(Sigh) just as I was feeling proud to have finally made it to the Nat Championship, Al points out I'm an imposter and have no right to be there 🤣🤣🤣

But, of course, I think you're right - b) is a better system than a) if you're looking to identify the best sides in the league. There's no world where Pekin are a superior side to St. Charles (as our 0-2 record testifies) so it would more accurately reflect relative strength. 

My only caveat is that the NC games can skew performance versus Q1/2 sides. This year I've only played the top teams in games they care about, hence the abysmal record. Last year Pekin were a lot worse but still recorded some famous victories over better sides in NC matches because they were otherwise engaged in Cup Finals, while I had nothing else to play for.

Nearly forgot your other question - Amy, Kelsey, Max?

I like the proposed idea. For a start it has built in bonuses for WGM and Conference Tourney semi-final and final winners.

As for non-conference games they get skewed anyway: Quincy are 1-5 here which gives them -7 points. We are 3-3 which I think gave us +4 points (Jacob has taken his tablet back so I can't easily check).I also think Al used Quad I/II partly because he's actually 1-5 vs Quad I alone but that 1 win outscores all the losses as it was the Tourney final win.

I agree on Kelsey, Amy and Max. I find the 4 girls behind them amusing: 3 are applauding the point win but the 4th either doesn't care about Volleyball (there supporting her boyfriend?) or is checking out one of the opposing players.

I like the proposed model. It mirrors reality which is good. It gives more emphasis to the NET Rankings, which is good (despite them being biased against Antioch because we can't be anyone not in a "good quadrant").

I also like using the NET Rankings. My thoughts on "why" echo the comments above. One thing we should also decide is what happens in the event there is a tie in the NET rankings. This may be unlikely, but it's not impossible. I note that Mat-Su and Albuquerque are currently tied; although that would not have an impact on the National Championship if we were using NET this season, it would impact their seeding in the Holyoke. 

I like the proposed idea for the reasons stated by the others above. If there is a tie in NET rankings, the only fair way to resolve it is to give the higher place to the team with more ferocious mascot.

I think using the NET rankings is fair. The only drawback I can see is that they are so fluid especially around the middle of the pack where this would affect the most teams that it can be hard to see a week or two out what you need to do to make it in (e.g., if you are 3 wins above 15th in the overall standings you know where you definitely stand two weeks from the cutoff in the current setup).

For ties I'd imagine using things like Head to Head, NET Power Ranking, Performance against common opponents...

I also like the proposed solution and with the same arguments as others. Maybe last seasons winner should be given a place (maybe they will get a high NET anyway)

Ok, this might be a crazy idea but I was thinking about how there's a possible advantage to teams that have a lot of Quadrant I/II games and this is something at the moment we have little control over because the schedule is quite formulaic. For instance, St. Charles this year would take my spot in the National Championship with his 7 wins over Quad I/II teams. But he also had 18 Quad I/II games to my 7! 

In real college sports you have some autonomy over your non-conference opponents and can try to schedule games based upon who's willing to play you but also whether you want to increase your strength of schedule (equivalent to trying to schedule an extra Quad I/II game to increase your postseason chances with a win). What I'm thinking here is we could have one of our non-conference games be scheduled by the teams instead of formula through a sort of auction process. We could use RPs as the currency. I haven't thought all the way through this (and maybe this is a just a way to get the ideas flowing if people like this idea) but I'm thinking you could put in a bid on how much an extra game against all non-conference teams is worth to you and we could use a matching algorithm where the two teams that combine for the most RPs would be scheduled first and taken out of the draw and then we would keep doing that down until all teams had scheduled their non-conference opponents. And those RPs that you bid would go to the opponent you picked so there would be no net loss of RPs and an extra layer of gamesmanship. Anyway, just a thought to maybe kickstart some discussion.

To be clear you would put in a separate bid for each non-conference team (or your top 5 choices) and those would be matched up against everybody else's bids.

I choose b).

Why? Because we'll always qualify.

Boom.

I think this is a good idea...

But does it create an incentive to lose big?  It is better to lose to a quad I or II team for Net rankings, so  losing by 30 points is better than by very little.  If you already have lost twice to a team and play them for a third time could you lose so badly that you can move them 2 quads up and gain points?  Moving them 1 would limit the damage.

It is quite difficult to lose very badly of course, so perhaps this is a moot point!

Well it might be quite difficult for you to lose very badly James, some of us are experts.

I like the NET option over straight record.  An amended version of NET is in the works which will also try to take in account 3-set wins, 5-set losses, and include tie breaks for season head-to-head or overall record (still working that out for preference).  That should hopefully account for the tie-scenario without Al having to do extra work on his end.

And... because the NET already has anti-Antioch formulae worked in, so again there's no need for extra coding.

I prefer to use the overall standings.

One of the great things about MSWL is that everything is predictable, whereas the NET rankings aren't. With the standings, if I'm on the bubble, I know I'll need to win the next X games to make it into the national championship. With NET rankings, I'll have absolutely no idea what I'll need to do apart from "winning is helpful." While this seems like an odd situation, it may appear like I have a chance to qualify when I don't, whereas the standings are simple math. And as Bryce says there's no way to control who you play.

My alternative proposal is using the standings to qualify the at-large teams, then the NET ratings to seed both tournaments.

Also, if we do switch to this, non-conference games should be seeded, if they aren't already.

I don't care how we rank the teams in the conference, but it looks like you play six total non-conference games, two against each other conference.

#1 seeds should play #1 seeds and #8 seeds from the other conference

#2 should play #2s and #7s

#3s should play #3s and #6s

#4s should play #4s and #5s

et cetera

I agree with John that non-conference should be seeded.

Non conference games are seeded from the previous season 1st and 2nd in each conference at the end of this season wil play 1st and 2nd from the other 3 conferences next season.

Steve is that a new change or how we've always done it?

It was introduced with the expansion. We used to play 8 non conference games but the expansion meant extra WGM Cup and TMVL games so it made sense to drop to 6 non conference game as it was easy to work out.

In the first expansion season the old and new teams were kept apart.

Can we ensure the #1 and #2 teams from last season didn't get a ratings boost from playing the better teams?

Assuming that strong teams stay strong and weak teams stay weak from one season to the next  (Columbia might end up as a Holyoake team but nit overnight) then as far as NET ratings are concerned it's definitely an advantage to be strong. A 3-3 record against Quad I teams is far better than against Quad IV teams. 

I don't think it's necessarily a disadvantage for a weak team to play a strong team. The latter often treat the non- conference games less seriously as they're after bigger titles.

So I like John's idea of #1 playing all #1 and #8 seeds.

But do we seed based on Conference position or NET ratings? We are 8 points behind Quincy in NET but a win in the TMVL Cup final is worth 8 points (I believe as we're playing a Quad I team) so we could finish 2nd in the Conference but ahead of Quincy in NET.

I would prefer conference position since it's something predictable. It's a little less advanced, but you'll know exactly who your opponents are based on where you finish.

Ir's also a lot easier for Al to code.

Will there be a change this year Al?

Thanks for asking.  

There will be no change this season as I ran out of time to try to put something in place and update rules.   I believe we have some solutions above though (such as how we schedule non conference; which requires coding) to make this happen for season 10.

This is on my list for planned season 10 changes that I'll need to firm up.