Poll |
Would you like playing Grey Box games? |
Yes, it would be a challenge playing. |
|
17% |
[ 7 ] |
No, I think I have enough options for playing even games (Leagues, Ranked, etc) |
|
74% |
[ 29 ] |
Other (see below) |
|
7% |
[ 3 ] |
|
Total Votes : 39 |
|
neoliminal
Joined: Aug 02, 2003
|
  Posted:
Jan 29, 2013 - 19:42 |
|
I was contemplating a new kind of Black Box system, for which I've coined the term "Grey Box".
Currently you are paired against a random opponent that is close to you in Team Value. It does not appear that they system takes into account coach rating at all. I could be wrong, as I don't think the secret sauce of this system has been disclosed.
However;
What if there was a Grey Box and it took the following variables into account when making pairings:
• Team Value (with assumption of no inducements)
• Coach Rating with that race.
• Overall Win/Loss rating of the individual team (more recent games weighted more heavily)
The result would be an even game between two coaches, even if the teams and coaches were on different levels. For example a coach with a Chaos team and coach rating of 165 and Team Value of 1250 might face off against a Skaven coach rated 145 and Team Value of 1500. The game would be a fairly even game. (Don't get caught up in the statistics of the example numbers, they are there for explanation. Real Database calculations would be needed to find the right pairings.)
Would you like playing in that kind of game with that kind of challenge? |
_________________ *
<BBRC>retired</BBRC><NAF>founder, 1st, 2nd Presidents</NAF><BB-Developers>retired</BB-Developers><Game-Developer>active</Game-Developer> |
|
Qaz
Joined: Apr 28, 2004
|
  Posted:
Jan 29, 2013 - 19:48 |
|
What I would not like is a community who is about half the size it was in 2008 (played games wise) split up into even more divisions. |
_________________ Superstition brings bad luck.
"he who has relied least on fortune is established
the strongest"
Niccolo Machiavelli |
|
Azure
Joined: Jan 30, 2007
|
  Posted:
Jan 29, 2013 - 19:52 |
|
Err...so if I understand, this means people that lose a lot and have a low CR (like the noob clawpomb bashers in the box) would get easier games giving them more and more of an advantage so that they can just bash bash bash - but still the game would be tilted enough to still give them a chance to win...I am confused here - how is this better? |
|
|
Woodstock
Joined: Dec 11, 2004
|
  Posted:
Jan 29, 2013 - 19:52 |
|
Coach rating was removed from the formula for a reason, most people didn't want it there. |
|
|
Hitonagashi
Joined: Apr 09, 2006
|
  Posted:
Jan 29, 2013 - 19:58 |
|
Azure wrote: | Err...so if I understand, this means people that lose a lot and have a low CR (like the noob clawpomb bashers in the box) would get easier games giving them more and more of an advantage so that they can just bash bash bash - but still the game would be tilted enough to still give them a chance to win...I am confused here - how is this better? |
+1.
As Woodstock says, this is the way the Box started, and nobody liked it, so it was changed to how it is now. |
_________________ http://www.calculateyour.tv - an easy way to work out specific team builds.
|
|
pythrr
Joined: Mar 07, 2006
|
  Posted:
Jan 29, 2013 - 20:26 |
|
No. |
_________________
|
|
Calcium
Joined: Apr 08, 2007
|
  Posted:
Jan 29, 2013 - 20:29 |
|
good effort, but ultimately flawed. Until the rules are sorted no amount of tinkering with the scheduling parameters in the box will make any difference.
Or put another way....LOL...weakling |
_________________
|
|
uuni
Joined: Mar 12, 2010
|
  Posted:
Jan 29, 2013 - 20:53 |
|
People do not really want 50:50 forced, but instead they want to win less skilled opponents with perceived equality over TV.
Personally, my favorite proposition is pairing by FF, which gets similar results with reasonably wide random variance. It also removes the noob-jacking flavor of minmaxing.
Some recent random events have me guessing, that the box pairing formula is not frozen, but Christer may make some tweaks here and there. For example, I have witnessed some pairings outside of the 15% TV difference, which was supposed to be a hard limit. Still, I may have a wrong hunch, as I have really no hard evidence of that. |
|
|
Wreckage
Joined: Aug 15, 2004
|
  Posted:
Jan 29, 2013 - 21:27 |
|
uuni wrote: |
Personally, my favorite proposition is pairing by FF, which gets similar results with reasonably wide random variance. It also removes the noob-jacking flavor of minmaxing. |
Isn't the presence of FF itself an inbuild match making system?
I'm not sure it's wise to go any further than that. I have been monitoring my FF development on some teams. And while it reliably drops if you lose and if you tie, you can easily play 10-12 wins in a row without ever hitting FF 10.
On the other hand your FF may just increase to 15 in the same time.
In the end it is an even more fleeting thing than CR and I wouldn't dare to tell anything based on it. |
|
|
The_Murker
Joined: Jan 30, 2011
|
  Posted:
Jan 29, 2013 - 21:40 |
|
I picked Christer's brain on the formulas for pairing and suitability a few months ago on irc and he answered several questions. If you will allow me to indulge a bit and paraphrase..
He wasn't exactly sure of what the formula's still were, as he hadn't looked at them in a while, and I got the feeling he didn't really care to at that time, or in the near future. His mind was focused on other areas. He was content that the Box was performing quite well, was popular enough, and was reasonably happy with it for the time being. So I wouldn't hope for any changes to the system anytime soon. I have my own pretty well thought out ideas for Box, but after that conversation I put my ideas and arguments on the shelf for a while to await my moment at a later date.
What he didn't say was that he would never change anything. Also notable, he said he monitors forum traffic and hot topics periodically, so be sure he knows what way the wind is blowing most days.
But nothing is going to change until he feels like changing it, and wants to spend the very considerable amount of free time required changing it. And I don't think that is now, with Kickstarter in full swing.
Maybe a fun topic to mull over for some.. but I'm going to save my energy for another day or another subject.
As for the forum opinions you will get... yea.. more choice than you have already would be bad. If it ain't broke don't fix it. Don't pee into the wind. And one other I can't think of right now.
Cheers. |
_________________
Join the wait-list. Watch the action. Leave the Empire. Come to Bretonnia! |
|
pythrr
Joined: Mar 07, 2006
|
  Posted:
Jan 29, 2013 - 22:01 |
|
uuni wrote: | People do not really want 50:50 forced, but instead they want to win less skilled opponents with perceived equality over TV.
|
With all due respect, this is bollocks. sir.
What "people" want is an even playing field where skill (and by this i mean both playing skill and team building skill) are the basis for wins and losses, with the added randomness of dice (this being a dice game, after all). What they do not want is some kind of handicapping system that patronizes people who are less skillful, thus infantilizing them and giving them no incentive to improve. |
_________________
|
|
Pastorn
Joined: Jun 13, 2004
|
  Posted:
Jan 29, 2013 - 22:53 |
|
I must say that´s "intertesting" to read this kind of threads. The formula is the same every time (almost). Step one: inspired guy come up with an idea. Step two: several "knowers" chaisaws the idea into pieces. Step 3: Thread dies out. Step 4 (hopefully not): Inspired guy, stops being inspired.
In my world a community is the sum of all its members thoughts, ideas and engagment. What we have is a result of inspired guys crazy ideas and descisions to spend time "here" instead of doing "important" stuff IRL.
If the community should outlive us all, there must be some possibilities for new (younger) guys to develop and maintain this community. I think all ideas are great - not saying I think the ideas always are great, if you can see the distiction.
But the game shouldnt either be a crazy mix of every ones crazy ideas - then it will be impossible to steer and keep playable. But maybe there should be some "playground" for testing all sort of things. If I and 10 other guys want to play games of some sort of style - maybe the community should have some space for us to test it?
The league games are some sort of playground; awfull skill league, 7vs7, 3vs3 etc. But maybe there should be some sort of open code for inspired guys and gals to test things. Maybe this kind of open codes can make it possible to change roosters, additional skills etc.
Ex. In our league every players start with dirty player and no ref. We want to play games there you can have elves and dwarfs in the same team or skaven and beastmen. All big guys teams. etc. Maybe there will come great ideas out of this, ideas that could be used in the "real" games.
My point is: "if it isnt broke - dont try to fix it" isnt an argument if we want the game to envolve and develop. If this coach wants to try some "grey box" let there be possibilites so create a league with this options.
I have respect for that there could be some technical issues, that im not aware of that make some of these things not doable - but the answer should never be - bad idea. It should be: great! Try it out and come back and tell us how it went. |
|
|
Panda_
Joined: Jul 14, 2004
|
  Posted:
Jan 29, 2013 - 23:01 |
|
Not that thread again |
_________________ "Rien ne sert de partir a point, il vaut mieux courir." |
|
Woodstock
Joined: Dec 11, 2004
|
  Posted:
Jan 29, 2013 - 23:20 |
|
Pastorn, you couldn't be more wrong... There is no room to test things. We have limited resources and should be critical about every thing suggested. So if we can identify the flaws of suggestions early on, we can move on to some thing that might be worth it. All about being realistic and efficient. |
|
|
uuni
Joined: Mar 12, 2010
|
  Posted:
Jan 30, 2013 - 01:21 |
|
@Pastorn: I think I remember that League custom rosters and custom rules are on Christer's todo list, but that is a long and ambitious list indeed, as we all know. At least if I remember some chat in #fumbbl correct.
@pythrr: Thank you for your opinion, I respect that. You may actually know the situation better than I do, as I have only second hand knowledge from these forums about the CR-based pairing. I understand it was widely loathed, it seems.
Still, there seems to be an inherent conflict between the notions of trying to attain balanced pairings (15% rule etc) and not to attain perfect pairings (50:50, CR-base pairings).
I think this is not an unique phenomenon, as it has been said that the main design goal of inducements was to make the match more balanced but retain overdog advantaged position (transform the odds to -> [2:1,1:1[ in math notation).
Perhaps a compromise could be found in somewhere? Something like including CR into the suitability equation but also relax the hard limits and perhaps even prefer a little lopsided matches such as odds 5:4 or something over 1:1.
Oh, and thank you The_Murker for your interesting insights!
And for new guys: Don't get depressed, things change and both you and the site change, it just takes time... |
|
|
|
| |