Duel2.Com  
•   Home  •  Rules  •  Your Account  •  Forums  • Newsletters  •
Navigate
· Home
· Content
· Encyclopedia
· Forums
· Members List
· Newsletters
· Old Newsletters
· Private Messages
· Setup
· Tourneys
· Your Account
User Info
Welcome, Anonymous
Nickname
Password
(Register)
Membership:
Latest: TheWanderer
New Today: 0
New Yesterday: 0
Overall: 19800

People Online:
Visitors:
Members:
Total: 0
Duel2.Com: Forums

Duel2 :: View topic - Roll Up Pool
 Forum FAQ  •  Search  •  Memberlist  •  Usergroups   •  Register  •  Profile  •  Log in to check your private messages  •  Log in

 
Post new topicReply to topic
View previous topic Log in to check your private messages View next topic
Author Message
pipthetroll
Advanced Master Poster
Advanced Master Poster


Joined: Nov 04, 2002
Posts: 447
Location: In my underwear, in front of my computer

PostPosted: Thu Nov 02, 2006 4:14 am Reply with quoteBack to top

Woody wrote:

I did a bias test on this one, too.

There is no bias in the 7-17 range.

The confidence interval is more than (.90).

In English: It is more than 90% certain that the rollup pool you drew from has no size bias in the 7-17 range.


This sample is polluted. It includes warriors from team sheets(a static set), and replacements(random).
View user's profileSend private messageSend e-mailICQ Number
Woody
Grandmaster Poster
Grandmaster Poster


Joined: Nov 01, 2005
Posts: 989
Location: Lake Powell

PostPosted: Thu Nov 02, 2006 8:10 am Reply with quoteBack to top

pipthetroll wrote:

Whats with the big surprise crack? Theres no surprise at all--the team sheets are a static pool, theres somewhere between 150-250(could be more now), then they repeat.


It's called sarcasm. Wink

Given the huge sample size and the various frequency spikes,
conclusions from Guardian's info could easily be reached by inspection.

I was just curious what my little bias test would say about the confidence interval.
View user's profileSend private message
Woody
Grandmaster Poster
Grandmaster Poster


Joined: Nov 01, 2005
Posts: 989
Location: Lake Powell

PostPosted: Thu Nov 02, 2006 8:25 am Reply with quoteBack to top

pipthetroll wrote:
Woody wrote:

I did a bias test on this one, too.

There is no bias in the 7-17 range.

The confidence interval is more than (.90).

In English: It is more than 90% certain that the rollup pool you drew from has no size bias in the 7-17 range.


This sample is polluted. It includes warriors from team sheets(a static set), and replacements(random).


Again, I was just testing the bias. Nothing more.
View user's profileSend private message
guardian
Advanced Master Poster
Advanced Master Poster


Joined: Nov 05, 2002
Posts: 334

PostPosted: Thu Nov 02, 2006 2:37 pm Reply with quoteBack to top

the bias lies in the fact that sz 15-17 are handicapped by the skill set they recieve and therfore should not fall along the same curve as sz 7-9 unless of course yu can fix the way they are generated to make them as usefull as the average sz 7-9 warrior is ,

the whole reason its wrong is becasue less of them are usefull than the rollups generated at sizes 7-9 .


what part of from a math standpoint it is all correct , but from an in game standpoint it is not are you not getting ?

guardian

_________________
im guardian who the f... are you !.
View user's profileSend private messageSend e-mail
Woody
Grandmaster Poster
Grandmaster Poster


Joined: Nov 01, 2005
Posts: 989
Location: Lake Powell

PostPosted: Thu Nov 02, 2006 8:38 pm Reply with quoteBack to top

Edited

Sorry Guardian,

Frustration getting the better of me.

_________________
...
Not enough time on my hands anymore...
...

Last edited by Woody on Fri Nov 03, 2006 7:52 pm; edited 1 time in total
View user's profileSend private message
The Consortium
ArchMaster Poster
ArchMaster Poster


Joined: Nov 23, 2002
Posts: 8805
Location: on the golf course, in the garden, reading, traveling, and now Consulting

PostPosted: Fri Nov 03, 2006 12:55 pm Reply with quoteBack to top

Rollups from the DA rec'd this week are these sizes:

5
6
7
9
13 (2)
14 (1)
15
18

avg size 11
SZ 14 will make us the best warrior; size 13 next

the "small" ones were
17-10-5-9-9-13-7 (runnable, but in tournament arena)
9-16-7-10-12-9-7 (same)
9-7-6-5-16-14-13 s(ame - would have made an "interesting" burner)


FYI

_________________
Ye Old Consortium Scribe & Crapmaster 2013, Crapgiver 2014
1000 ADM graduates (with 40+ manager IDs) including 105K+ fights and 68K+ wins plus 4 teams with 1000+ wins (Animal Farm DM11 @1518*; Bulldogs DM11 @ 1460; Lenpros DM30 @ 1424; Fandils @ 1336
View user's profileSend private messageSend e-mail
Woody
Grandmaster Poster
Grandmaster Poster


Joined: Nov 01, 2005
Posts: 989
Location: Lake Powell

PostPosted: Fri Nov 03, 2006 8:51 pm Reply with quoteBack to top

guardian wrote:
the bias lies in the fact that sz 15-17 are handicapped by the skill set they recieve and therfore should not fall along the same curve as sz 7-9 unless of course yu can fix the way they are generated to make them as usefull as the average sz 7-9 warrior is ,

the whole reason its wrong is becasue less of them are usefull than the rollups generated at sizes 7-9 .


what part of from a math standpoint it is all correct , but from an in game standpoint it is not are you not getting ?

guardian


I've seen your posts and I understand fully.


Here's what I'm trying to point out:
The problem is the perception of "unfair" dice.

Here's a simplified example:

Let's say the chance of getting a 15+ SZ rollup is 30%
(estimated from normalized dist of LHI's data)

There's a 30% chance of getting hosed.

A player gets a DA rollup back. He's one of the 30% hosed.
No biggie, he'll just DA's this one, too.

The next DA rollup is one the the 30% hosed.
Now the player is upset. The rollup generator is broken. Change something.

The pool is changed. The chance of getting hosed is now 25%
(the amount given by your proposition)

A player has the same double-hosing. He's upset.


What's the difference between 30% and 25% for this double hosing?

(30%)*(30%)= 9%
(25%)*(25%)= 6.25%

Only 2-3 players out of 100 are affected by changing the pool.
In more severe cases of getting hosed, it affects even fewer.

I am not against changing the pool.

I am saying that it will not change the underlying problem.

The problem is the perception that everyone gets their rollups in a percentage consistent with the overall distribution.

This will probably happen...after about 1000 DA'd warriors.

For the level of warriors that most players will DA, there is no way to guarantee the same percentages.

Probability is not fair with small sample sizes.

_________________
...
Not enough time on my hands anymore...
...
View user's profileSend private message
guardian
Advanced Master Poster
Advanced Master Poster


Joined: Nov 05, 2002
Posts: 334

PostPosted: Sun Nov 05, 2006 10:15 am Reply with quoteBack to top

woody


exactly , its not a big change , its alittle one , the kind rsi is more liklely to make , but it is moving in the right direction , rsi makes a tiny change , some managers notice a small difference and that it is a difference for the better , not the worse , and


hope begins



then if it all goes well maybe rsi ups it a little more , and no one complains becasue , everyone sees it as an improvment , one of the frustrating things in the game is , less frustrating

and so on....

_________________
im guardian who the f... are you !.
View user's profileSend private messageSend e-mail
Woody
Grandmaster Poster
Grandmaster Poster


Joined: Nov 01, 2005
Posts: 989
Location: Lake Powell

PostPosted: Sun Nov 05, 2006 6:41 pm Reply with quoteBack to top

guardian wrote:
woody


exactly , its not a big change , its alittle one , the kind rsi is more liklely to make , but it is moving in the right direction , rsi makes a tiny change , some managers notice a small difference and that it is a difference for the better , not the worse , and


hope begins



then if it all goes well maybe rsi ups it a little more , and no one complains becasue , everyone sees it as an improvment , one of the frustrating things in the game is , less frustrating

and so on....


I do not object to the goal.
However, I do not think that this will accomplish the goal.
The problem is that probability will still leave some people hosed.

I crunched a few numbers yesterday and here's an example of how this problem is affected by both perception and reality.

For the simplicity of the math and descriptions, I have made some generalizations and assumptions. They are:

-The 50% best (subjective) rollups in the pool are "good" rollups.
-Redundant draws for the same manager are ignored (sampling without replacement).
-The DA draws are normally distributed.
-The DA pool is relatively large (a few thousand).

Given these initial assumptions:

If 100 players DA 10 warriors each,
15 of the players will get no more than 2 "good" rollups.

These 15 managers are quite sure to complain about the rollup pool, even after a shift.

In English: The DA can, and will, hose some people. Badly.
--
--
--

The difference with the team sheet pool.

Not as many managers feel like their getting hosed by the team sheets because: You generally don't get hosed with all of the warriors on a given team sheet AND you don't get hosed by probability as badly with the team sheets.

The reason probability doesn't hose you as badly with the team sheets is because of something called the "finite population correction factor" (fpc).
The effect of the fpc is probably why RSI only has 200-250 sheets.
After you've bought about 10-12 team sheets, the fpc starts kicking in, and your chance of getting hosed reduces.
When players expect the DA to act similarly, they get disappointed.
--
--
--

I'll answer the next question before someone posts it...

-Why not just reduce the DA rollup pool to 200-250 warriors, too?
--Because then you start eliminating links in the "luck chain".
--I described this problem in the "Unintended Consequences" article, so I won't revisit it here.

The problem with the game isn't the rollup pool. It's economics.
Buying more rollups is a linear function; the distribution of those rollups is not a linear function.

If I knew easy fixes for these various problems, I would post them.

The only idea I've had for increasing arena play (I think this was the goal) is: reward arena play by giving 87 point rollups(or something similar) as replacements for graduates with 20+ arena fights.


Please let me re-iterate:
-I am not trying undermine this, or any, game change; nor am I otherwise trying to flame.
-I am just trying to show that these changes might not have the desired effects.

_________________
...
Not enough time on my hands anymore...
...
View user's profileSend private message
guardian
Advanced Master Poster
Advanced Master Poster


Joined: Nov 05, 2002
Posts: 334

PostPosted: Sun Nov 05, 2006 7:26 pm Reply with quoteBack to top

woody

perhaps not , but doing nothing certainly is not going to make things better in the basic arena's . some change to get better rollups through the arena rather than through buying team sheets , is bound to have some effect , even if it only means that managers become more active for tourney prep in those arena's instead of just 82 .

the point of a discussion is to bring out all the ideas and to perhaps also see others points of view , i can say that while my basic premise remains the same , everyone elses comments , casue me to add more information to my posts and to re evaluate my opinion as the discussion moves along its course, which is in fact the purpose of the discussion .

i post something knowing what my full point of view is , but in my inability to fully express that point of view or opinion in one post , it leads to peple replying , and then in turn casues me to more fully express what i am trying to say or to clarify it . and occasionally even to change my opinion . ( as stubborn as i can be it does happen ).


my point all along was about the actual situation not just the math but it took lots of people replying , before i was actually able to expreess it all , and clarify it , it is so easy to assume peole will understand exactly what you mean , when the reality is everyone will understand it a little bit differently .

over the years i have learned to get better at comunicating in here and try harder to express myself correctly , though on occasion i too can be a smart ass .

not letting things escalate beyond a smart ass comment here or there is important to any real discussion .

thanks btw for all the math help i certainly could not have ecpressed it all , or any where near as well as you did

guardian

_________________
im guardian who the f... are you !.
View user's profileSend private messageSend e-mail
Woody
Grandmaster Poster
Grandmaster Poster


Joined: Nov 01, 2005
Posts: 989
Location: Lake Powell

PostPosted: Sun Nov 05, 2006 9:30 pm Reply with quoteBack to top

guardian wrote:
woody

perhaps not , but doing nothing certainly is not going to make things better in the basic arena's . some change to get better rollups through the arena rather than through buying team sheets , is bound to have some effect , even if it only means that managers become more active for tourney prep in those arena's instead of just 82 .

the point of a discussion is to bring out all the ideas and to perhaps also see others points of view , i can say that while my basic premise remains the same , everyone elses comments , casue me to add more information to my posts and to re evaluate my opinion as the discussion moves along its course, which is in fact the purpose of the discussion .

i post something knowing what my full point of view is , but in my inability to fully express that point of view or opinion in one post , it leads to peple replying , and then in turn casues me to more fully express what i am trying to say or to clarify it . and occasionally even to change my opinion . ( as stubborn as i can be it does happen ).


my point all along was about the actual situation not just the math but it took lots of people replying , before i was actually able to expreess it all , and clarify it , it is so easy to assume peole will understand exactly what you mean , when the reality is everyone will understand it a little bit differently .

over the years i have learned to get better at comunicating in here and try harder to express myself correctly , though on occasion i too can be a smart ass .

not letting things escalate beyond a smart ass comment here or there is important to any real discussion .

thanks btw for all the math help i certainly could not have ecpressed it all , or any where near as well as you did

guardian


guardian

I agree wholeheartedly that some change will help.
I think your idea on break points not matching distributions is a very good insight.

I just don't want anyone to have unrealistic expectations and get disappointed.

Sorry about the intensity some of my posts, I get worked up sometimes. As far as the general "energy" that goes along with some of these strings, I think it is a good thing. The energy, the argument, the back-and-forth; it helps the process along.

I just hope everyone can understand that an intense debate, where emotions (and emoticons) sometimes get out-of-control, doesn't have to result in permanent grudges.

In general: I apologize to any whom I may have offended in this (and related) strings.
To guardian: Sorry about the earlier post (now edited). It went too far.

...here's a mental image of how some of these debates go: imagine two PR's in full fencing gear gracefully dueling...gradually it devolves to the point where they're grabbing improvised weapons and bash modifying...

Anyway, I think this string has been helpful and I'm glad you kept the issue moving.

_________________
...
Not enough time on my hands anymore...
...
View user's profileSend private message
pipthetroll
Advanced Master Poster
Advanced Master Poster


Joined: Nov 04, 2002
Posts: 447
Location: In my underwear, in front of my computer

PostPosted: Sun Nov 05, 2006 11:26 pm Reply with quoteBack to top

Woody wrote:

For the simplicity of the math and descriptions, I have made some generalizations and assumptions. They are:

-The 50% best (subjective) rollups in the pool are "good" rollups.
-Redundant draws for the same manager are ignored (sampling without replacement).
-The DA draws are normally distributed.
-The DA pool is relatively large (a few thousand).


Your assumptions are wrong.
1.Nowhere near 50% are even decent. 15-20% of the DA replacements will not be able to achieve 15 wit or wil after adding.
2.The DA pool isnt a pool, its random. There are hundreds of thousands of possible replacements, if not millions. Effectively, your chances of getting the same replacement are very slim.
View user's profileSend private messageSend e-mailICQ Number
Maximillian
Grandmaster Poster
Grandmaster Poster


Joined: Jun 24, 2002
Posts: 964
Location: Kelowna, BC CAN

PostPosted: Mon Nov 06, 2006 6:05 am Reply with quoteBack to top

pipthetroll wrote:
Woody wrote:

For the simplicity of the math and descriptions, I have made some generalizations and assumptions. They are:

-The 50% best (subjective) rollups in the pool are "good" rollups.
-Redundant draws for the same manager are ignored (sampling without replacement).
-The DA draws are normally distributed.
-The DA pool is relatively large (a few thousand).


Your assumptions are wrong.
1.Nowhere near 50% are even decent. 15-20% of the DA replacements will not be able to achieve 15 wit or wil after adding.
2.The DA pool isnt a pool, its random. There are hundreds of thousands of possible replacements, if not millions. Effectively, your chances of getting the same replacement are very slim.



I thought I'd heard somewhere that RSI only had around a thousand possible warrior designs.

Anyone know how many different RU's it possible to get?

_________________
Maximillian, DJI

"Only the dead have seen the end of war."- Plato
View user's profileSend private messageSend e-mail
Managerr
ArchMaster Poster
ArchMaster Poster


Joined: Jul 12, 2002
Posts: 4059
Location: Omaha

PostPosted: Mon Nov 06, 2006 7:10 am Reply with quoteBack to top

There are 856'ish team RU sheets.
View user's profileSend private messageSend e-mailAIM AddressICQ Number
Woody
Grandmaster Poster
Grandmaster Poster


Joined: Nov 01, 2005
Posts: 989
Location: Lake Powell

PostPosted: Mon Nov 06, 2006 7:25 am Reply with quoteBack to top

pipthetroll wrote:
Woody wrote:

For the simplicity of the math and descriptions, I have made some generalizations and assumptions. They are:

-The 50% best (subjective) rollups in the pool are "good" rollups.
-Redundant draws for the same manager are ignored (sampling without replacement).
-The DA draws are normally distributed.
-The DA pool is relatively large (a few thousand).


Your assumptions are wrong.
1.Nowhere near 50% are even decent. 15-20% of the DA replacements will not be able to achieve 15 wit or wil after adding.
2.The DA pool isnt a pool, its random. There are hundreds of thousands of possible replacements, if not millions. Effectively, your chances of getting the same replacement are very slim.


-1. The "good" rollup assumption was to communicate the results more easily, and, it was "subjective".

-2. If the pool is random, the redundant draw and large pool assumptions are no longer necessary, and the math still comes out the same.

_________________
...
Not enough time on my hands anymore...
...
View user's profileSend private message
Display posts from previous:      
Post new topicReply to topic
View previous topic Log in to check your private messages View next topic
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum



Powered by phpBB 2.0.10 © 2001 phpBB Group

Version 2.0.6 of PHP-Nuke Port by Tom Nitzschner © 2002 www.toms-home.com
Forums ©
:: fisubsilver shadow phpbb2 style by Daz :: PHP-Nuke theme by coldblooded (www.nukemods.com) ::