44 coaches online • Server time: 16:46
Forum Chat
Log in
Recent Forum Topics goto Post Creating a custom to...goto Post Secret League Americ...goto Post DOTP Season 4
SearchSearch 
Post new topic   Reply to topic
View previous topic Log in to check your private messages View next topic
cdassak



Joined: Oct 23, 2013

Post   Posted: Oct 22, 2014 - 13:10 Reply with quote Back to top

Fair enough, although I’d use the saw on thralls and/or fouls and attack the vamps with the tackler.
In any case, the total turns (163 to his 158) and total cas (2 to his 4) are a abnormal, given you have 4 MB and he had none. Your only chance was to win heavily the numbers game which you didn’t.
So I think the result was determined more by dice than by the match up.

_________________
Image
Garion



Joined: Aug 19, 2009

Post   Posted: Oct 22, 2014 - 13:17 Reply with quote Back to top

I disagree, I know I was unlucky in that match up cas wise but I never rely on cas basically because I never get them, I win through positional play, but I also know that there was no way I could stop his vamps, I couldn't even hit them with tackle as they were st5, so there was no chance to hunt them, also chainsaws are rubbish especially av7 ones, av9 one are ok.

The point of my post on the previous page was simply that the match up its self on paper was an impossible one to win because his team was soooooooooo much more developed than mine. With +stats and blodge, which if you do not have the tools to deal with leaves you hopeless.

cdassak wrote:
Your only chance was to win heavily the numbers game which you didn’t.

Relying on blind luck of cas dice does not sound like much fun to me. I'd rather a fair match up.

_________________
Image
cdassak



Joined: Oct 23, 2013

Post   Posted: Oct 22, 2014 - 13:23 Reply with quote Back to top

We agree that we disagree then Smile

_________________
Image
Christer



Joined: Aug 02, 2003

Post   Posted: Oct 22, 2014 - 13:25
FUMBBL Staff
Reply with quote Back to top

The match in question was given a suitability of @699, which isn't great.
The other options found in that round were:

Dwarfs, 1580 TV, @590
Pact, 1280 TV, @529

The latter is low due to the rookie protection factor (8/15 or roughly 53%, without it, the suitability would be near max (ie, @999 or @1000). Given that you were out of the rookie protection zone, the higher team was assigned a better suitability.

Is the blackbox scheduler perfect? Not by any means, and I'm certain there are improvements that are possible. Would the pact team be a more reasonable opponent for you in this case? Perhaps, but it's very hard for the scheduler to know that for sure. The whole point of the rookie protection aspect is to discourage minmaxing, which was pretty significant at this particular TV.

I'm open to suggestions and improvements on the rookie protection formula if someone is up for analyzing it and doing a well thought-out writeup on it.
Balle2000



Joined: Sep 25, 2008

Post   Posted: Oct 22, 2014 - 13:36 Reply with quote Back to top

koadah wrote:
Christer's idea seems simpler. As I understand it, the CPOMBers play each other more often the other play CPOMBers less often.


Or alternatively, instead of making it race specific, link it to a percentage threshold of total games played in the division lately.

E.g. a suitability rating is added equal to a race's excessive game share over the last X days/games.

This way, if, say Chaos' game share over the last X games/days drops below the threshold, they will "resume normal rotation".

This way there's a clear incentive to play other races (adding variation), while keeping it optional what race you play; and it doesn't doom a set of races to be permanent outcasts in the box.

And most importantly, the goal: lesser played races will more often be paired more with the other races keeping below the threshold.


Last edited by Balle2000 on Oct 22, 2014 - 13:42; edited 1 time in total
Garion



Joined: Aug 19, 2009

Post   Posted: Oct 22, 2014 - 13:37 Reply with quote Back to top

Interesting. I agree with the rookie protection side of the formula as that was B's biggest problem, and it is hard to tell from a forum post the tone in which it is intended, just to clarify I wasn't having a whinge on there. I was just trying to say why not have the best from both worlds?

So the old system worked pretty well apart from the rookie protection side of things, while the new scheduler creates the aforementioned match up issues. Why not use the 15% one we used to have but introduce a rookie protection forumla to that. As I said on the previous page, something like First 30 games you cannot play a team that has played twice as many games or more than your team.

I should say, I do still prefer the current one compared to the old system, because rookie bashing was awful. I just think there could be a happy medium here somewhere.

_________________
Image
Christer



Joined: Aug 02, 2003

Post   Posted: Oct 22, 2014 - 13:38
FUMBBL Staff
Reply with quote Back to top

If I end up implementing a diversity factor, it'll be race independant and be based on the actual distribution of races in the last X number of days. Thus, it'd do the same for elves should they end up being dominant.
Christer



Joined: Aug 02, 2003

Post   Posted: Oct 22, 2014 - 13:44
FUMBBL Staff
Reply with quote Back to top

Also, for a bit of insight into the scheduler for that particular round, you can check out the schedules it considered on the Blackbox Scheduler History page.
Garion



Joined: Aug 19, 2009

Post   Posted: Oct 22, 2014 - 14:00 Reply with quote Back to top

sorry if this has been asked before, but I couldn't see anything in this thread, is the scheduler forumla that is currently being used available for us to look at somewhere?

_________________
Image
Balle2000



Joined: Sep 25, 2008

Post   Posted: Oct 22, 2014 - 14:03 Reply with quote Back to top

Christer wrote:
If I end up implementing a diversity factor, it'll be race independant and be based on the actual distribution of races in the last X number of days. Thus, it'd do the same for elves should they end up being dominant.

For what it's worth (and if it wasn't obvious already), I think this is a clever idea that is likely to have a positive effect on blackbox. Would love to see this implemented.
xnoelx



Joined: Jun 05, 2012

Post   Posted: Oct 22, 2014 - 14:21 Reply with quote Back to top

Garion wrote:
sorry if this has been asked before, but I couldn't see anything in this thread, is the scheduler forumla that is currently being used available for us to look at somewhere?


https://fumbbl.com/p/notes?op=view&id=522

(from the 'How does the scheduler work?' link on the activation page)

_________________
Image Nerf Ball 2014
koadah



Joined: Mar 30, 2005

Post   Posted: Oct 23, 2014 - 00:06 Reply with quote Back to top

The issue I see is that the more you mess with it the more big aTV gaps you get.

You need more coaches for any of this to work

_________________
Image
O[L]C 2016 Swiss! - April ---- All Stars - Anniversary Bowl - Teams of Stars - 13th March
mrt1212



Joined: Feb 26, 2013

Post   Posted: Oct 23, 2014 - 00:08 Reply with quote Back to top

Christer wrote:
Also, for a bit of insight into the scheduler for that particular round, you can check out the schedules it considered on the Blackbox Scheduler History page.


fwiw, i feel like the scheduler history should show when there are enough coaches (4+) but no matches were made, for whatever reason.
nufflehatesme



Joined: Nov 02, 2011

Post   Posted: Oct 23, 2014 - 01:32 Reply with quote Back to top

looking at the graphs posted by balle, my conclusion is there is a soft cap on teams at 1700-1800tv. the big 3 only make up 35% of games at 1700, then accelerate to 60%+ by 2000. my theory is the rule of 5, and base cost of rosters has a large part to play in this.

chaos, nurgle, and chaos dwarf roters are 100k+ more expensive than most other rosters, making their cap 100k+ higher than everyone else.

i.e. 13 man chaos team
4 CW = 400k
9 beastman = 540k
3 rerolls = 180k
1 apo = 50k
FF 7 = 70k
total = 1240k
add 5 superstars with normal skills and you are at 1740k, with still plenty of room to grow with stats, doubles, ball handler and skills on rookies.

human 13 man roster
4 blitzers = 360k
1 catcher = 70k
1 thrower = 70k
7 linesmen = 350k
3 rerolls = 150k
1 apo = 50k
7 ff = 70k
total = 1120k

by default and no other factors included -
if you are a disciple of the rule of 5, the big 3 will find it easier to get over 1800 and stay there than others. pretty simple really.

edit: elves are a bit of an outlier, but struggle to run with more than 11 players for extended periods
koadah



Joined: Mar 30, 2005

Post   Posted: Oct 23, 2014 - 06:05 Reply with quote Back to top

Maybe disciple is the wrong word.

You get the rule of five because you cannot keep the grunts alive
Weaker cpomb and fewer cpombers would help

_________________
Image
O[L]C 2016 Swiss! - April ---- All Stars - Anniversary Bowl - Teams of Stars - 13th March
Display posts from previous:     
 Jump to:   
All times are GMT + 1 Hour
Post new topic   Reply to topic
View previous topic Log in to check your private messages View next topic