80 coaches online • Server time: 14:39
Index Search Usergroups Profile
Log in
Recent Forum Topics goto Post Bb2020 Will you reti...goto Post Passing Madness!goto Post Help testing bb2020 ...
SearchSearch 
Post new topic   Reply to topic
View previous topic Log in to check your private messages View next topic
Poll
Are you ready to come full circle?
Yes! Everything old is new again!
26%
 26%  [ 8 ]
Probably, I'm ready for some change.
13%
 13%  [ 4 ]
Maybe... I don't want to lose too much.
3%
 3%  [ 1 ]
nO! cHaNgE iS bAd!!!
13%
 13%  [ 4 ]
I've been circling that pie like a hawk!
43%
 43%  [ 13 ]
Total Votes : 30


pizzamogul



Joined: Jun 13, 2005

Post   Posted: Dec 17, 2020 - 06:27 Reply with quote Back to top

Anyone been around long enough to remember this from... *checks notes*... 2004?!

AFK_Eagle wrote:
Here's the key concept to the league: you are awarded points for playing a "new" race. Each team will have records kept for which non-elven races are played. Once you play a given race, you are not given any further points for having played them until you've played ALL other races once. For example, your first game is vs a skaven opponent. You can play three consecutive games vs a skaven team, but points are awarded only for the first game. You do not get points for playing another skaven coach until after you've played chaos, orc, human, etc, at least once each. After all races played (yes, to include Ogres atm, subject to debate), all races become open once more for point play.


With BB2020 on the horizon I thought it would be an opportune time to do a rules review. There are a few things about the competitive balance of the league that I'd like to address and other issues to possibly correct with the benefit of hindsight. Some of the items on my agenda:

  1. Despite the efforts to arrange the Categories in a fair manner at the time, Box teams have a tough time completing the Cycle. With that plus the stuff below, it's been over 25 seasons since a Box team has claimed the Grand Championship.

  2. Allowing teams to replay matchups to improve scores benefits Ranked teams as they can focus on repeating a given race/category. It also favors a coach with the ability to play lots & lots of matches over a more time-constrained (but also talented) coach... something present in the league from its inception.

  3. The sliding scale of Baseline scoring favors high TV teams - one of those "hindsight" regrets I mentioned before. I had wanted to "slightly favor experience instead of encouraging serial team creation," but this again benefited teams playing in the open environment of Ranked.

  4. The point system really makes it beneficial to be able to pick opponents that can be beaten 5-0 for maximum score... again favoring an open environment. Also, a fact that leaves a bad taste in my mouth as commissioner.


So, where to go from here? Much depends on whether or not Match Reports will denote how a match was arranged (open/auto-scheduler/tournament) and/or the progress for the current FUMBBL Season of each team.

My current ideas for Cycles (addressing items 1 & 2):

  • Option A (if nothing changes regarding Match Reports). Limit a Cycle to a team's first run through of their 10 scoring matches. Ranked(Open) & League teams only count their first 2 matches in each of the 5 Categories. Box(Strict) teams are freed from the constraints of Categories and will count their first 10 games played against 10 different races.

    • Tweak #1. After a team has played its Cycle, it can begin a new one. Either allow the following Cycle(s) with better scores to replace the earlier ones, or make it so each Cycle counts and a single team will show up in the standings more than once.

    • Tweak #2. Allow a team to play unlimited matches during an E.L.F. season with its best string of 10 consecutive matches comprising a Cycle being the ones that count. For example, a Strict team plays 25 matches during the E.L.F. season. It's best consecutive score comes from games 7 through 16, so those are the matches that count towards its Cycle.

    • Tweak #3. As in Tweak #1, but each coach gets only one crack at the Cycle with each team each E.L.F. season. Now it's time to play a different team and we'll have an award by coach with the highest total points among all their teams... maybe only allow one score added from each race of elves to encourage playing all four.


  • Option B. If FUMBBL seasons wind up being 15 matches AND the progress is noted in Match Reports, make the Cycle = Season and truly come full circle. This would allow each team to score up to 15 regular-season matches PLUS end-of-season tournament matches. This would require longer E.L.F. seasons.

  • Possible wildcard for all above Options: If Match Reports tell when an Open team plays an auto-scheduled (or tournament) match, then any such matches follow the Strict freedom from Categories.
Note: A bit of this is stolen from my experience playing in the ARR! meta-group. Credit where credit is due.

Ways I'm thinking of altering the scoring (addressing items 3 & 4):

  • Baseline. The idea is to encourage Open teams to choose quality opponents. Strict teams could be exempt from some of the requirements.

    • I might just set this to 20 points for all matches. Maybe "slightly encourage" seasoned teams by offering a higher tier above like TV 1300 at 26 points or some such.

    • Another option (if the Match Report complies) might offer full points for choosing an opposing team with a winning record vs. partial points for a losing record.

    • Dedicated Fans/Gate could return to the mix to encourage coaches to choose opposing teams that are well run. Might add a bit of a wildcard too with the roll. I'd like this idea much more if the game designers stopped allowing fans to be bought and instead have them be earned.


  • Scoring/Casualties/Bonuses.

    • I'll probably remove the 5+ TD and Shutout bonuses.

    • No more double points for certain Categories too with the other changes to the Cycle.

    • Ideally, I'd like to weigh these amounts against the Baseline in such a way that choosing a quality opponent and eking out a 2-1 victory results in more points than can earned by blowing out a bad one.


What else?

The Categories will need some adjustment but it can be done with only Open in mind if Strict teams can "build their own cycle" so to speak. With Slann going away, D will need some more agility teams. Another Human race means E could be split up some. Suggestions for tweaking or a complete rebuild of the 5 Categories?

Feedback on these ideas and suggestions welcome.

_________________
"Don't expect mercy."
-Woodstock
policeshades



Joined: Oct 31, 2017

Post   Posted: Dec 17, 2020 - 09:40 Reply with quote Back to top

Thanks for posting this. Are you thinking of one big change and one change only? Or of testing an approach and modifying as the 'meta' and the new site changes become more familiar and their implications are more obvious?
Dalfort



Joined: Jun 23, 2008

Post   Posted: Dec 17, 2020 - 10:29 Reply with quote Back to top

When the new rules drop I believe it is Christer's plan the create a C (Competitive) Division removing Ranked and Box, I think then only League Division teams would be able to game this meta and they thus far have had a limited pool of opponents and notoriously appear low on the final lists.

Just my initial thoughts, you run E.L.F. fantastically and like the new rules themselves I am happy to accept the new group ones just as readily. Thank-you for all the time and effort you have lavished on FUMMBL's greatest META Group.

_________________
Image
The_Demon



Joined: Nov 18, 2012

Post   Posted: Dec 17, 2020 - 11:38 Reply with quote Back to top

If Box remains, they should be given some boost, given their poor recent record.

Your comment of a 2-1 win against a better opponent is a good idea, can this be linked to CR in some way perhaps (maybe a multiplying factor or additional points for playing "Better" coaches)?

I like the double points in some categories, those teams are more of a challenge, so I'd like to see that kept (or something equivalent) if possible.

But, will be interesting to play whatever rules/changes come in.

Thanks for running it - E.L.F. is Born to Live Forever!! Very Happy Laughing Cool
tussock



Joined: May 29, 2011

Post   Posted: Dec 17, 2020 - 12:09 Reply with quote Back to top

Ooh, considering just one natural 15-game season, but taking the "best" 10 from it?

And perhaps 1-3 per group count (preferring A, then B, then C, then D) for strict scheduler teams, so you're at least reasonably likely to have 10 games count.


If I was doing the roster groups it'd be ...

A: Elf killers: Chorfs, Dorfs, Lizards, Orcs, Shambling.
B: The Middle Road: Zons, Humans, Imps, Norse, OWA.
C: Claw Patrol: Chaos*, Regegades*, Necro*, Nurgle*.
D: Agility teams: Delfs, ElfUs*, HElfs*, Skaven, Woodies.
E: Fouls R Us: Black Orc*, Tomb Kings*, Underworld*, Vampires*, plus max one stunty.

* = official tier 2.

Sort of reasonably similar difficulty within groups and different builds and tactics work better against the different groups.

If there's bonus points, for TV, a lot of teams get much better with just a few skills, some DPs, couple guard, some block at least, a couple T/MB, I agree about 1400+ is solid, but there is a bit of a step again around 1600+ where teams are more complete.

Really though, use coach CR for a bonus, 160+, 165+, or even 170+ if you take on Malmir, those are worth a bit extra surely. That's leaving the forest!

Whatever you choose, just imagine someone doing waaaaaay too much of it to win. Is that a good thing, or a bad thing? Very Happy

--

Could accept Skaven and Underworld teams to E.L.F., leaving the shroom forests of the deeps, they face similar problems vs similar groups of teams, and playing bash is an adventure for them. Even can pass about as well as elves in BB20. Smile

_________________
ImageImage
razmus



Joined: Jun 23, 2017

Post   Posted: Dec 17, 2020 - 15:49 Reply with quote Back to top

Dalfort wrote:
When the new rules drop I believe it is Christer's plan the create a C (Competitive) Division removing Ranked and Box, I think then only League Division teams would be able to game this meta and they thus far have had a limited pool of opponents and notoriously appear low on the final lists.


I think this is relevant and probably of interest in this discussion...
Chister in the fumbbl_podcast channel on Discord wrote:
Let me clarify:
- Strict teams will be able to only play through the random scheduler, and not through the gamefinder
- Strict teams may "unflag" and switch to Open at any point
- Open teams will be able to play through both random scheduler and game finder.
- Open teams may not go back to Strict at this point

- Current Blackbox teams will automatically become Strict in the conversion
- Current Ranked teams will become Open

- The Blackbox Trophy equivalent will require new Strict teams, and while running the trophy, teams can not switch to Open


So really Box/Ranked will become Strict/Open ... with differences... and I assume Open teams still able to 'game the meta' as it were.
Nelphine



Joined: Apr 01, 2011

Post   Posted: Dec 17, 2020 - 15:55 Reply with quote Back to top

I'd rather not use CR - in particular, when looked at historically (as has been suggested before for E.L.F.) it didn't actually correspond well to difficult matchups.

The problem is that a high CR coach can be using a team they are bad with (for instance, the coach gets 170 CR from playing dark elves, and then against your E.L.F. team they played halflings). In the past (aside from a newbie team that bought fan factor), fan factor was much more reliable as an indicator for E.L.F. difficulty.

Personally, I'm not worried about newbie teams buying up high fan factor, but I'd agree with Pizza's other suggestion that there is a single boost from tv around 1300 or so. Newbie teams would not have that much tv, so even if they bought fan factor, it should already have started to normalize by the time they get to that tv. At the same time, its low enough that in the box, maintaining that tv should be REASONABLY reliable (even if not 100%).
Nelphine



Joined: Apr 01, 2011

Post   Posted: Dec 17, 2020 - 16:10 Reply with quote Back to top

Additionally, I'd keep the bonus for shut outs and 5 touchdowns (but I'd reduce it to 1 instead of 2).

Instead, I'd remove the points for touchdowns themselves (that's way bigger part of the problem of needing 5-0 wins in order to be competitive), as well as the points for causing casualties (this just leads to the highest scores coming from playing against the lowest av, particularly ones where the opposing coach has kept massive numbers of injuries for fluff reasons).
steinerp



Joined: Sep 18, 2005

Post   Posted: Dec 17, 2020 - 17:00 Reply with quote Back to top

This probably won't surprise anyone given my comments elsewhere but I really dislike any concept that interacts with seasons or limits games/team development. I do agree that playing poor opponents is rewarded. After one of my season wins, I did some analysis of my games compared to others that finished high and some box teams of MattDakka to see if I was cheery picking opponents. The results were mixed. Overall the average CR of my opponents was higher than that of the the box teams and most of the of the other ELF coaches. HOWEVER the average CR of the games that counted in the league was pretty comparable to the other ELF coaches and lower than that of Matt's. This makes sense to me as the scoring does select for lopsided wins which are more common vs poorer opponents. The same effect happens with box teams that play enough games.

To me the ELF league from Season 1 has been about encouraging elves to play everyone and not just other elves. It was the bravado of starting a 9 elves vs a healthy chaos dwarf team (https://fumbbl.com/p/match?id=1255194) or proving that 7 elves could defeat some rotters (https://fumbbl.com/p/match?id=1122494) <= TBF rotters only had 9 players instead of hiding and rebuilding. Box did not exist at the time but really fits into what the league was looking for. The league also has hints of story behind it. The challenge of guiding a team through the gauntlet of races is part of the team's history and team building aspects of the game. IMO BB2020 and Christer's proposed design of the [C] division discourage history and team building. I dislike that but others see it as a good thing. My suggests are designed to try to allow both. Finally the league was about not just playing elf ball.

Not related to the original structure of the league but in full disclosure. I will give [C] a try but I anticipate playing 100% league after seasons are implemented as I have 0 interest in non-development teams so I admit there is bias here in keeping League a viable part of the league.

Proposed structure.
Essentially the same structure (categories, 2 month seasons, etc) with the following changes:
1. Random (box-like) matchup teams ignore categories and just select 10 best games but only 2 games from Category A will be doubled in score.
2. No extra points based on team value (and as a result no category B bonus points)
3. The CR change is added to the game's score. (Not sure if CR is reported for league games)
4. If open League includes a box type scheduler maybe restrict to only box scheduled games (Note: this is dangerous as here in NA it is almost impossible to get a box game most nights).
5. Remove 5 TD bonus and replace with 5 casualties against bonus.
6. All casualties (for and against) earn you 1 league point. Kills count for 2 points instead of 1.
7. I don't know that I like this but move Ogres and Stunties into Category F. Those teams give up more points than the other Category E teams. Of the top teams for the last five seasons, 9 of them counted ogres/stunties, 5 played either ogres or stunties and 3 played neither. (Note: 2 of the ogre/stunties were actually ogres/Throne of Blood Underworlds and the last short season saw 1 of the neither and 2 of the either games which I think can be blamed on the short season)
pizzamogul



Joined: Jun 13, 2005

Post   Posted: Dec 18, 2020 - 04:20 Reply with quote Back to top

Thank you for the feedback everyone! I'll respond to bits of these initial posts.

policeshades wrote:
Are you thinking of one big change and one change only? Or of testing an approach and modifying as the 'meta' and the new site changes become more familiar and their implications are more obvious?

I'm wanting to set the general structure we will use for the next X years. I'm planning to play a bit and see how everything operates on FUMBBL post BB2020 before bringing the league out of its hiatus.

Dalfort wrote:
...Christer's plan the create a C (Competitive) Division...

Razmus shared how that will look in his follow up post.

The_Demon wrote:
If Box remains, they should be given some boost, given their poor recent record.

I'm hoping that freeing them from Categories paired with the elimination of replaying match ups might help.

The_Demon wrote:
Your comment of a 2-1 win against a better opponent is a good idea, can this be linked to CR in some way perhaps (maybe a multiplying factor or additional points for playing "Better" coaches)?

What CR level is displayed for potential opponents when in Gamefinder? Is that Racial or Overall?

The_Demon wrote:
I like the double points in some categories, those teams are more of a challenge, so I'd like to see that kept (or something equivalent) if possible

With the Strict Category changes, I'm thinking the multipliers will have to go away or it will result in an imbalance.

tussock wrote:
...one natural 15-game season, but taking the "best" 10 from it?

If an E.L.F. Cycle used a team's season, I'd count all 15 matches plus tournament. Strict teams would have all matches counted and Open teams would only have to worry about race variety in non-autoscheduled or tournament matches.

tussock wrote:
If I was doing the roster groups it'd be ...

What about truly Coming Full Circle and throwing out Elves as an option for E.L.F. opponents? What would 4 Categories look like? 3?

tussock wrote:
Really though, use coach CR for a bonus, 160+, 165+, or even 170+ if you take on Malmir, those are worth a bit extra surely.... just imagine someone doing waaaaaay too much of it to win.

I'd love to see Open coaches begin hunting for matches against Legends. I'd prefer to use the Racial CR though if only it appeared on the Match Reports.

razmus wrote:
...Chister in the fumbbl_podcast channel on Discord...

I'm also interested in knowing if the way a match is arranged will be noted in the Match Report? Otherwise we won't know when an Open team uses the Random Scheduler unless they pair with a Strict team. Can somebody politely pose that question on Discord?

Nelphine wrote:
Personally, I'm not worried about newbie teams buying up high fan factor...

Can a team purchase Dedicated Fans on redraft?

Nelphine wrote:
Additionally, I'd keep the bonus for shut outs and 5 touchdowns... remove the points for touchdowns themselves... as well as the points for causing casualties...

I feel that keeping only the bonuses would really fuel a drive to generate 5-0 match results in Open. In ARR! the only scoring that was done was a 4/2/1 W/D/L system over 16 matches, I've toyed with the idea before of doing away with TDs & Cas and just awarding points based on the "quality" of the win. Hmmm.

steinerp wrote:
I do agree that playing poor opponents is rewarded.

Part of the reason I'd like to reward playing a quality team with the Baseline and lower the value of scoring & casualties.

steinerp wrote:
...discourage history...

Hopefully participation in the E.L.F. helps to add to the story of a team as it winds through its seasons and will give coaches a reason to keep playing!

steinerp wrote:
Not sure if CR is reported for league games.

It is not. League team participation is more of a statement than a competition though. Also, Sea Elves make me laugh.

steinerp wrote:
...Ogres and Stunties...

Maybe the changes to fouling might make these matches more of a risk where picking is concerned...

Sounds like it might be a bit longer than anticipated before BB2020 is up and running on the site. Maybe I'll do some back testing of ideas to see what results might have looked like. Keep up the discussion!

_________________
"Don't expect mercy."
-Woodstock
Nelphine



Joined: Apr 01, 2011

Post   Posted: Dec 18, 2020 - 05:52 Reply with quote Back to top

On my particular note, that's why I reduced the value to 1. If you think about a match right now, a 5-0 win is worth 19 points. I'm suggesting to make it worth 12. If the bonuses are only one for each, then compared to the value of the win, they are largely nonexistent - so they would be used akin to tiebreakers.

Admittedly that could still encourage picking to try to earn those tiebreakers, so I'm also ok with the bonuses simply going away. But the bonus for touchdowns and cas are by far the worse culprit for those scores.
Java



Joined: Jan 27, 2018

Post   Posted: Dec 18, 2020 - 06:42 Reply with quote Back to top

If ranked teams racking up games and cycling through all races is a problem, you can steal a trick off Throne of Blood

Only one match, the highest scoring, is counted for every single race. Playing the same race again only gives you a chance to improve on your previous best score.

Too harsh?
Nelphine



Joined: Apr 01, 2011

Post   Posted: Dec 18, 2020 - 07:31 Reply with quote Back to top

No, the opposite - in ranked you can pick who to face to actively improve your score. In box you can't.
ArrestedDevelopment



Joined: Sep 14, 2015

Post   Posted: Dec 18, 2020 - 07:46 Reply with quote Back to top

Great to see some initiative here Smile

Some basic thoughts:
TV should really be somewhat removed/downgraded as a metric for points scoring given that outside of L no-rebuy leagues, the vast majority of teams will spend their lifetimes floating between 1350-1700tv past the first season. I actually thought it was somewhat backwards anyway - the current/old scheme rewards someone who plays a load of builder games then plays vs dwarves at high tv (where they may have all the tools to make the game easier) much more than someone who plays dorfs game 1 with a 0 apo, max 2 rr elf team with no skills, no bench etc.
Yes, it's nice to reward longevity and experience, but frankly it's complete bollocks, and very frustrating to potentially play a new ELF team in a random scheduler and find yourself getting pittance scores for hard fought wins vs extremely competent coaches using minmax bash at low tv where your team has nothing but raw stats to fight with.

One thing that has always stuck out firmly like a thorn about E.L.F. in some ways is this - the existence of the box/strict scheduler actually somewhat does what ELF is supposed to do. Obviously it's impossible for League teams to use the scheduler, so something would need to be worked out there, but is it not possible to simply make only games played via the scheduler/in tournaments in the C div count towards ELF scores for C-div teams?
It might seem a bit restricting, but it would put the entire div on even footing in terms of ease of finding games/completing the cycle/level of opponent in measured games. Open teams would of course be free to play builders, recoveries, elfball, whatever between games counting towards their ELF season and thus would still have several potential advantages, but the actual selection of games would be on the level.

I think Dedicated Fans is too low a variable to base an idea of how well teams perform as opposed to FF - it will normalise a lot more readily than FF did, giving less of an idea on actual performance levels.

And on races - actually i think no agile replacement for slann is necessary at all, the "agile" category was already overloaded in comparison to the 3 before it, and if humans & stunties were split in half (into, wait for it, a humans group and a stunty group) all the groups would be roughly equal. Although it may very well be a time to regroup or even re-tier some teams...

_________________
Image
razmus



Joined: Jun 23, 2017

Post   Posted: Dec 18, 2020 - 14:29 Reply with quote Back to top

ArrestedDevelopment wrote:
Obviously it's impossible for League teams to use the scheduler, so something would need to be worked out there, [..]


But maybe. There has been talk of using the scheduler for Secret League as well. I know it was mentioned once that Secret League E.L.F. games are tracked by the algorithm currently, but matches between known rosters in secret league matches are rare. Any chance all the secret league teams could be categorized? (I.e. add Avelorn and the other elves to being tracked... and put all the Secret League rosters into an A-E (A-F) category?)
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