RDaneel
Joined: Feb 24, 2023
|
  Posted:
Aug 18, 2023 - 23:43 |
 
|
Hell All
This evening i have activated several teams from TV 900 till TV 1300 and among them I have one team zero match played TV 995.
Then the scheduler run, there was several coach activating ( I think we were 4 or 6) and decision was to pair one of my Team 900 TV zero match played to an opposite team TV 995 zero match played.
For the sake of completeness I will tell you that my team TV 900 were Dwarfs. My team TV 995 were Undead. And my opponent's Team TV 995 were Humans. All zero match played.
My opponent rank was 1390 ( Rookie) where my Rank was 1618 (SuperStar overall... undeserved perhaps  )
The more reasonable choice would have been TV 995 vs TV 995 zero match played
So seems there are situation where TV gap is not the main decision factor of the algorithm
Looking in the old forum i founded this thread
https://fumbbl.com/p/notes?op=view&id=522
if this is still valid the only thing come to my mind is linked to the Ranking: because the other coach was a "rookie" and thus triggers the "rookie protection factor"...so the TV 995 vs TV995 association was judged unfair and algorithm decided to go for the next team with TV less still with 10% difference (and 90 is exactly the maximum 10% allowed for TV 900 team with zero match played)
is my assumption correct?
Any comments is welcome!
thx! |
|
|
Christer
Joined: Aug 02, 2003
|
|
RDaneel
Joined: Feb 24, 2023
|
| Posted:
Aug 19, 2023 - 08:34 |
|
|
Sp00keh
Joined: Dec 06, 2011
|
| Posted:
Aug 19, 2023 - 12:38 |
|
IIRC the rookie protection was for new teams, not for rookie coaches |
|
|
RDaneel
Joined: Feb 24, 2023
|
| Posted:
Aug 19, 2023 - 15:08 |
|
| Sp00keh wrote: | | IIRC the rookie protection was for new teams, not for rookie coaches |
Ok
Both team were new (zero games played) . |
|
|
intrepidwilson
Joined: Feb 09, 2020
|
| Posted:
Jan 27, 2024 - 00:23 |
|
| Sp00keh wrote: | | IIRC the rookie protection was for new teams, not for rookie coaches |
This, IMHO, is the issue. There should (ideally) be a differentiation between rookie teams and rookie coaches. I've an undeveloped Human team and, after 2 Black Box games, they're a busted flush for the next match and I've just desperately rolled for skills to give them some fighting chance.
I have been in 145 but found it was a) hard to find equally incompetent coaches to play against and b) less engaging than a cup / league with clear fixtures and deadlines. |
|
|
The_Murker
Joined: Jan 30, 2011
|
| Posted:
Jan 27, 2024 - 04:32 |
|
I don't know how to start on Christer's link, but once upon a time, in Black Box pairing, each team entered into the draw was given a 'score' of some sort, based loosely on TV, and other things.. one of these other things being a very small random factor.
ALL the possible matches between ALL the possible teams were assessed for fairness, by comparing each team's 'score', looking for similar scores. If this random factor still exists.. this would explain why equal TV is not always paired with another equal TV.
Additionally, the pairing code tried to make as many 'good' matches as possible.. meaning sometimes one of the matches might be slightly less 'even' to make a total box draw with 3 or for 'good' games. |
_________________
Join the wait-list. Watch the action. Leave the Empire. Come to Bretonnia! |
|
Joost
Joined: Mar 17, 2014
|
| Posted:
Jan 27, 2024 - 10:32 |
|
I asked a similar question not long ago, and Christer explained that any team in the box under 1000 TV counts as 1000 TV for matchmaking purposes. So to the box algorithm your TV 900 team is exactly the same as your TV 995 team. There is a max TV gap that is taken into account though and for that your team real TV is taken. |
|
|
MattDakka
Joined: Oct 09, 2007
|
| Posted:
Jan 27, 2024 - 11:33 |
|
|
MerryZ
Joined: Nov 28, 2005
|
| Posted:
Jan 27, 2024 - 13:16 |
|
TV cap should be removed from box if more than 15 games played so you can spin old stuntty teams and big teams there also. |
_________________
Kaptain Awasoam, Dicer of All Men and Women and Children and Puppies. |
|
|
|
Automatic Time Out O...
Search
