Starting Variance - Some Analysis

[Jaded]

<p>I've been super-excited for starting variance having multiple levels because I love anything that makes the game world a little different each time. I wanted to take a look into how much each level actually DID vary and I figured I may as well post it here on the off-chance anyone else is interested.</p><p> </p><p>

To test, I started one game with each level of starting variance, filtered to bisexual workers to get a somewhat random sample of people rather than try and deal with a couple of thousand (my computer's not great and I was worried Excel would crash on me!) and set up a spreadsheet to calculate the difference between each level and none.</p><p> </p><p>

<strong>Notes</strong></p><p> </p><p>

Health of body parts DOES NOT vary.</p><p> </p><p>

Stats that are 0 (so announcing/colour/refereeing for most workers) DO NOT vary. </p><p> </p><p> </p><p>

<strong>Skills</strong></p><p> </p><p>

At minor, skills only vary by 1 point either way. Looking at skills and excluding announcing/colour/refereeing/booking, my test analysed 400 skills. (16 workers with 25 skills each.)</p><p> </p><p>

78.8% of them stayed the same, compared to no variance.</p><p>

10.8% of them went up 1 point.</p><p>

10.5% of them went down 1 point.</p><p>

(All values rounded to 1 dp so add up to 100.1% not 100%, in case anyone notices it!)</p><p> </p><p> </p><p>

At medium, skills varied by up to 5 either way, although these extremes were rare. However most stats varied at least slightly.</p><p> </p><p>

29.5% of them stayed the same, compared to no variance.</p><p>

26.5% of them went up between 1 and 3 points.</p><p>

34.3% of them went down between 1 and 3 points.</p><p>

5% of them went up 4 points.</p><p>

3% of them went down 4 points.</p><p>

0.3% of them went up 5 points.</p><p>

0.5% of them went down 5 points.</p><p> </p><p> </p><p>

At major, skills varied by up to 9 either way, again with the extreme differences being rare.</p><p> </p><p>

21.8% of them stayed the same.</p><p>

19.8% of them went up by between 1 and 3 points.</p><p>

23.8% of them went down by between 1 and 3 points.</p><p>

9% of them went up by 4 or 5 points.</p><p>

9.8% of them went down by 4 or 5 points.</p><p>

6.5% of them went up by 6 or 7 points.</p><p>

5.3% of them went down by 6 or 7 points.</p><p>

2% of them went up by 8 or 9 points.</p><p>

2.3% of them went down by 8 or 9 points.</p><p> </p><p> </p><p>

At extreme, skills varied by up to 14 either way! Similar distribution.</p><p> </p><p>

18.3% of them stayed the same.</p><p>

18.9% of them went up by between 1 and 5 points.</p><p>

26.5% of them went down by between 1 and 5 points.</p><p>

17.8% of them went up by between 6 and 10 points.</p><p>

14.3% of them went down by between 6 and 10 points.</p><p>

4.25% of them went up by between 11 and 14 points.</p><p>

2.75% of them went down by between 11 and 14 points.</p><p> </p><p> </p><p> </p><p>

<strong>Overness</strong></p><p> </p><p>

Looking at overness, I'm pretty certain that if a worker has 0 overness in an area, it doesn't vary at all. Also, while skills seem slightly more likely to go up that down, overness seems far more likely to stay the same than skills do, but if it DOES change (at least by more than a few points) it's significantly more likely to increase than decrease.</p><p> </p><p> </p><p>

At minor, overness only varies by 1 point either way. Looking at all areas where people started with overness, there are 449 in total between the workers.</p><p> </p><p>

93.5% of them stayed the same, compared to no variance.</p><p>

3.1% of them went up 1 point.</p><p>

3.3% of them went down 1 point.</p><p>

(All values rounded to 1 dp so add up to 99.9% not 100%, in case anyone notices it!)</p><p> </p><p> </p><p>

At medium, skills varied by up to 5, although these extremes were rare. Much more overness stayed the same compared to stats varying at medium.</p><p> </p><p>

63% of them stayed the same, compared to no variance.</p><p>

15.3% of them went up between 1 and 3 points.</p><p>

20% of them went down between 1 and 3 points.</p><p>

1.5% of them went up between 4 and 5 points.</p><p>

0% of them went down between 4 and 5 points.</p><p> </p><p> </p><p>

At major, overness varied by up to 9, with the most extreme differences being incredibly rare.</p><p> </p><p>

56.8% of them stayed the same.</p><p>

12.7% of them went up by between 1 and 3 points.</p><p>

21.3% of them went down by between 1 and 3 points.</p><p>

3.4% of them went up by 4 or 5 points.</p><p>

1.3% of them went down by 4 or 5 points.</p><p>

2% of them went up by 6 or 7 points.</p><p>

1.8% of them went down by 6 or 7 points.</p><p>

0.6% of them went up by 8 or 9 points.</p><p>

0% of them went down by 8 or 9 points.</p><p> </p><p> </p><p>

At extreme, skills varied by up to 12 although again the higher values either way were massively rare.</p><p> </p><p>

51.0% of them stayed the same.</p><p>

18.6% of them went up by between 1 and 5 points.</p><p>

21.6% of them went down by between 1 and 5 points.</p><p>

6.3% of them went up by between 6 and 10 points.</p><p>

1.5% of them went down by between 6 and 10 points.</p><p>

0.6% of them went up by between 11 and 12 points.</p><p>

0.2% of them went down by between 11 and 12 points.</p><p> </p><p> </p><p>

I'm assuming few people are as stats-focused as I am, but if anyone wants a look at the Access databases or Excel spreadsheet I used give me a shout.</p>

[lavelleuk]

<p>That's great info! I haven't got the game in front of me, is there an option for just skills/Just popularity variance or does it have to be both?</p><p> </p><p>

Data is based on 16 workers?</p>

[Jaded]

<blockquote data-ipsquote="" class="ipsQuote" data-ipsquote-username="lavelleuk" data-cite="lavelleuk" data-ipsquote-contentapp="forums" data-ipsquote-contenttype="forums" data-ipsquote-contentid="48258" data-ipsquote-contentclass="forums_Topic"><div>That's great info! I haven't got the game in front of me, is there an option for just skills/Just popularity variance or does it have to be both?<p> </p><p> Data is based on 16 workers?</p></div></blockquote><p> </p><p> Has to be both. And yeah, I filtered for 'bisexual' because I just wanted a range of people and that seemed as good a way as any of getting a reasonable mixture. I think it includes at least one person from each area except Japan, several women, people of varying ages, at least one retired worker. Obviously more stats would be more useful but it would rely on my PC not being quite so old!</p>

[lavelleuk]

Haha yeah I get that!

[Taker4455]

<p>This is amazing, Jaded. Very much appreciate the work you've put into this.</p><p> </p><p>

Looks like medium is the perfect compromise for running a historic mod. You still get a decent number of changes, the odd surprise or two, but the stars of yesteryear should still be recognisable.</p>

[Jaded]

This is amazing, Jaded. Very much appreciate the work you've put into this.

 

Looks like medium is the perfect compromise for running a historic mod. You still get a decent number of changes, the odd surprise or two, but the stars of yesteryear should still be recognisable.

 

Glad it's of use! Yeah, I know some people don't like variance in historical mods, but for those who do, medium looks like it should be a great setting.