Best documented MK vs XG experiment yet with videos.

[mu...@compuplus.net]

http://www.montanaonline.net/backgammon/xg.php

I've been working at redoing the entire site to
be more user friendly. I still want to add notes
for each set of experiments, once I get another
spurt of interest.

The last expeiment at the above link took about
6 hours in total playing time.

I captured the entire session onto video with a
clock and dicedll dialog box always on the screen.

I saved the session in 4 videos (25 games each)
but in a single .xg file.

Videos are at normal speed, with necessity pauses
between games. On an Intel-i5 4-core 3.2ghz cpu,
you can see that XG needed time to think ;) and
used up most of the playing time. You can speed
up the videos when playing back if you want.

At the above link, there is also an exported .pxg
user profile and a .pdf with my own analysis as a
spreadsheet and two graphs (I hope you can make
some sense out of them).

The result was MK +43 with 18.76 cube error, 9.56
checker error and insignificant +0.0013 luck.

Feel free to comment on what you make out of it.

MK

[Michael]

A)It's sessions of 25 games. As such they don't prove anything. Because as soon as you establish a lucky lead within the first few games it's very easy for anyone to maintain that lead. I am not very familiar with money games but I think it should have been sessions of say 25 pointers instead.
B) despite the clock and the dice roller shown on the videos we have no way to tell how many sessions and how many videos you could have possibly scrapped
C)There are ways to conduct a controlled experiment in which your audience can check whether you have hidden the sessions that didn't work for you, but I don't think you will ever accept such thing.
D)at the risk of losing my time (one more time) I propose you to use a dice roller that I would design for you. The rolls will be totally random, but will be marked sequentially. This way you will not be able to hide any sessions that you played using it.

[mu...@compuplus.net]

December 30, 2016 at 4:19:58 PM UTC-7, Michael wrote:

>> I saved the session in 4 videos (25 games each)
>> but in a single .xg file.

>> At the above link, there is also an exported .pxg
>> user profile

>> The result was MK +43 with 18.76 cube error, 9.56
>> checker error and insignificant +0.0013 luck.

> A)It's sessions of 25 games. As such they don't
> prove anything. Because as soon as you establish
> a lucky lead within the first few games it's very
> easy for anyone to maintain that lead.

Boy, you're really hard to fool smartie, aren't you.

As I suggested to you before, you need to learn to
read carefully, try to understand what you read and
not jump to false conclusions before you respond.

I played a session of 100 games without closing and
reopening XG, thus it's saved as "one single .xg"
file!

It was "sub-totalled" in the video recorder every
25 games to keep the video file sizes to about 10mb.

> I am not very familiar with money games but I think
> it should have been sessions of say 25 pointers instead.

I play long sessiosn ("money games" as they call it)
because it allows a better way to debunk the "cube
skill bs". If you paid attention, you woul have seen
that my cube error rate is twice my checker error rate.

> B) despite the clock and the dice roller shown on the
> videos we have no way to tell how many sessions and
> how many videos you could have possibly scrapped

If you were smart enough, you could have seen that the
dice rolls displayed by the dll flow from one video to
the next...

> D)at the risk of losing my time (one more time) I
> propose you to use a dice roller that I would design
> for you. The rolls will be totally random, but will
> be marked sequentially. This way you will not be able
> to hide any sessions that you played using it.

The dice rolls used in the session are "sequential" and
match the ones in the .xg file. If you knew the seed,
you could duplicate the entire sequence. Even if there
is no way to know the seed in XG, you can ask Tim Chow
for example, about how to figure out the seed given a
long enough sequence of rolls.

I also explicitly included an exported .pxg profile so
that you all could see there were exactly 100 games
played by the player "MK Dicedll++" and in "Iron Man"
settings.

I know you all bot worshippers are tormented and I do
feel your pain... Sorry! :(

MK

[Michael]

Still we have no way to tell how many more sessions of 100 games you probably had scrapped.
I wish I could convince myself beyond any reasonable doubt that your results are genuine.
Then we could go to next experiment that of playing series of normal 11 pointers and see what happens there.
Again I would wish that you were proven right in that 2nd experiment as well.
I would then personally scrap all bot's verdicts and forget all about them.
Unfortunately you are not cooperating.

[Tim Chow]

On Saturday, December 31, 2016 at 7:37:55 AM UTC-5, Michael wrote:
> Unfortunately you are not cooperating.

Murat is not a troll, of course.

On a completely different topic, let me randomly mention that trolls never
cooperate.

---
Tim Chow

[mu...@compuplus.net]

December 31, 2016 at 5:37:55 AM UTC-7, Michael wrote:

> Still we have no way to tell how many more sessions
> of 100 games you probably had scrapped.

You really made me curious now. Have you ever thought
about how many sessions of 100 games one would have
to scrap before ending up with a winner to keep?

And what would be the distribution of sessions scrapped
after N games (1<N<100)??

Surely it would be much easier to scrap a few games than
to scrap 80-90 games...(?)

Just in case you have difficulty figuring it out, you
can ask help from Chow who is the expert at concocting
such formulas that involve match length multiplied by
pi, divided by the sq root of ELO difference minus 1500,
etc... :)

> I would then personally scrap all bot's verdicts and
> forget all about them.

I'm not trying to scrap all bots but to create better
bots than the current ones built on fart-ass fantasies
and folk-science...

> Unfortunately you are not cooperating.

You are not either.

With your dice roller, you are not offering anything new
that can't accomplished now using the existing rollers
that come with the bots.

Still, I wouldn't mind using yours if you also take some
of my suggestions to make it a dll that can be used both
with gnubg and xg.

Why aren't you cooperating on that?

MK

[mu...@compuplus.net]

December 31, 2016 at 8:43:03 AM UTC-7, Tim Chow wrote:

> Murat is not a troll, of course.

> On a completely different topic, let me randomly
> mention that trolls never cooperate.

Trolling is what you are doing right here by making
useless, meaningless comments...!

Because you have nothing better to say on the subject.

And because you are so afraid of losing your unearned,
circumstantial credibility, you are scared shitless
of saying something wrong (as you have done countless
times in the past) in trying to make a meaningful and
logical argument.

Thus, you egg on others like Michael... :(

Your worries are unnecessary. Shit doesn't stick to
teflon assholes like you anyway... ;)

MK

[Tim Chow]

On Tuesday, January 10, 2017 at 8:14:43 PM UTC-5, mu...@compuplus.net wrote:
> Trolling is what you are doing right here by making
> useless, meaningless comments...!
>
> Because you have nothing better to say on the subject.
>
> And because you are so afraid of losing your unearned,
> circumstantial credibility, you are scared shitless
> of saying something wrong (as you have done countless
> times in the past) in trying to make a meaningful and
> logical argument.
>
> Thus, you egg on others like Michael... :(
>
> Your worries are unnecessary. Shit doesn't stick to
> teflon assholes like you anyway... ;)

Hilarious! Murat, you never fail to disappoint by providing free
entertainment. Thank you for bringing a smile to my day.

---
Tim Chow

[Michael]

On Wednesday, January 11, 2017 at 3:00:44 AM UTC+2, mu...@compuplus.net wrote:
> December 31, 2016 at 5:37:55 AM UTC-7, Michael wrote:
>
> > Still we have no way to tell how many more sessions
> > of 100 games you probably had scrapped.
>
> You really made me curious now. Have you ever thought
> about how many sessions of 100 games one would have
> to scrap before ending up with a winner to keep?

2 or 3. All you have to do is raise the cube to maximum possible to establish a race lead at the beginning. The rest is easy
btw you accuse everybody of mastur*g with bots but it looks to me you are the number one in playing with them for hooours. 6 more hours is nothing...

>
> Still, I wouldn't mind using yours if you also take some
> of my suggestions to make it a dll that can be used both
> with gnubg and xg.
>
> Why aren't you cooperating on that?
>

Because I don't know how to make it a dll. VB binds the code into an *.exe I haven't seen anything turning it into a dll yet..
I am new on VB and don't know enough for the moment.
Besides it will even be difficult for me to make it record the rolls sequentially and keep that recording inside the executable even AFTER you are done using it.The idea is to continue counting from the point it stopped the previous time.
If you promise you will use it after it's done then I will try do it.

[mu...@compuplus.net]

On January 11, 2017 at 1:32:35 PM UTC-7, Michael wrote:

>>> Still we have no way to tell how many more sessions
>>> of 100 games you probably had scrapped.

>> You really made me curious now. Have you ever thought
>> about how many sessions of 100 games one would have
>> to scrap before ending up with a winner to keep?

> 2 or 3.

Hundreds, you mean? Otherwise we must not be talking the
same language...

> All you have to do is raise the cube to maximum possible
> to establish a race lead at the beginning. The rest is easy

First: raising the cube to maximum possible to establish
a race lead in just 2-3 tries would be nearly impossible.
It can perhaps be achieved in a few dozen tries.

Decond: after you establish a beginning lead, it would be
even harder to maintain a winning score until the end of
a 100-game session. I would guess that it would take at
least several hundreds of scrapped (unfinished) sessions.

Maybe you are talking about something else, though..??

> btw you accuse everybody of mastur*g with bots but it
> looks to me you are the number one in playing with them
> for hooours. 6 more hours is nothing...

I used the term in referring to a bot's doing rollouts,
likening it to a bot's "playing with itself", not about
humans playing against the bots.

> Because I don't know how to make it a dll.

Oh, okay. Forget about it, then. Actually, it is something
that the bot developers should have done long time ago but
won't for some reason.

> Besides it will even be difficult for me to make it record
> the rolls sequentially and keep that recording inside the
> executable even AFTER you are done using it.

Nothing will prevent running another copy of it from fresh.

> If you promise you will use it after it's done then I will
> try do it.

Let's not waste time on this.

What you said above though, about establishing a beginning
lead made me think and play another session of 100-games.

Ironically, in the past, I had scrapped sessions early on
exactly because of having gained a big enough lead right
of the bat, which would make you all even more suspicious.

So, this time, I said why not try at it even harder because
it seems like I play bolder at the beginnings, possibly with
the idea of heaving nothing to lose (yet).

Once I gain a lead, I may be playing more to not lose what
I have gained instead of risking to lose it by trying too
hard to gain still more.

It seemd to work as I gained a bigger lead in several high
cube games, even though I lost some of those.

Then the same pattern of mostly low cube games... I haven't
quite figured out why. I even wondered if the bot starts to
cheat more in proportion and that there is nothing I can do
(i.e. can't cause the cube to go high by muself, unless the
bot is willing to "tango" also)...

However, I never stopped trying to win higher cube games and
in fact my lead peaked at the 59th game!

Before the last somewhat screwy 10 games, I was still +78
and finished the session with +50.

I posted the XG game and profile files and my stats at:

http://www.montanaonline.net/backgammon/xg.php

I think I should say something nice about your having a
positive influence on my attitude here. Instead of being
combattive all the time, I started looking back at my
own games to understand what I am doing right or wrong,
and just keep doing what I do regardless of who values
my efforts and comments.

One thing emerging from my experiments is that I was
able repeat my results four times in row, in 100-game
sessions, which proved to myself that:

1- Playing against a bot with a "masturbating style"
and a "one track strategy" checker skill is not so
pleasurable, exciting, challenging but not terribly
bad either.

2- The so-called "cube skill" is indeed a horse fart
and all attempts to derive any analyses, error rates,
etc. are meaningles, useless.

MK

[Michael]

On Friday, January 13, 2017 at 8:20:25 AM UTC+2, mu...@compuplus.net wrote:

>
> Nothing will prevent running another copy of it from fresh.
>

No, no, there are ways to prevent this. One way is to pass a single entry into your registry but I don't want to do such a thing. Another way is to make the program send each and every roll by email.I don't want to do that either.The third way is to use an encrypted file, you will know both it's name and location but you will not be able to delete it more than once otherwise it will start counting from 1 again and you will possibly need years to complete your 100 games session

>
> 2- The so-called "cube skill" is indeed a horse fart
> and all attempts to derive any analyses, error rates,
> etc. are meaningles, useless.
>
> MK

There is a possibility XG plays junk-better say it handles the cube totally wtong- on money sessions.Most probably it was not designed for that. Another possibility concerns the MET table used. It was actually derived from statistics on matches (upto 25A-25A) not from money sessions...

Why don't you switch to normal matches? I personally understand nothing from those money sessions, and I am totally in the dark about those beavers and racoons.

[Michael]

On Friday, January 13, 2017 at 8:20:25 AM UTC+2, mu...@compuplus.net wrote:
>

> 2- The so-called "cube skill" is indeed a horse fart
> and all attempts to derive any analyses, error rates,
> etc. are meaningles, useless.
>
> MK

Correct me if I am wrong, are your statistics saying you won 40 games and the bot won 60? If yes then that's a crystal clear indication the bot handles the cube totally wrong on money sessions.

Imo you should proceed to prove that the same is true on normal matches.

[mu...@compuplus.net]

On January 13, 2017 at 4:47:39 AM UTC-7, Michael wrote:

> There is a possibility XG plays junk-better say it
> handles the cube totally wtong- on money sessions.
> Most probably it was not designed for that.

I'd think it's probably the opposite. Remember that
these bots are marketed as "investments" in "tools
of the trade" for people aspiring to make a living
as a gambler...

> Another possibility concerns the MET table used.
> It was actually derived from statistics on matches
> (upto 25A-25A) not from money sessions...

I'm sure these "teaching bots" would be smarter than
that. A session is basically a bunch of 1-pointers
(i.e. 1A-1A) with unlimited cube, and thus, in other
words are played for the sole purpose of gambling.

> Why don't you switch to normal matches?

Initially I refused to use the cube at all and played
calssical cubeless 5 or 7 point matches.

Then I started playing cubeful matches to demonstrate
that the "cube skill" is not what it's touted to be.

And finally I started playing "money sessions" as an
even better way of debunking the "cube skill".

> I personally understand nothing from those money
> sessions, and I am totally in the dark about those
> beavers and racoons.

Neither do I! :)) That's exactly my point. You don't
need to (as long as you are strong enough with checker
play).

One proof I offered again and again, that has never been
argued against (let alone being disproven) is that you
can insert random/inane cube decisions into the bots'
code without affecting the results of cubeful rollouts
for very early positions, which proves that that such
rollouts are completely inaccurate and useles garbage.

My 4 sets of 100-game sessions results are a real-life
demontration of this.

My cube error rate and and actual win performance are
so far apart that the only conclusion is that cube skill
is horse fart...! :)

MK

[Michael]

On Saturday, January 14, 2017 at 2:08:21 AM UTC+2, mu...@compuplus.net wrote:
>
>
> My cube error rate and and actual win performance are
> so far apart that the only conclusion is that cube skill
> is horse fart...! :)
>

Sure but the point is to convince others that your results are genuine, and you have not hidden sessions that didn't work as "planned".
I said before that you will not need more than 2-3 trials to achieve a huge initial lead. Let's take an example:The bot Doubles at a position where it has 70-30% winning chance. You redouble, the bot re-redoubles/beavers whatever that's called, and this goes on almost indefinitely. Remember you still have 30% chances to win the game, so it will work fine for you 1 out of 3 times.

There is an indication from your own statistics that this is exactly what you are doing. Look at how high the cube was at game 3 and game 5 that established an initial lead for you of 96 Vs 16. The next such high cube value occurred at game 53 (48 games later!).In the meantime all cubes where between 2 and 4...

In a nutshell we have indications that your results might no be genuine, and indications that the bot might no be able to handle cube decisions correctly in money sessions.
But that's all indications,no proof whatsoever.
All's left is to provide evidence beyond any reasonable doubt for your results.

[Michael]

On Saturday, January 14, 2017 at 2:08:21 AM UTC+2, mu...@compuplus.net wrote:
>
>

> One proof I offered again and again, that has never been
> argued against (let alone being disproven) is that you
> can insert random/inane cube decisions into the bots'
> code without affecting the results of cubeful rollouts
> for very early positions, which proves that that such
> rollouts are completely inaccurate and useles garbage.
>
>

Have YOU inserted random/inane cube decisions into the bots' code and you found out that is does NOT affect the results of cubeful rollouts??
Sorry but this is just YOUR hypothesis... and the fact that nobody cared to disprove it does not make it right.
All you have to do (to prove your own hypothesis WRONG) is just do rollouts of any position and then look at the statistics and how the cubeful equity was calculated from those statistics. Then just insert random/inane cube actions into those statistcs and if you ever manage to keep the results UNAFFECTED I will cut my throat ;-)

[Michael]

On Saturday, January 14, 2017 at 2:08:21 AM UTC+2, mu...@compuplus.net wrote:
>

> One proof I offered again and again, that has never been
> argued against (let alone being disproven) is that you
> can insert random/inane cube decisions into the bots'
> code without affecting the results of cubeful rollouts
> for very early positions, which proves that that such
> rollouts are completely inaccurate and useles garbage.
>

I kept on scratching my head how you could possibly get that impression.
Notice the bots don't do rollouts when playing.At best they do 2-3 ply EVALUATION based on the 0-ply numbers they get from their database.
Now the 0-ply cubeful equity at EVALUATION level, will never change no matter what you do, even if you alter the bot's code in handling cube decisions.
Do you know why?

[Tim Chow]

On Friday, January 13, 2017 at 6:57:24 AM UTC-5, Michael wrote:
> Imo you should proceed to prove that the same is true on normal matches.

He has done this, but with very long matches (31 points or 33 points) where
he can apply the same strategy.

If you want to string him along further, ask him to do the same with five-point
matches. That should keep him busy for a while because it will take more effort
before he can find a cherry to pick.

---
Tim Chow

[Michael]

As long as he would not be able to hide any games or sessions that he played, then games of any length, be it 5 pointers, 10 pointers, 30 pointers would suffice.

Btw he already published 102 25 Pointers, 48 5 Pointers, and 49 15 Pointers that he played Vs GNUbg. Same results as always!

Since Murat is claiming the same things for 20+ years, I beleive it's about time he provides concrete proof that his results are genuine.

[Tim Chow]

On Friday, January 13, 2017 at 6:47:39 AM UTC-5, Michael wrote:
> On Friday, January 13, 2017 at 8:20:25 AM UTC+2, mu...@compuplus.net wrote:
>
> > Nothing will prevent running another copy of it from fresh.
>

> The third way is to use an encrypted file, you will know both it's name and
> location but you will not be able to delete it more than once otherwise it
> will start counting from 1 again and you will possibly need years to
> complete your 100 games session

This doesn't make any sense to me. Say I have two computers. I install on
Computer A and start playing. I manually record all the dice rolls that I
see. Then I do a totally separate and fresh install on Computer B and play,
now knowing all the dice rolls in advance. I can easily beat any bot this
way.

---
Tim Chow

[Michael]

Not at all.It doesn't use any specific seed, so it doesn't repeat the rolls.It is truly random.
I am halfway finishing it. in fact one of the buttons will enable you to save your own log (of rolls) anytime and as many times as you wish. what you will not be able to do though, is mess around with the encrypted file (containing all the history from day 1)that the program will automatically create and update by itself. Actually the only thing you can do about it is delete it...
Nothing will be secret -more details about it tomorrow.

[Michael]

Oh ..and you cannot join two separate encrypted files taken from 2 different computers. I already tried it ;-)

[Tim Chow]

On Monday, January 16, 2017 at 5:44:43 PM UTC-5, Michael wrote:
> I am halfway finishing it. in fact one of the buttons will enable you to save
> your own log (of rolls) anytime and as many times as you wish. what you will
> not be able to do though, is mess around with the encrypted file (containing
> all the history from day 1)that the program will automatically create and
> update by itself. Actually the only thing you can do about it is delete it...
> Nothing will be secret -more details about it tomorrow.

Then what I don't understand is this. Can't I study your code and figure
out what encryption algorithm it uses? If so, why can't I generate my own
dice rolls, and then write my own implementation of your algorithm to
generate an encrypted file that makes it look like I used your program to
generate the rolls and encrypt them?

---
Tim Chow

[Michael]

First of all the code will be in the executable so nobody would be able to read it.The encryption algorithm will also be in the executable and it might be a very simple one or one of the complicated ones that Microsoft provides. (it is actually a Microsoft one with some minor modifications). The password will also be unknown. That's too many unknown factors for someone to start an effort of cracking it.
Imo the encryption is strong enough that it might take someone years to crack it. Despite of that I admit there are still some things that worry me, but I won't tell publicly. Let's wait and see. They are some additional tricks to make it uncrackable if necessary.

[Tim Chow]

On Tuesday, January 17, 2017 at 8:52:48 AM UTC-5, Michael wrote:
> Imo the encryption is strong enough that it might take someone years to
> crack it. Despite of that I admit there are still some things that worry
> me, but I won't tell publicly. Let's wait and see. They are some additional
> tricks to make it uncrackable if necessary.

Do you save the checker plays and cube actions in the encrypted file as well,
or only the dice rolls? If you don't store the plays then it still seems to
me that I could play a session and generate an encrypted list of rolls, and
then manually copy the dice rolls into an ordinary copy of GNU (without your
add-on) and play it out knowing the rolls in advance.

---
Tim Chow

[Michael]

Damn it and I was telling myself to add one more button for cubes and I didn't.
No, the moves are not recorded,since the user should pass them on to the program and that would distract him too much.
It only records and encrypts the start (and end of each game) and who plays first. What you are saying is not possible because the normal use of the program includes the Black/or White plays first line.

However you just gave me a clue of how possibly someone could cheat. And that's by rolling a few rolls in advance... Perhaps Murat's idea to backup his evidence with a video recorder would still be a must.
Please continue this discussion on the other topic as I already uploaded the program.

[mu...@compuplus.net]

On January 14, 2017 at 9:02:42 AM UTC-7, Tim Chow wrote:

> He has done this, but with very long matches (31 points
> or 33 points) where he can apply the same strategy.
> If you want to string him along further, ask him to do
> the same with five-point matches. That should keep him
> busy for a while because it will take more effort
> before he can find a cherry to pick.

You are an idiot!

MK

[mu...@compuplus.net]

On January 17, 2017 at 10:11:37 AM UTC-7, Michael wrote:

> However you just gave me a clue of how possibly someone
> could cheat. And that's by rolling a few rolls in advance...
> Perhaps Murat's idea to backup his evidence with a video
> recorder would still be a must.

Michael, there is no hurry. Take your time. This an old
subject that has been hashed countless times...

Using XG's dice roller, which doesn't show or allow user
to set the seed, along with a real-time video capture is
probably the farthest you can go without requiring actual
physical presence. And it doesn't look like you are
offering anything better.

> Please continue this discussion on the other topic as
> I already uploaded the program.

You need to convince Chow and his ilk before we can move
ahead with your ideas...

MK

[Tim Chow]

On Wednesday, January 18, 2017 at 2:22:50 AM UTC-5, mu...@compuplus.net wrote:
> You are an idiot!

Come on, you can do better than that! You disappoint me, Murat. Where's the
fire? Where's the energy? Are you getting old and tired? Do I have to dangle
my math Ph.D. in front of you again to get you to do better? If this is the
best that you can do then the BGO monkeys might come and repopulate r.g.b.,
and then where will we be?

This is no time to give up. It is your duty to keep r.g.b. up to snuff. Get
off your backside and do some serious ranting. Now!

Oh, and you said "Chow and his ilk" instead of "Chow and his ilks". You must
really be losing your touch if you're making grammatical mistakes like that.

---
Tim Chow

[Dmitriy Obukhov]

On Friday, December 30, 2016 at 2:46:09 PM UTC-5, mu...@compuplus.net wrote:
> http://www.montanaonline.net/backgammon/xg.php
>
> I've been working at redoing the entire site to
> be more user friendly. I still want to add notes
> for each set of experiments, once I get another
> spurt of interest.
>
> The last expeiment at the above link took about
> 6 hours in total playing time.
>
> I captured the entire session onto video with a
> clock and dicedll dialog box always on the screen.

>
> I saved the session in 4 videos (25 games each)
> but in a single .xg file.
>

> Videos are at normal speed, with necessity pauses
> between games. On an Intel-i5 4-core 3.2ghz cpu,
> you can see that XG needed time to think ;) and
> used up most of the playing time. You can speed
> up the videos when playing back if you want.

>
> At the above link, there is also an exported .pxg

> user profile and a .pdf with my own analysis as a
> spreadsheet and two graphs (I hope you can make
> some sense out of them).

>
> The result was MK +43 with 18.76 cube error, 9.56
> checker error and insignificant +0.0013 luck.
>

> Feel free to comment on what you make out of it.
>
> MK

Why not just connect to someone's computer via services like Join.me or Webex or TeamViewer? You can control their XG from the distance and record everything simultaneously. Seems like this would solve the problem. No?

[mu...@compuplus.net]

On February 18, 2017 at 4:27:09 PM UTC-7, Dmitriy Obukhov wrote:

> On December 30, 2016 at 2:46:09 PM UTC-5, mu...@compuplus.net wrote:

>> I captured the entire session onto video with a
>> clock and dicedll dialog box always on the screen.

> Why not just connect to someone's computer via services like
> Join.me or Webex or TeamViewer? You can control their XG from
> the distance and record everything simultaneously. Seems like
> this would solve the problem. No?

Yes! Excellent suggestion. Although somenone will still need to
contribute their time and a "server" machine, the overall effort
needed would be much less than other methods previously suggested
and technically this should eliminate concerns about cheating, etc.

In the past, videotaping and instant messaging had been talked
about separately and were considered to be either not secure
enough or too cumbersome.

Video capturing, which should be proving enough by itself in my
opinion, was not as feasible back then either.

Nor was remote access/control so easy and available, which I now
use regularly (i.e. Teamviewer) to help my customers.

Your suggestions combines the two and would be acceptable to me.

Let's see what "the church flock" says...??

MK

[Dmitriy Obukhov]

I have a joun.me acount and would be willing to set it up so Murat is remotely controls my computer and several people are able to watch in real time and make comments. Let me know if interested.

[mu...@compuplus.net]

As I have time to dedicate to RGB/backgammon again,
checking this thread was the first the first thing
that I did.

I had left it at saying:

"Your suggestions combines the two and would be acceptable
"to me. Let's see what "the church flock" says...??

Amazingly, my response was the last post in this thread.

Just when I thought we found a way to conduct an experiment
in a way that would be acceptable to everyone, everyone
except me have gone silent...?

Doesn't surprise me a bit.

Let me do my part once more by repeating that Dmitriy's
suggestion is acceptable to me with one conditional
reservation which is that if having spectators becomes
d'stracting to me, I should have the option to play by
screeing them out, at least by a "one-way-mirror"...

Initially, however, not only that I will not object to
it, I would like to extend my personal and cordial
invitation to the cocksucking, cheating thief to be our
guest of honor... ;)

Come on, already, folks. Let's get at it!

MK

[Michael]

Probably nobody is interested in subjects whose responses are 1 or 2 months apart...

[Michael]

The ball was actually in your hands. Dimitri proposed you something to which you never responded. Get in touch with him, make it work, and then let us watch a short trial session. Only after seeing what it really is and how it works, the rest of us could ever express an opinion.

[Tim Chow]

On Tuesday, May 2, 2017 at 3:10:16 AM UTC-4, mu...@compuplus.net wrote:
> On March 8, 2017 at 1:47:30 PM UTC-7, Dmitriy Obukhov wrote:
>
> > I have a joun.me acount and would be willing to set
> > it up so Murat is remotely controls my computer and
> > several people are able to watch in real time and
> > make comments. Let me know if interested.

[...]

> Amazingly, my response was the last post in this thread.

Amazingly, you simultaneously quote Dmitriy's post, which was the last post
in this thread, and claim that your own post was the last post. Idiotic even
by your standards.

---
Tim Chow

[mu...@compuplus.net]

On May 2, 2017 at 7:19:49 AM UTC-6, Michael wrote:

> Probably nobody is interested in subjects whose
> responses are 1 or 2 months apart...

Dmitriy suggested using services like Join.me or
Webex or TeamViewer on February 18.

I responded to him only six days later on February
24 by saying "Yes! Excellent suggestion.... Your

suggestions combines the two and would be acceptable
to me. Let's see what "the church flock" says...??"

Nobody responded.

On March 8, Dmitriy posted again offering the use of
his computer, referring to me (not addressing me!),
and finished by asking the readers: "Let me know if
interested".

Nobody responded.

Even as of today, nobody else except me responded to
the point.

MK

[mu...@compuplus.net]

I knew exactly what he was talking about and
I had responded to him. There is no question
about whether it will work and most people in
here should already know it. If you didn't,
you should have made your above comment back
then.

Regardless, better late than never. Actually,
I like the ide of doing some demostration games
without the purpose of winning but with the
purpose of exposing what kind of cheating shit
XG is and its developers, contibutors, admirers,
etc. must be.

I think it would be an excellent way to brake
the ice for everyone to be involved and set the
tone of the experiment...

MK

[mu...@compuplus.net]

On May 2, 2017 at 5:55:59 PM UTC-6, Tim Chow wrote:

>> Amazingly, my response was the last post in this thread.

> Amazingly, you simultaneously quote Dmitriy's post,
> which was the last post in this thread, and claim
> that your own post was the last post.

You could have read it as it was intended (but misworded)
to say "my post was the last reponse in this thread" (to
Dimitriy's suggestion).

But you had to be your cocksucking self to post without
saying anything about the actual subject at hand... :(

> Idiotic even by your standards.

Whatever bullshit... What do you say about Dimitriy's
suggestion?

Obviously you prefer to distract from the real subject
because you have since acquired personal, insider info
that your idol is indeed a cheater and a thief, who keeps
his silence in order to avoid being a liar also... But
fortunately for some, unfortunately for others, that same
silence says enough.

MK

[Tim Chow]

On Thursday, May 4, 2017 at 1:57:42 AM UTC-4, mu...@compuplus.net wrote:
> On March 8, Dmitriy posted again offering the use of
> his computer, referring to me (not addressing me!),
> and finished by asking the readers: "Let me know if
> interested".
>
> Nobody responded.
>
> Even as of today, nobody else except me responded to
> the point.

Sure they responded. Silence is a response. It means, "Not interested."

On a totally different topic, I will remark that it is amazing that trolls
are amazed when nobody is interested in them.

---
Tim Chow

[mu...@compuplus.net]

On May 4, 2017 at 12:52:16 PM UTC-6, Tim Chow wrote:

> On May 4, 2017 at 1:57:42 AM UTC-4, mu...@compuplus.net wrote:

>> Even as of today, nobody else except me responded to
>> the point.

> Sure they responded. Silence is a response. It means,
> "Not interested."

Fair enough. However, it would be more accurate to say
"No longer interested"...

> On a totally different topic, I will remark that it is
> amazing that trolls are amazed when nobody is interested
> in them.

It was you indeed you who was trolling along unsuspecting
sincere people, knowing that you could always fall back
on a security, etc. excuse to weasel out of any experiment
proposed in the past.

This time you know directly from Mr. Ed's mouth or have a
better founded expectation of what the results will be and
so you can't troll no more (i.e. "you lost interest" :))

The results will flap you people's "flat bot world of bg"
like a pancake and there is nothing you can do to prevent
it.

If Dmitriy delivers on his offer, I would be willing to do
this experiment with just the two of us, hoping that his
credibility in the bg circles will be good enough.

MK