[FSPA] Preplays are better

[Kevin Stone]

Anyone know the rule on how many forfeits this year before getting kicked out?  Used to be 8 but that was with up to 3 weeks of pre-plays for a no show of 5 weeks out of 10.  I’m guessing you don’t show 4 weeks and you’re out?

From: FSPA [mailto:fspa-bounces at fspazone.org] On Behalf Of Dave Hubbard via FSPA
Sent: Wednesday, June 30, 2021 10:57 PM
To: FSPA main discussion list
Cc: Dave Hubbard
Subject: Re: [FSPA] Preplays are better

> There are events that could change that like people dropping out or joining in late, but otherwise it’s set.

That's why it's incorrect to say that next week's groups are set in stone.  They can change.  In fact, section 6.4 lays out how to deal with drop-outs.

Whether this is a good system or not is of course a matter of opinion.  I'm personally not a fan of declaring next week's groups ahead of time and potentially playing with a physical 2 or 1-player group (preplays or not) but that's the system we have.

           --- Dave

On Wed, Jun 30, 2021 at 9:57 PM Stephen Jonke via FSPA <fspa at fspazone.org<mailto:fspa at fspazone.org>> wrote:
Groups do matter. I don’t understand why we are arguing. I just dropped another group because I’m playing really poorly so far. There are events that could change that like people dropping out or joining in late, but otherwise it’s set.

Steve
On Jun 30, 2021, 9:29 PM -0400, Bob K Mertz via FSPA <fspa at fspazone.org<mailto:fspa at fspazone.org>>, wrote:

All that may be true but that's literally what this argument seems to be
about... The groups don't matter so we should all just ignore them .....
but here's a list of the groups that don't matter anyway because it's
interesting but, remember, they don't matter.

I'll be honest with you, I've been in FSPA for years and until this
discussion began I always thought it was set in stone that if you win
your group you move up to the next group and if you lose you move down a
group and that the ladder was a completely seperate thing (perhaps this
is why it seems like I feel the "taking 0s" thing is more absurd than
others). My understanding of that was based solely on what I saw every
week - for all of those years. It's easy to say that it's just always
been about the ladder but for those of us that have watched the grouping
formula consistently remain true week after week it *feels* like you're
rewriting the rules even though you aren't.

I agree that it's interesting to know where you are playing next week
but if that's not set in stone then it's a "fantasy". Those fantasies
may essentially be fact right now since no one ever seems to actually
change the groups before league but watching the FSPA leaders get upset
because no one seems to comprehend that groupings don't matter while
they simultaneously say they are important because they are interesting
only adds to the confusion.

Obviously I've not understood this system and I never had any reason to
expect that I didn't. Previously I just ignore emails that turn into
these huge discussions but I got caught up in this one..... From a
practical sense others in the league may be making the same assumptions
I had been and having no reason to ask how things really work and aren't
reading this discussion to learn otherwise.

I've never felt comfortable really voicing my opinions about these huge
issues because it always seems to turn into a battle with those that
have been in FSPA for more than 10 years..... It usually doesn't seem
the take away ever is "hey, new comers or even those who have been
around for less than 15 years may see things differently". This time I
might have actually learned something but I'm not expecting the outcome
to be any different than any other huge discussion about the league.


On 6/30/21 3:46 PM, steve wrote:

Bob, what you are highlighting is a byproduct of practical
implementation details.

Sure we could not publish groups as part of prior week's results, but
knowing where people are playing is very interesting stuff to most
people.  And in the VAST majority of cases, what was projected when last
week's results, will still ring true when the next week starts.  So, for
the majority of cases, it's GOOD info and accurate to the future.  I
mean.. if you wanted to get technical, maybe the right takeaway is to
label the table "Projected Groups"

The software is just presenting a simplified view of the world for easy
consumption.  Prior implementation were more precise and dumped more
details, but most didn't use the info anyway, so it was streamlined.
The 'next week' table really shows two things - The stacked ladder, and
groupings overlaid on that if things stay as they did when the
scoresheet was published.

People are just holding onto it 'too tightly' vs what the information
really is.


TLDR: SLOs scream that the groupings don't matter but every week we
get emails that paint a

picture of how "crucial" the groupings actually are.

That is a misunderstanding of what is being said.  No one said
'groupings don't matter'.  It's that groupings are not what is SET first
nor are the 'fixed'.  They come last and are laid over whatever other
decisions are made ahead of it.  The tail doesn't wag the dog...

-Steve

On Wednesday, June 30, 2021, 02:21:29 PM EDT, Bob K Mertz via FSPA
<fspa at fspazone.org<mailto:fspa at fspazone.org>> wrote:


I think there is another huge problem being overlooked.... Well, the
creation of that problem is being overlooked. You're right that the
ladder doesn't gel well in the minds of many (myself often included) and
the one thing that seemingly is repeated over and over here is that
groups are never assigned until just before league starts......
...... Except that we get an email that tells us what group we are in
next week as soon as the results are posted.

I trust the ladder and I trust that the software is doing is thing so
why do the results emails we get try to paint a picture in the hopes of
explaining something that, apparently, doesn't at all seem to be the
case? Just don't tell us on advance what our groups are (or are supposed
to be) and tell us what groups we are in when we start league play and
this "false narrative" eventually disappears.

TLDR: SLOs scream that the groupings don't matter but every week we get
emails that paint a picture of how "crucial" the groupings actually are.


On June 30, 2021 1:45:54 PM EDT, steve via FSPA <fspa at fspazone.org<mailto:fspa at fspazone.org>> wrote:


The difference with FSPA seems to be that you guys try to
'pre-make' your groups of 3,

which obviously causes scrambling when people don't show

There are no 'pre-made' groups.  This is a common misunderstanding
about Groups in the FSPA design.  Groups are not pre-set before
league... a SLO could define groups at 5mins before league start if
they needed to.  What is 'fixed' in the FSPA system is the LADDER -
which is the stacking of players in order which serves as the
fundamental handicapping system in the FSPA match play system.  The
ladder starts with everyone's initial seeding, and is refined each
week by the movement defined by player performance against their
peers.  Group movement is the feedback loop that 'refines' the
ladder each week.  (Group Movement also gives people more variety in
competition you see.. but that's more a side-benefit)

FSPA is a match play system.. so a fundamental concept in match play
is "Who do you play in your match"

You could have
- random assignments
- groupings based on skill assessment/external rankings
- some non-random scheme designed to rotate or assign people by some
distribution (Example: Round robin)

The first and third systems generally will mix up players of
different skill, and there are pros and cons to such models.  But
over the long haul, you can expect the better players to score
better at the expense of their under matched opponents.

The second model tries to group similar skilled players together.
But is highly dependent on how you measure/assess/assign those skill
rankings.  This is difficult to do with new players, new locations, etc.

Systems like Pinburgh were 'skill' based, but rather have you only
play against your peers, it used a converging model.  So #1 didn't
always play against #1, #2, #3, they played against other ranks
until ultimately converging there.  Like the FSPA model, their
'ranking' was not fixed, but refined each round based on
performance.  (They used total points, FSPA instead uses group
promotion/demotion to refine your rank in the ladder).

FSPA fundamentally is setup to allow players of different skill
levels to compete together.  This isn't just about breaking into
divisions.  It's the fundamental concept that we have inherent
handicapping built into the system by grouping similar skill levels
together.  This functions so it makes sense to compare the 12 points
I got in a week in group 1, to someone in group 4 who also got 12
points.  We both got 12 points...  are we equal players?  Probably
not, but our handicapped output is the same.. and you win/lose the
overall league based on your handicapped output - your match points.

The premise of allowing players of different skill level to compete
and enjoy league is one of the core premises behind the FSPA rules
design.

Everyone generally agrees getting creamed in a group is not fun long
term - thus there is extra attention to the idea of not placing
players where they would be setup for failure for no fault of their
own.  Additionally, on the competitive side, it creates significant
disparities when people are not aligned with their skill group,
while other players are.  (easier/harder to get points, etc)

The point of all this is to try to explain why the Ladder is a
central pillar of the concept of how we pair players to play
together.  The more you move people around, the more you shake the
principal of how people are intended to be 'fairly' matched up which
is how we establish the equality of match points between different
groups.

Match points are comparable across groups as the measure of success
because of handicapping.  Instead of adding/subtracting to scores as
a handicap, we use who you compete against as the handicapping in
the system.  The ladder is the construct to do this.


*So, TLDR - what the f are you talking about?*  It's important you
keep people playing against similar skill (within our ability) as a
fundamental construct of how the competitive and fun factors of our
scoring model operates.  Obviously you can start over with another
model entirely, and do away with the handicap FSPA is built on, but
understand it's not just a 'group change' but cuts much deeper when
you move people around the ladder or who they are grouped with.

Skipping people not present when grouping is entirely feasible, the
messier part is how to systematically handle group movement after
the fact.  It would probably look a LOT more messy to players to
understand.  (only move winners and losers from their initial
position, not where they actually played, etc).

But players don't generally understand ladder movement in the first
place... so...  :)


On Wednesday, June 30, 2021, 10:16:19 AM EDT, Elliott Keith via FSPA
<fspa at fspazone.org<mailto:fspa at fspazone.org>> wrote:


Even though I had to drop out of MOM's league this season due to
personal stuff, I felt the need to chime in with a 'simple' solution.

Pretty much every other league I've played in besides FSPA had a
drop 2 format, with no headache. The difference with FSPA seems to
be that you guys try to 'pre-make' your groups of 3, which obviously
causes scrambling when people don't show. The 'drop 2' and 'pre-made
groups' just don't seem compatible to me. Most all other leagues
randomly dole out groups at the start of the night, and
Crabtowne's league had a 'division split' halfway through so you'd
play with people somewhat around your skill level.

My solution idea is that instead of trying to 'pre-make' groups, why
not just keep track of ladder position? Like, after confirming who's
not there, the top three present are group 1, next three group 2,
etc. all the way down, with the last group being a 4 or 2 if
necessary. Keeps the similar skill level thing going, and seems
pretty easy to implement, since you guys keep track of that anyway.

Just my two cents. Hope to be back in a league next season, and
happy 4th!

-Elliott

On Wed, Jun 30, 2021 at 9:57 AM Rob Wintler-Cox via FSPA
<fspa at fspazone.org<mailto:fspa at fspazone.org> <mailto:fspa at fspazone.org<mailto:fspa at fspazone.org>>> wrote:

On Wed, Jun 30, 2021 at 9:38 AM Daniel Northover
<northovr at verizon.net<mailto:northovr at verizon.net> <mailto:northovr at verizon.net<mailto:northovr at verizon.net>>> wrote:

Just ribbing you Rob your doing a awesome job


After last week my ribs are pretty sore. ;)
_______________________________________________
FSPA mailing list
FSPA at fspazone.org<mailto:FSPA at fspazone.org> <mailto:FSPA at fspazone.org<mailto:FSPA at fspazone.org>>
http://lists.fspazone.org/mailman/listinfo/fspa

_______________________________________________
FSPA mailing list
FSPA at fspazone.org<mailto:FSPA at fspazone.org> <mailto:FSPA at fspazone.org<mailto:FSPA at fspazone.org>>
http://lists.fspazone.org/mailman/listinfo/fspa


--
Sent from my Android device with K-9 Mail. Please excuse my brevity.
_______________________________________________
FSPA mailing list
FSPA at fspazone.org<mailto:FSPA at fspazone.org> <mailto:FSPA at fspazone.org<mailto:FSPA at fspazone.org>>
http://lists.fspazone.org/mailman/listinfo/fspa
_______________________________________________
FSPA mailing list
FSPA at fspazone.org<mailto:FSPA at fspazone.org>
http://lists.fspazone.org/mailman/listinfo/fspa
_______________________________________________
FSPA mailing list
FSPA at fspazone.org<mailto:FSPA at fspazone.org>
http://lists.fspazone.org/mailman/listinfo/fspa
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.fspazone.org/pipermail/fspa/attachments/20210701/707292f8/attachment.htm>