Ok - I’ve jumped in and derailed the thread: now here is my attempt to re-right as well as contribute a bit of the discussion.
My understanding/Summary:
The core of the issue relates to @tua’s question:
Q: How do we have the most competitive playoffs possible?
An element of this is the identification of which teams are “best” suited to participate in playoffs. Mothership states the current system is a “compromise of both” referring to wins and rounds. This system (MP) should be emphasized as a COMPROMISE - this conclusion is also the essential result of the Swiss vs RR discussion (Thread and subsequent discussion) - Swiss isn’t ideal, but chopping divisions for RR can be worse (Or just too difficult).
@Cayorne in the OP states that he believes “W-L should be everything”.
In a latter post (Can we please get post #'s on these forums?) cayorne states that the current system is confusing to new players and the lack of ease in understanding the system can cause players to feel “wronged” or “punished for…winning matches and not rounds”.
We deal with two examples in this thread:
- A (theoretical?) example in OP between team A and B - this dilemma is well summarized by mothership:
"Team A had 2 rolls and 5 close games. Team B had 5 close games and was completely rolled in 2. Which team is better?"°fa-info°(Mothership)
- A current season example in AM - where a team with two forfeit wins results in them taking a playoff spot over a team with a better record (It seems unclear yo me if both teams got forfeit wins and the match page is missing information, but the main point stands).
I think a good quote to summarize the general issue with both:
match points isn’t an unusable system, nobody is saying that, what we’re saying is that a swiss style bracket paired with match points can give results that don’t make sense and can seem unfair.°fa-info°(cayorne)
Michlele identifies the issue with 2) as FFW and suggests point decay as a fix for the MP system.
My Thoughts
On W-L vs MP:
I think W-L makes for a much simpler system and one that is unarguably more intuitive, but I think this paired with the swiss format doesn’t really solve an issue of “unfair” playoff placements. While Example 1 attempts to indicate an unfair placement as a result of MP, an equivalent example surely exists should W-L be the only consideration - in fact I think BBL’s thread essentially highlights a belief that SWISS is the issue, not MP vs W-L.
What interesting to me is that I think the base PURPOSE of MP was to try to correct some of the inaccuracies of swiss. I’m not sure of the history or implementation, but I’d imagine the purpose of MP was to emphasize the value of “Stealing rounds off of top teams” - trying to work toward that better evaluation of teams to create more competitive playoffs.
On FFW
The essential problem with Example 2 that both mothership and Micahlele agree with is the FFW giving full match points. Micahlele offers a potential solution with multiple FFW decay, but I don’t think this addresses the root of the problem - a team that wins via FFW even once has an inherent advantage (Via full MP), in short his solution focuses on dealing with multiple FFW, but doesn’t any other aspect of the issue.
My instinct is to resolve forfeits by putting a resulting greater weight on the provable accomplishments of the team - the equivalent thought would be say two midterms and an final for a class accounting for 50% of a grade with 15%, 15%, and 20% weighting turning into one midterm and a final weighted at 20% and 30% respectively.
Unfortunately, I can’t instinctively see an intuitive way to fandangle this - so instead for practicality I’d suggest the approach of attempting to trying to ensure that a FFW doesn’t result in that team having a greater advantage over other “winners”. A quick and dirty way to do this might be to take the “average” points obtained by all teams that “won” their matches and give that average to the forfeit winner. It probably doesn’t really make sense that the amount of points won by a forfeit is tied to literally every other match - but I think it is the closer to a fair reward than the minimal 2 where the other team might not even exist anymore, or maximum 3 while other teams have to struggle for fractions of a point.
Note: I think it’s hilarious that this could mean even more fractional points, but I do think that if you’re going to use MP, emphasis should be placed on balancing those arbitrary points so that they are as accurate as possible when it comes to representation of a team.