achievements but fun ones. And say for instance the end of the year is close you've used all your 20 picks and a great lease comes up that you really want a specific gender from and random gives you the opposite when you know a colt would have been better from a line, or a filly because that guy sires better girls. Just random thoughts here. And no one should be kicked out for not liking sticking to 50/50, as many new players don't know that rule, and many longer running players want to do something different.

[Radical Edward]

I want to add a "me too" to Anna's comment. Games should be fun and it's OK to have a limit on the realism.

[Bradley Davis]

Remember that this year is a test case, and that we may not be 'stuck' with it. As far as I can tell, this new method came about not to 'enforce' a realistic gaming experience, but rather as a response to the overpopulation in the SIM resulting from too many fillies being bred.

[J K Rowling]

I feel like the people not in favor of this are reacting the same way as those who got upset about insurance and random injuries, and yet there's only one person still upset about that.

[Ronnie Dee]

...Just for kicks, I've been revisiting this topic throughout the week. I do think that a quick glance at early returns indicates that the new 20-choice limit isn't having an immediate result on gender distribution. Obviously comparing the results of approximately six days to previous years of accumulated 16-week chunks is an apples-to-oranges comparison, but players are still apparently intentionally skewing "random" to a pretty significant degree. With 454 fillies from 833 foals bred so far in the SIM, the probability of having 454 or more fillies if each foal were actually bred as a 50/50 random probability would be 0.4% ... which is to say from a Statistics 101 standpoint (95% confidence interval) we could be reasonably confident that something is intervening in the process to make this result distinctly "non-random."
...
For anyone who wants to have fun with SIM numbers, visit a binomial probability calculator here. I haven't gone through and checked the math on anything myself, which means that now that I've posted it publicly Ronnie will surely test it and prove it isn't calculating properly (would be just my luck!) but it seems to work with small numbers so I assume it works with large ones. I'm not doing anything that involves 833! by hand.

I regularly use this calculator. It is accurate and very easy to use.

One very minor note on the p-value from the statistical test using the exact Binomial distribution: you should include the "equal" in your probability reporting because you are reporting the "observed or more extreme". That is, you should have reported P(X>=454)=0.005 rather than P(X>454)=0.004 (but this extremely minor difference of 0.1% does not change the interpretation).

[Paul Heinrich]

I'm not surprised we are heavily skewed filly early on in the season; this is usually true, as people often do a lot of leases and their best mare breedings early in the season. I would expect this to start ticking down somewhat around week 12, as it usually does.

For what it's worth, I'd expect something around 53% fillies at the end of the year, with the current system. There were 33000+ foals born last year, meaning there were ~33000 broodmares in the game, distributed mostly amongst something like 300 active and established players. If all 300 of those players used their allotted 20 breedings exclusively for fillies, we'd end up with something like 59% filly (if we assume close to the same number of breedings as last year). I don't think it will be that severe - lots of players will not choose 20 fillies, either because they like random, or will select colt for some mares. But I would guess it will land between 53% and 54%, as it currently stands.

Considering that we're including mixers too here, I'm not as alarmed as I would be if we were only talking about thoroughbreds. If the powers that be want to produce something close to 50/50, the best way to tweak this system, IMO, would be to allow for unlimited colt selections, and to adjust the *random* generator to favor colts something like 53/47. I'm not saying that needs to be the goal - that's up to Em and team. But insofar as this current system limits individual players ability to manipulate the fake gene pool, whatever the reasons, I approve, even if it still results in more fillies.

[Ronnie Dee]

If the powers that be want to produce something close to 50/50, the best way to tweak this system, IMO, would be to allow for unlimited colt selections, and to adjust the *random* generator to favor colts something like 53/47. I'm not saying that needs to be the goal - that's up to Em and team. But insofar as this current system limits individual players ability to manipulate the fake gene pool, whatever the reasons, I approve, even if it still results in more fillies.

If the female/male ratio is too skewed in favor of females with 20 gender selections per player in Year 35, a simple solution would be to reduce the number from 20 to 10 in Year 36. I think a step-by-step approach is a much better way to go rather than trying radical solutions.

[Friedrich Barbarossa]

I've bred 7, chosen 3 fillies due to leases, and the remaining 4 are split evenly. So from that insanely large sample size I'm right were I should be.

I don't expect to use all 20, it would be cool if we could sell leftover foal gender selections. Hehe...

[Paul Heinrich]

I've bred 7, chosen 3 fillies due to leases, and the remaining 4 are split evenly. So from that insanely large sample size I'm right were I should be.

I don't expect to use all 20, it would be cool if we could sell leftover foal gender selections. Hehe...

Like the carbon tax credits. Laura could be the Al Gore of gender selection. I like it!

[Laura Ferguson]

Is the goal to make the split 50/50, or is the goal to make sure fillies don't run amok? The two goals aren't necessarily the same. The lesser of the evils, to me, is to allow you to breed all the colts you want, select up to X fillies, and then after you hit X, you have to go random if you want to shoot for a filly.

Picking the gender is a huge part of the fun factor to me, and I know I'm ending up with more fillies than I want for both last year and this year as a result of this change, which is NOT fun. It's much easier for me to cull underperforming colts than underperforming fillies, so I usually skew 65-70% colts for the horses I breed to race under the Nursery Stud silks. I get tired of the realism argument - if you want realism, a much larger chunk of foals should never make it to the track, and a huge percentage of 2yos should have nagging problems that prevent them from running until they turn 3, not bucked shins that they recover from early in their 2yo year. There's a happy medium between realism and fun. I totally get the need to rein the percentage of fillies in, and that there has to be some way to address it, so I'm trying to keep an open mind about this, but when the fix to contain the number of fillies has me ending up with MORE fillies (even though I get that the SIM as a whole may have less fillies), I can't help but shake my head.

[Matt Feldman]

it would be cool if we could sell leftover foal gender selections.

This would also help with mare leasing in those cases where the lessor (mare owner) wants to be breeder. Want to choose the gender? Send one of your 20 selections to the lessor.

[J K Rowling]

...allow you to breed all the colts you want, select up to X fillies, and then after you hit X, you have to go random if you want to shoot for a filly.

I actually really, really like that idea. For most of my mares I do 2 colts first to see if they can produce. If they can, I get awesome stud prospects/money earners. If they can't, I'm not letting junk fillies run around and I can dump the mare.

[Stormy Peak]

Does someone have a year by year account of how many more fillies there were over colts?

Also, I don't understand why it even makes a difference?

So what if there is 60% fillies? They race - they retire...and can only have 1 foal (sometimes 2 foals a year). Where as the colts, a male can sire 1 on up into the hundreds each year. Those that can't become sires become pretty much worthless geldings upon retirement - whose only hope to do anything for a stable is to become a rodeo, event horse or something and even then...older mares that don't produce nice foals can do that too.

I just don't see the need for gender control here in this game...even if the gender does slant more in one way then the other.

I like a lot of the realism in the game but at some point if it gets too real it loses a lot of the fun and fantasy aspect that drew so many here....including myself.

Stormy