This video was posted to YouTube a couple of months ago,
but I only just discovered it.

https://www.youtube.com/watch?v=2ie9PmNM678

I thought it was excellent. Highly recommended. I'm also
extremely impressed that Mochy has hit the Super-Grandmaster 2
level. Incredible! Especially since nobody else has reached
Super-Grandmaster 3 yet.

I remember having an email conversation with Rick Janowski
back in 2014, when they were trying to figure out how to set
up the title system for the BMAB. In his initial proposal,
the highest title you could get was Grandmaster 1, with a
threshold of 3.0. I went to great lengths to argue that higher
titles were necessary, and suggested adding a 2.0 threshold
and a 2.5 threshold. At the time, apparently a lot of people
thought that 2.0 was unattainable and didn't see the need.
But I said that people like Mochy needed a stretch goal, and
I pointed to the fiasco with the rock climbing rating system
as an analogy. I don't remember with 100% certainty, but
I think that BMAB started off with G1 being the highest level,
and they only added the Super-Grandmaster level later, when
(predictably) more people than they expected were achieving
the 3.0 level. It's good to see that people are finally coming
to realize that 2.0 is attainable and that giving it a name
helps motivate them.

---
Tim Chow

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 2/5/2022 7:35 PM, Timothy Chow wrote:

This video was posted to YouTube a couple of months ago,
but I only just discovered it.

https://www.youtube.com/watch?v=2ie9PmNM678

Mochy makes some interesting comments about whether PR is the best
basis for something like the BMAB or the UBC.

https://youtu.be/2ie9PmNM678?t=4965

I used to argue for MWC instead of ER/PR, but I changed my mind, based
in part on a discussion on BGOnline in which Mochy took part. In my
view, the strongest argument in favor of something like ER/PR is what
I'd call the "Hiawatha" argument.

https://mathshistory.st-andrews.ac.uk/Extras/Hiawatha/

MWC has the advantage of being unbiased, but for something like UBC,
lack of bias is less important than being right more often, and I do
believe that PR/ER is right more often.

On the other hand, there is one point about PR that it seems very few
people understand, which is that if you mess with the definition of a "decision" then it introduces weird biases; e.g., errors in positions
where you have a high chance of getting blitzed are magnified. When
you're on the bar, XG says that you have no decisions to make, and so
your equity loss is divided by a small denominator. More generally,
an important feature of ER/PR is that it tests how well you make
decisions that come up in actual play. If there's a certain type of
decision that you don't understand but it never comes up in practice,
then it stands to reason that your ER/PR shouldn't penalize you.
Conversely, if some type of decision comes up all the time, then your
ER/PR should reflect that. If you declare that certain types of
moves are not "decisions" and you remove them from consideration, then
errors in games from which you have removed many moves are going to
be artificially magnified.

Ultimately, though, what's more important for something like BMAB is
that the standards remain stable. It's better to stay with something
that has some minor defects than to tinker endlessly with it. So I'd
also be in favor of staying with the current definition instead of
changing it.

---
Tim Chow

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Saturday, February 5, 2022 at 7:35:46 PM UTC-5, Tim Chow wrote:

This video was posted to YouTube a couple of months ago,
but I only just discovered it.

https://www.youtube.com/watch?v=2ie9PmNM678

I thought it was excellent. Highly recommended. I'm also
extremely impressed that Mochy has hit the Super-Grandmaster 2
level. Incredible! Especially since nobody else has reached Super-Grandmaster 3 yet.

I remember having an email conversation with Rick Janowski
back in 2014, when they were trying to figure out how to set
up the title system for the BMAB. In his initial proposal,
the highest title you could get was Grandmaster 1, with a
threshold of 3.0. I went to great lengths to argue that higher
titles were necessary, and suggested adding a 2.0 threshold
and a 2.5 threshold. At the time, apparently a lot of people
thought that 2.0 was unattainable and didn't see the need.
But I said that people like Mochy needed a stretch goal, and
I pointed to the fiasco with the rock climbing rating system
as an analogy. I don't remember with 100% certainty, but
I think that BMAB started off with G1 being the highest level,
and they only added the Super-Grandmaster level later, when
(predictably) more people than they expected were achieving
the 3.0 level. It's good to see that people are finally coming
to realize that 2.0 is attainable and that giving it a name
helps motivate them.

---
Tim Chow

I said 2 was attainable forever ago. I'm sure we could find it on BGO if we really wanted (I don't). There's no real motivation to get there. A lot of players, the ones who could make a dent into the higher levels of BMAB, just don't care about it.
Maybe that's sad but it's certainly true. I couldn't even imagine a scenario where I'd ever bother with it again and I've talked to quite a few other Giants/should be Giants who feel the same.

Stick

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 2/5/2022 9:46 PM, Stick Rice wrote:

I said 2 was attainable forever ago.

Yes, I remember that.

There's no real motivation to get there. A lot of players, the ones who could make a dent into the higher levels of BMAB, just don't care about it.

There's always a younger generation of players that could decide
that they care.

I seem to remember that back in the 1990s, there was a small core
of Rubik's Cube speedsolvers who were thinking that maybe some
people could achieve what seemed to be the Holy Grail at the time
(12 seconds maybe?) but the best people weren't motivated. Rubik's
Cube seemed to be basically dead. Of course, today, 12 seconds is
a joke to serious speedcubers.

---
Tim Chow

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Sunday, February 6, 2022 at 12:54:04 AM UTC-5, Tim Chow wrote:

On 2/5/2022 9:46 PM, Stick Rice wrote:

I said 2 was attainable forever ago.

Yes, I remember that.

There's no real motivation to get there. A lot of players, the ones who could make a dent into the higher levels of BMAB, just don't care about it.

There's always a younger generation of players that could decide
that they care.

I seem to remember that back in the 1990s, there was a small core
of Rubik's Cube speedsolvers who were thinking that maybe some
people could achieve what seemed to be the Holy Grail at the time
(12 seconds maybe?) but the best people weren't motivated. Rubik's
Cube seemed to be basically dead. Of course, today, 12 seconds is
a joke to serious speedcubers.

---
Tim Chow

I love your optimism that there will be a 'younger generation of players' nevertheless that they'll be good enough to make a charge at BMAB. (in the US at least)

Stick

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Sunday, February 6, 2022 at 5:54:04 AM UTC, Tim Chow wrote:

On 2/5/2022 9:46 PM, Stick Rice wrote:

I said 2 was attainable forever ago.

Yes, I remember that.

There's no real motivation to get there. A lot of players, the ones who could make a dent into the higher levels of BMAB, just don't care about it.

There's always a younger generation of players that could decide
that they care.

I seem to remember that back in the 1990s, there was a small core
of Rubik's Cube speedsolvers who were thinking that maybe some
people could achieve what seemed to be the Holy Grail at the time
(12 seconds maybe?) but the best people weren't motivated. Rubik's
Cube seemed to be basically dead. Of course, today, 12 seconds is
a joke to serious speedcubers.

The 12-second sceptics probably were only thinking from the dexterity viewpoint, and ignoring the inevitability of better algos emerging.

How long does it take you to solve it?
I used to do it in a minute approx, but I forgot the solution.
I never worked it out -- I just memorized a solution that I saw when
I was on an exchange visit in Germany. That may well be the same trip
where I unwittingly ordered too much ice-cream, which I described on an
earlier thread.

Paul

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Sunday, February 6, 2022 at 5:22:44 PM UTC, Stick Rice wrote:

On Sunday, February 6, 2022 at 12:54:04 AM UTC-5, Tim Chow wrote:

On 2/5/2022 9:46 PM, Stick Rice wrote:

I said 2 was attainable forever ago.

Yes, I remember that.

There's no real motivation to get there. A lot of players, the ones who could make a dent into the higher levels of BMAB, just don't care about it.

There's always a younger generation of players that could decide
that they care.

I seem to remember that back in the 1990s, there was a small core
of Rubik's Cube speedsolvers who were thinking that maybe some
people could achieve what seemed to be the Holy Grail at the time
(12 seconds maybe?) but the best people weren't motivated. Rubik's
Cube seemed to be basically dead. Of course, today, 12 seconds is
a joke to serious speedcubers.

---
Tim Chow

I love your optimism that there will be a 'younger generation of players' nevertheless that they'll be good enough to make a charge at BMAB. (in the US at least)

Stick

Young people might get interested in bg when "Squid Game" gets less interesting.

Paul

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 2/6/2022 12:22 PM, Stick Rice wrote:

I love your optimism that there will be a 'younger generation of players' nevertheless that they'll be good enough to make a charge at BMAB. (in the US at least)

I was actually thinking of Japan more than the U.S.

---
Tim Chow

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 2/6/2022 1:48 PM, [email protected] wrote:

The 12-second sceptics probably were only thinking from the dexterity viewpoint, and ignoring the inevitability of better algos emerging.

Well, maybe that was part of it. But I think there was a strong
sociological factor. Stick looks around him and makes a prediction
about BG based on what he sees. That's totally reasonable, but it
can also be wrong. If you're a speedcuber in the late 1990s, what
do you see? You see a small coterie of enthusiasts, which shows no
sign of growing significantly. The cube has been "dead" for over
ten years. You see the top people with solving times of 15-16 seconds,
or maybe the occasional 14 or even sub-14. You know that in theory
there could be better algorithms, but they're going to require an
enormous amount of research, memorization, and practice. Who is going
to be motivated to do that?

Such reasoning is actually pretty accurate *if* nothing happens to
cause the population of enthusiasts to expand tremendously. In the
case of Rubik's cube, it's not totally clear what caused the
renaissance, but it's widely speculated that the growth of YouTube
and the web was the spark. Once enough people got interested, it
didn't take too long for the solving times to drop dramatically.

Backgammon faces a somewhat steeper hurdle. The good part is that
in principle, a player can train in isolation to achieve low PRs,
but that's not enough. One has to perform on demand in public
competitions, and for backgammon, the entry fees are traditionally
rather high, due in large part to its traditional association with
gambling. Again, in principle, that could change, but the barriers
to change are not trivial. I think that Japan has a better shot at
it than the U.S. does.

How long does it take you to solve it?
I used to do it in a minute approx, but I forgot the solution.
I never worked it out -- I just memorized a solution that I saw

I'm in more or less the same boat as you are, except I was probably
taking more like a minute and a half. I learned a solution when
I was in school. Later, I managed to figure out on my own the
principle that is explained in this excellent video---

https://www.youtube.com/watch?v=-NL76uQOpI0

---and I almost managed to figure out a complete solution to the 4x4x4
as a result. I think you need one more trick for the 4x4x4, which I
never bothered to work out. But anyway, if you're not familiar with
the above trick, you might find it interesting to try out. For once,
the click-baity title, "Solve all twisty puzzles with this one weird
trick" is (almost) entirely true.

---
Tim Chow

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)