uuni
Joined: Mar 12, 2010
|
  Posted:
Dec 14, 2011 - 02:00 |
|
What if the box would take Fan Factor in account as a main factor in calculating the pairing preferences?
I think that Fan Factor is a quite good, but lenient, measure of two things: 1) team's age in games and 2) teams general success.
My feeling is that with factoring the Fan Factor into the pairing points, teams would get more games against similar FF teams than against other teams. I perceive there to be about 4 different team classes: Beginners (0 FF), Less-than-averages (6 FF), Averages (8 FF), Winners (10 FF). Greatest difference would be between the Beginners and everyone else.
It is also noteworthy, that FF can only be manipulated by not playing to win (which is against the site rules) and it is not possible to get back to 0 after it once has risen up from there. I feel this could dissuade n00b-bashing, even though minmaxing would not lose its meaning.
I feel that some time has gone by since the last Blackbox pairing talk, so perhaps we can ponder it again. Of course, as always, any actual changes would be in the exclusive territory of Christer and not us forum dwellers.
What do you think about this kind of system? |
|
|
uuni
Joined: Mar 12, 2010
|
  Posted:
Dec 14, 2011 - 11:22 |
|
Ah the fifth FF group would apparently on FUMBBL be the Grandfathered (16 FF)! |
|
|
JimmyFantastic
Joined: Feb 06, 2007
|
  Posted:
Dec 14, 2011 - 11:27 |
|
Very bad. Bigger TV differences makes for more unfair games so lets keep it TV matched. |
_________________ Pull down the veil - actively bad for the hobby! |
|
uuni
Joined: Mar 12, 2010
|
  Posted:
Dec 14, 2011 - 11:40 |
|
Would you like to elaborate more on your point, JimmyFantastic? I do not remember mentioning TV in any point of my posts. I asked if the box would take Fan Factor in account.
I thank you for your contribution and I think it has a good point. In general, I would hope that we could discuss the heated TV discussion in other threads, if possible. |
|
|
JimmyFantastic
Joined: Feb 06, 2007
|
  Posted:
Dec 14, 2011 - 11:45 |
|
Well you said "take Fan Factor in account as a main factor" and the poll is "Would you like Box pairing by Fan Factor?".
This implies that TV would be a lesser consideration and more unfair games would result.
Sure you could add FF to the equation that the box uses for suitability but if that's what you mean you have worded it very badly.
Adding it to the existing formua somehow could be pretty good as it would discourage min maxers vs starting teams. |
_________________ Pull down the veil - actively bad for the hobby! |
|
uuni
Joined: Mar 12, 2010
|
  Posted:
Dec 14, 2011 - 11:51 |
|
I apologize that my language was not better, JimmyFantastic. |
|
|
King_Ghidra
Joined: Sep 14, 2009
|
  Posted:
Dec 14, 2011 - 11:56 |
|
For me the biggest positive change to scheduler would be adding CR to the mix. |
|
|
Emeric
Joined: May 25, 2011
|
  Posted:
Dec 14, 2011 - 13:23 |
|
JimmyFantastic wrote: | Adding it to the existing formua somehow could be pretty good as it would discourage min maxers vs starting teams |
I fully agree with that. However, finding the good mix is another problem...
As i see it, preventing high FF to meet very low FF (<4) for rookie teams is good, but preventing a FF6 from meeting a FF11 team is against the spirit of competitive division. |
_________________ Fill the BOX grid !
Francophone? #fumbbl.fr est là pour vous ! |
|
Juff
Joined: May 24, 2006
|
  Posted:
Dec 23, 2011 - 08:32 |
|
FF essentially becomes a factor of a team's recent form (with minor element of randomness, as per all things BB). This system would tend to put recent winners against recent winners, and developed teams against developed teams (and vice versa for both).
I think such a system could be beneficial, but some folks would see winning a match and getting +FF as a bad thing, that 'penalises' them by making it more likely they play a winning coach/team next.
And as we all know, that would be terrible, the whole idea of [B] is to game the system to avoid even matches at all costs! |
|
|
divino77
Joined: Jun 23, 2011
|
  Posted:
Dec 23, 2011 - 09:02 |
|
Im sure this is not relevent to all but when i make teams i start them with higher FF, for me winning kick off rolls and earning more money earlier on out weighs other factors for fresh teams IMO, also being a rookie coach i have had teams start with 9 FF that are now only only on 5 after 29 matches so i think its not so clean cut to say fresh teams will have low tv and higher TV teams higher ff |
_________________ Nothing is only something if it is contained otherwise it is eveything. |
|
Wreckage
Joined: Aug 15, 2004
|
  Posted:
Dec 23, 2011 - 09:22 |
|
high FF is already a disadvantage in itself. And people who start with a high FF shouldn't be punished either. Apart from that it raises and falls pretty randomly at a certain size. DOn'T really see much point in using such a random value as measurer. You could aswell go with the number of guarders on the team. Or the number of rerolls. Holds absolutely no particular meaning. |
|
|
sann0638
Joined: Aug 09, 2010
|
  Posted:
Dec 23, 2011 - 09:57 |
|
divino77 wrote: | Im sure this is not relevent to all but when i make teams i start them with higher FF, for me winning kick off rolls and earning more money earlier on out weighs other factors for fresh teams IMO, also being a rookie coach i have had teams start with 9 FF that are now only only on 5 after 29 matches so i think its not so clean cut to say fresh teams will have low tv and higher TV teams higher ff |
Erm, that's probably not the best strategy though. And I think Wreckage is being a bit harsh - winning teams do generally have higher FF than losing ones. If Ranked didn't exist and people needed a quick way of finding opponents and didn't care too much about TV differences then this would be pretty good, imo. |
|
|
Wreckage
Joined: Aug 15, 2004
|
  Posted:
Dec 23, 2011 - 10:26 |
|
So a team with a TV of 880 and 2300 are a better match because they have both FF6 then the other team at 1050 and 2200 wich both have ff9. |
|
|
uuni
Joined: Mar 12, 2010
|
  Posted:
Dec 23, 2011 - 10:45 |
|
I would again like to stress my original post:
"as a main factor" does not mean "use only".
***
One example that I could think of would be to use FF for the initial ordering of the teams in box pairings.
Currently box arranges teams to a random order and then starts to calculate the @/1000 numbers for each pairing. If a pair is found to not fit to the 15% TV, it is ignored. After calculating the sum of @ for a schedule, the team order is resuffled and another sum is calculated. After calculating a given amount of schedule proposals, the one with biggest @ sum is chosen. At least this is how it used to work.
If the round would be seeded from preselected order, such schedule would appear in the calculated list more often than pure random.
Of course, FF could be integrated to the pairing algo with some other way. What way do you guys suggest? |
|
|
Wreckage
Joined: Aug 15, 2004
|
  Posted:
Dec 23, 2011 - 11:02 |
|
I just don't get why you don't suggest win percentage or BWR or maybe some team dependend variant of BWR for your proposal. My previous comment still stands: High FF puts a team already at a disadvantage. Try to compete with a FF14 team and you will see what I mean.
If it was just for the won and the lost games FF wouldn't be very random but the tie games... mate... the tie games... they make the whole thing pretty wayne. |
|
|
|