Jump to content

The Ultimate Stat Balancing Thread


Recommended Posts

Posted

I've been calculating on Excel exactly how the data in the game is distributed. Because on the whole the game's women are significantly more talented than real life women, I have excluded them, and am only counting active male wrestlers who have already debuted.

 

First, a breakdown:

 

q1, median and q2 are the 25th, 50th and 75th percentiles of the data group respectively. If you were to sort the data from highest to lowest and go 25% down the list, you would find q1. The iq range is the interquartile range, and is simply q2-q1. These give you an idea of the central tendency of the data, and have the advantage of not being greatly altered by extreme values.

 

To calculate these in Excel, use these:

q1 =QUARTILE(A1:A50, 1)

median =QUARTILE(A1:A50, 2)

q3 =QUARTILE(A1:A50, 3)

 

Where A1:A50 is the all the range of data you are looking from. (When using data from the game, this range may go from A1:A1000 or even more!)

 

The max and min are the biggest and smallest values in the group of data. These will let you know the extremities of the data.

 

=max(RANGE)

=min(RANGE)

 

The mean is simply the average value of the data. While useful, it shouldn't be used on its own as it can be greatly biased by extreme data.

 

=Average(RANGE)

 

 

SD (standard deviation) and variance let you know the spread of the data. Generally all data should be encompassed between 2 standard deviations of the mean. For example, if the mean is 50 and the standard deviation is 15, most, if not all, of the data should be between 20 and 80 (15 x 2 = 30, so 30 above and below 50). SD is the square root of the variance

 

=VAR(RANGE)

=sqrt(VAR value)

 

The 'Sum Average' Column is a column I made myself. It is the average of every stat a person can have. I made it as a very rough idea of how 'good' the best workers are, and how 'bad' the worst workers are.

 

So if you've managed to understand all of this so far, good!

 

So as to not stretch the screen, I have split the data up a bit.

 

http://i47.tinypic.com/30mpmch.png

 

Our main notes to gether from this are:

1. There are some people who have absolutely zero skill in a certain attribute.

2. Excluding Submission, absolutely nobody has 100 in any skill from this set.

3. Puroresu seems to have a low distribution. This is because the puro stat is biased towards Japanese workers and very few workers from the rest of the world will have much (if any) skill in this stat.

4. Q1 and Q3 tell us where the majority of workers will lie between for these stats

 

http://i45.tinypic.com/vzx1km.png

 

Our main notes to gether from this are:

1. Nobody has any Announcing skill. If you are a wrestler, you are not a play by play man.

2. Most workers have decent athleticism, safety, selling, stamina, basics and toughness, with the majority lying between 50 and 75.

3. Psychology tends to be at around 50 on average. 65 would be considered good.

 

http://i49.tinypic.com/11vr8lt.png

 

Our main notes to gether from this are:

1. Very few wrestlers make good business men or bookers (shocker, I know)

2. Reputation is almost always 100.

3. Star Power is quite varied. The middle 50% of workers lie between 40 and 70 and the average is around 60.

4. Most wrestlers will have around 40-50 in menace, unless they look like a girl or a monster.

 

http://i50.tinypic.com/2ensxsw.png

 

Our main notes to gether from this are:

1. 40-70 is about average for stiffness (just like in EWR!)

2. Active wrestlers aren't good referees.

3. Unless injury prone, a worker will have at least 80 resilience.

4. If you add up every stat a worker has and get an average, it should almost always lie between 40 and 50, with very very few exceptions. This goes to show that generally every good worker has weaknesses, every bad worker has some redeeming qualities, and average people tend to be average across the board!

Posted
I've been calculating on Excel exactly how the data in the game is distributed. Because on the whole the game's women are significantly more talented than real life women, I have excluded them, and am only counting active male wrestlers who have already debuted.

 

First, a breakdown:

 

q1, median and q2 are the 25th, 50th and 75th percentiles of the data group respectively. If you were to sort the data from highest to lowest and go 25% down the list, you would find q1. The iq range is the interquartile range, and is simply q2-q1. These give you an idea of the central tendency of the data, and have the advantage of not being greatly altered by extreme values.

 

To calculate these in Excel, use these:

q1 =QUARTILE(A1:A50, 1)

median =QUARTILE(A1:A50, 2)

q3 =QUARTILE(A1:A50, 3)

 

Where A1:A50 is the all the range of data you are looking from. (When using data from the game, this range may go from A1:A1000 or even more!)

 

The max and min are the biggest and smallest values in the group of data. These will let you know the extremities of the data.

 

=max(RANGE)

=min(RANGE)

 

The mean is simply the average value of the data. While useful, it shouldn't be used on its own as it can be greatly biased by extreme data.

 

=Average(RANGE)

 

 

SD (standard deviation) and variance let you know the spread of the data. Generally all data should be encompassed between 2 standard deviations of the mean. For example, if the mean is 50 and the standard deviation is 15, most, if not all, of the data should be between 20 and 80 (15 x 2 = 30, so 30 above and below 50). SD is the square root of the variance

 

=VAR(RANGE)

=sqrt(VAR value)

 

The 'Sum Average' Column is a column I made myself. It is the average of every stat a person can have. I made it as a very rough idea of how 'good' the best workers are, and how 'bad' the worst workers are.

 

So if you've managed to understand all of this so far, good!

 

So as to not stretch the screen, I have split the data up a bit.

 

http://i47.tinypic.com/30mpmch.png

 

Our main notes to gether from this are:

1. There are some people who have absolutely zero skill in a certain attribute.

2. Excluding Submission, absolutely nobody has 100 in any skill from this set.

3. Puroresu seems to have a low distribution. This is because the puro stat is biased towards Japanese workers and very few workers from the rest of the world will have much (if any) skill in this stat.

4. Q1 and Q3 tell us where the majority of workers will lie between for these stats

 

http://i45.tinypic.com/vzx1km.png

 

Our main notes to gether from this are:

1. Nobody has any Announcing skill. If you are a wrestler, you are not a play by play man.

2. Most workers have decent athleticism, safety, selling, stamina, basics and toughness, with the majority lying between 50 and 75.

3. Psychology tends to be at around 50 on average. 65 would be considered good.

 

http://i49.tinypic.com/11vr8lt.png

 

Our main notes to gether from this are:

1. Very few wrestlers make good business men or bookers (shocker, I know)

2. Reputation is almost always 100.

3. Star Power is quite varied. The middle 50% of workers lie between 40 and 70 and the average is around 60.

4. Most wrestlers will have around 40-50 in menace, unless they look like a girl or a monster.

 

http://i50.tinypic.com/2ensxsw.png

 

Our main notes to gether from this are:

1. 40-70 is about average for stiffness (just like in EWR!)

2. Active wrestlers aren't good referees.

3. Unless injury prone, a worker will have at least 80 resilience.

4. If you add up every stat a worker has and get an average, it should almost always lie between 40 and 50, with very very few exceptions. This goes to show that generally every good worker has weaknesses, every bad worker has some redeeming qualities, and average people tend to be average across the board!

 

Thanks for all this. I personally never noticed this:

 

4. Most wrestlers will have around 40-50 in menace, unless they look like a girl or a monster.

 

before, so my data will be just a little bit better now. ^_^

Posted
If people wish, I can break it down further into weight classes, region bases, styles...any sortable category really!

 

I'm actually curious about weight class and the aerial/power cutoffs between them. :)

Posted

OK, by region:

 

USA

http://img408.imageshack.us/img408/4120/usamales.png

 

UK

http://img94.imageshack.us/img94/4189/ukmales.png

 

Japan

http://img51.imageshack.us/img51/711/japanmales.png

 

Europe

http://img709.imageshack.us/img709/5920/euromales.png

 

Australia

http://img291.imageshack.us/img291/2981/ausmales.png

 

Mexico

http://img695.imageshack.us/img695/9747/mexicomales.png

 

And by weight....

 

Giant/Super Heavyweight

http://img35.imageshack.us/img35/3716/giantandsuper.png

 

Big Heavyweight

http://img442.imageshack.us/img442/3527/bigh.png

 

Heavyweight

http://img85.imageshack.us/img85/7974/heavyweights.png

 

Light Heavyweight

http://img708.imageshack.us/img708/4042/lighth.png

 

Middleweight

http://img717.imageshack.us/img717/1958/middleweights.png

 

Lightweight

http://img16.imageshack.us/img16/4766/lightweights.png

 

Small/Very Small

http://img692.imageshack.us/img692/9356/smallandvsmall.png

Posted

C-Verse isn't RW, it's set at a different time, a different wrestling history and climate. It's good to use as a rough guide but imo nothing more. Take into consideration how many non workers each different mod has sways numbers one way, add to that the average age of a worker in said mod which will also effect lots of different stats in a big way.

 

I think these spreadsheets are popping up now give the impression mod makers don't need to play C-Verse and can just cross reference to make sure they're in line.

Posted

I think its important that mod makers dont try to match up all there workers to these stats just to make there mod match the c'verse because you'll end up with workers having the wrong stats just match the c'verse. What is good for tho is checking how balanced and how to compares to the c'verse for a great gameplay experience

 

Edit - damn genadi u got in there before me, typed faster and worded it better :(

Posted
What is good for tho is checking how balanced and how to compares to the c'verse for a great gameplay experience

 

Like I said though there are so many variables between any two mods (number of active workers, average age, time in history) that sometimes it could give you a wrong impression.

Posted
C-Verse isn't RW, it's set at a different time, a different wrestling history and climate. It's good to use as a rough guide but imo nothing more. Take into consideration how many non workers each different mod has sways numbers one way, add to that the average age of a worker in said mod which will also effect lots of different stats in a big way.

 

I think these spreadsheets are popping up now give the impression mod makers don't need to play C-Verse and can just cross reference to make sure they're in line.

 

As stated, non-workers aren't included in the equations, so there's no swaying. And these things can even be worked out for ages ranges. I do agree that they should only be used as a guide, but they shouldn't be underrated.

 

In fact, I was just comparing your 2002 data to the original figures before you replied. I've always considered it the most balanced mod, and thought the figures might help any future improvements:

http://tinypic.com/r/6r3j8x/6

 

2002 is in yellow, Cverse in white. As you can see the differences are minimal. Most people just look at the mean, but central tendencies and deviations are actually more important as they tell you where the majority of the stats lie and how spread out the figures are.

Posted
As stated, non-workers aren't included in the equations, so there's no swaying. And these things can even be worked out for ages ranges. I do agree that they should only be used as a guide, but they shouldn't be underrated.

 

In fact, I was just comparing your 2002 data to the original figures before you replied. I've always considered it the most balanced mod, and thought the figures might help any future improvements:

http://tinypic.com/r/6r3j8x/6

 

2002 is in yellow, Cverse in white. As you can see the differences are minimal. Most people just look at the mean, but central tendencies and deviations are actually more important as they tell you where the majority of the stats lie and how spread out the figures are.

 

:eek: I didn't realise you went that in depth. I appologise if I sounded like I didn't appreciate your work as it's obvious you've put some time into this and broken down the stats into numbers we can apply. Thank you :)

Posted
:eek: I didn't realise you went that in depth. I appologise if I sounded like I didn't appreciate your work as it's obvious you've put some time into this and broken down the stats into numbers we can apply. Thank you :)

 

Hey don't worry about it! I'm doing a stats module for my degree at the minute and figured I would apply what I know to a game that I like! I knew people wouldn't believe the accuracy of it all so I had been planning on using your data for comparison all along :)

 

 

If anyone else wants their data compared, I'd be happy to oblige.

Posted
Hey don't worry about it! I'm doing a stats module for my degree at the minute and figured I would apply what I know to a game that I like! I knew people wouldn't believe the accuracy of it all so I had been planning on using your data for comparison all along :)

 

 

If anyone else wants their data compared, I'd be happy to oblige.

 

Please do. My data is attached. :)

 

Any suggestions are welcome. ^_^

Kobeverse data.zip

Posted

Wow. Your numbers are almost identical to the C-Verse, distribution-wise!

 

I should note I have removed Booking, Business, Announcing, MMA and refereeing as these are essentially specialist skills and don't apply to most wretlers.

 

http://i50.tinypic.com/35jlzyd.png - Men

http://i46.tinypic.com/21k9k6p.png - Women

(your data is in grey, C-verse is in white)

 

For the men, puro is a little high in the third quartile - basically the top 25% at puro in your game is 50 while in the C-Verse that number is 28. If you look at it in the C-Verse, 75% of workers don't have more than 30!

 

You could bump up flashiness a little for your high fliers and reduce psychology slightly across the board, but they aren't huge things and aren't necessary.

 

I'm sure you can figure the rest of the comparisons pretty easily just by looking, but that's a pretty great job of balancing the data!

Posted

A short while ago I took over the New Power is Rising mod and would you to check that if you could. Do you just need me to zip the worker file up? or do you need it all.

 

Thanks for doing this, very interesting.

Posted
If you go to workers from the editor screen, filter by 'already debuted', 'male' and 'active worker', then hit spreadsheet and send me the xls file that's created!
Posted
If you go to workers from the editor screen, filter by 'already debuted', 'male' and 'active worker', then hit spreadsheet and send me the xls file that's created!

 

I did it for women as well as men. Here

Posted

One thing to note though is the relation of size & style to stats.

 

So an "average" super heavyweight brawler will have a much higher brawling stat then the C-verse average.

Posted

Thanks for all the feedback. ^_^

 

My puro is really high because Japan has a national promotional, two cult promotions and then a throw away T&A promotion - basically just to stick somewhere over there you need to be really good. If I ever add some small and local Japanese promotions that should average things out, I think. :)

 

I just never noticed the flashiness thing because you couldn't sort it in 08' and I haven't gotten around to it in 2010 yet. Thanks for letting me know. ^_^

 

Finally, also appreciate the psychology catch. I actually re-balanced that stat twice since starting the data in mid 2007. I guess I could have another look at it again. :)

 

Side notes:

 

I'm sure my booking, business, MMA and refereeing skills are high (particularly MMA) as I created a glut in business/booking skill to accommodate dojos/prevent promotions from ever opening without an owner. There are also a few workers I wanted to open promotions that I gave business skill. Pet peeves of mine. :)

 

I really want people to be able to successfully open any sort of promotion in the game world, so MMA is really overrepresented compared to CVerse in my data.

 

With refereeing, I just used a different distribution model than the CVerse did. I don't really know the effects to be quite honest. I'll probably tweak it a bit sometime.

 

Will check the announcing stat out and see if there is anything quirky going on there that I didn't notice before. :o

Posted
Thanks for all the feedback. ^_^

 

My puro is really high because Japan has a national promotional, two cult promotions and then a throw away T&A promotion - basically just to stick somewhere over there you need to be really good. If I ever add some small and local Japanese promotions that should average things out, I think. :)

 

I just never noticed the flashiness thing because you couldn't sort it in 08' and I haven't gotten around to it in 2010 yet. Thanks for letting me know. ^_^

 

Finally, also appreciate the psychology catch. I actually re-balanced that stat twice since starting the data in mid 2007. I guess I could have another look at it again. :)

 

Side notes:

 

I'm sure my booking, business, MMA and refereeing skills are high (particularly MMA) as I created a glut in business/booking skill to accommodate dojos/prevent promotions from ever opening without an owner. There are also a few workers I wanted to open promotions that I gave business skill. Pet peeves of mine. :)

 

I really want people to be able to successfully open any sort of promotion in the game world, so MMA is really overrepresented compared to CVerse in my data.

 

With refereeing, I just used a different distribution model than the CVerse did. I don't really know the effects to be quite honest. I'll probably tweak it a bit sometime.

 

Will check the announcing stat out and see if there is anything quirky going on there that I didn't notice before. :o

 

 

Oh, I didn't remove them due to anything you did! Its just that in the C-Verse they are usually reserved for non-workers so I didn't see any point including them in analysis!

 

Basically the data is fine as-is! Most tweaks you do won't improve the game greatly because you have it all covered pretty much.

 

And prague, post #5 has sizes covered. Styles I've yet to get around to.

Posted

Wow... just Wow...

 

I am a guy who hates maths with a passion, could never do it, but you sir do it well and make it understandable...

 

Now a good element of Chaos in there might help but I guess thats what the variance option is for :)

Posted
Wow... just Wow...

 

I am a guy who hates maths with a passion, could never do it, but you sir do it well and make it understandable...

 

Now a good element of Chaos in there might help but I guess thats what the variance option is for :)

 

I'm glad to hear that actually, since I plan on becoming a high school maths teacher!

 

Stealth, just saw your post, I'll get right on it!

Posted

http://i41.tinypic.com/6gc0uw.png

There's the Men

 

 

http://i40.tinypic.com/2lnc076.png

And there's the women

 

 

C-Verse is in white, your data in grey. I'll put up a mini-analysis tomorrow!

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...