Computerchess Misc (11)
Rolf Tueschen
Kg1 Qg4 Ra1 Re3 Nf5 Ng3 Pb2 b3 e4 f2 g2 h2
Kh7 Qd7 Rc8 Rd8 Bb7 Bf6 Pa6 b5 d6 f7 g6
BTM; Last move was Re1-e3.
This is a position I had with the Whites.
My opponent played 1...Rc2.
Now Fritz4 still needed over 1 hour to find the move I played. The
FRITZ5.32 needs only 50 seconds. Which is still a lot of time, I needed
also a longer time, over 15 minutes.
2. Nh5 with a winning advantage.
In a first test Crafty 16.5 had played:
1...Bxb2
Now this is interesting. Fritz 5.32 plays immediately:
2.Rd1+-
Crafty needs for this move 8'37'' It always prefers to play:
[2.Qh4+
but that doesn't lead to a clear advantage 2...Kg8 3.Ne7+ Kf8 4.Nxc8
4...Bxa1 5.e5 Rxc8 6.Qh8+ Ke7 7.exd6+ Kxd6 8.Qxa1 Kc7 9.Qe5+ Qd6
0.41/10]
Back to 2. Rd1:
2...Kg8
[Crafty 16.5: 2...Rc1 3.Qh4+ Kg8 4.Rxc1 Bxc1 5.Qf6! strong move, but
Crafty cannot see him long enough before to make the choice early for
2.Rd1]
3.e5! +-
-------------------------------------------------------------------------------------------
I chose this position for Misc because I was somewhat surprised how long
the tactical monsters need. And Crafty 16.5 is not at all satisfying.
When I repeated the analysis for the article I saw that some former
statements were not so purely true. Either because of the exact mode in
ChessBase (I always analyse with the modules in ChessBase) or due to
differences in the RAM.
Can you confirm my impressions? On faster hardware, are the programs
more similar in their solutions or do the differences I observed still
exist?
Is this a typical (good) position for our machines?
Miranda
of Re1-e3. (my pc is IBM 600 notebook PII266Mhz)
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote in message
news:7eotae$tlq$6...@news01.btx.dtag.de...
Rolf Tueschen
<7epjpp$aqp$1...@reader1.reader.news.ozemail.net>:
>I use ChessBase with Junior 5 engine, it takes 4 seconds to find the move
>of Re1-e3. (my pc is IBM 600 notebook PII266Mhz)
Yes, well possible, but how about the speed it finds Rd1 after 1...Bxb2?
Or Nh5 after 1...Rc2?
Could you give us the results on your high-speed machine?
Another question, and most of all the reason for my posting, for me as a
layman in computer chess techniques, why is the move Nh5 so much more
difficult than Re1-e3?
Explanation for this and also the chessic situation as such, that is
what I asked for. It's interesting to see how one's own games change
their "spirit" in view of always new and stronger machines and programs.
I wished more people would report their experience. Analysing own games
is the most important thing to become stronger!
Thanks for your contribution.
Carl Tillotson
> Kg1 Qg4 Ra1 Re3 Nf5 Ng3 Pb2 b3 e4 f2 g2 h2
> Kh7 Qd7 Rc8 Rd8 Bb7 Bf6 Pa6 b5 d6 f7 g6
It would help if you post the FEN (no excuses from you, you do have
Chessbase!), I had setup a pawn on e6 when I first had a go at this and
wondered why Rd1 didn't lose straight away.
Anyway, the FEN I have know is :
3r4/1b1q1p1k/p2p1bp1/1p3N2/4P1Q1/1P2R1N1/1Pr2PPP/R5K1 w - - 0 1
Anyway.....
On my humble machine, Fritz finds Nh5 in 15 seconds (5827KN) with an
evaluation of +0.69 which has turned to +3.16 after 28 seconds (11302KN)
Your evaluation appears to be off, unless I am missing the thread of your
analysis (which is easy since you spin off on tangents as per usual), after
Nh5 if Black plays Bxb2 he loses after Qh4 - he can't handle the threat of
Nf6+ which leads to mate or lost of Queen.
However, let's assume that you are saying Bxb2 instead of Rc2......
Yes Rd1 is played instantly.
Now we come to the crux of the matter. What purpose does this thread serve,
apart from confirm that different kit makes moves quicker. Or do I detect
another little swipe at Crafty somewhere. RH has left the NG and STILL RT
wants to keep the anti-crafty jibes going.
So I decided to fire up Crafty myself, and to be fair decided it needed
just about the same amount of time as Fritz (maybe more, since Fritz is a
Chessbase product and Crafty is not but is interfaced). So after 17 seconds
Crafty decides that Rc2 is too risky and plays Kg8 instead - obviously
aware of "King Safety", and there it stayed Kg8 instead is preferred.
So I am now thinking to myself, hang on a minute Carl, if Rolf's Fritz
found Nh5 why can't his Crafty find Kg8 after all the timeslice is about
the same. Maybe Rolf forced Crafty to follow Bxb2.....
You have to admit, Kg8 blows a lot of holes in your analysis.
--
Carl Tillotson
Lancashire Chess Association
Rolf Tueschen
<VA.00000045.00b5b78b@carl>:
>Anyway, the FEN I have know is :
>3r4/1b1q1p1k/p2p1bp1/1p3N2/4P1Q1/1P2R1N1/1Pr2PPP/R5K1 w - - 0 1
>Anyway.....
Thanks, anyway for all the comments. I will analyze Kg8 now for myself.
This is really interesting. Carl, a short remark in general. We have
different machines and by force we have different evaluations and speed.
If you ask me why Crafty at all, then I can only say that I'm a user of
Fritz and Crafty modules. I have Junior 4.6 too. Bbut it's not my style
to use the tools mecahnically. It's the game that interests me. And
Crafty is always updated by ChessBase, I hope you know that, and
therefore I'm mostly interested in _new_ implementations like that King
safety feature.
If you ask for the purpose at all I must admit that I'm too weak myself
to analyze my own games properly enough. And it's like magic for me when
the little programs suddenly approve or reject moves I had thought over
myself in my youth. This is so satifying. But it takes a lot of time,
because you must give them at least some 3 or 5 minutes per move. So, in
the meantime I have "flagged" my games with the CB symbols and some have
that golden sign of real masterpieces. What am I taliking about here,
you tore me into the public, but that is totally private. No, I have
these games with in the meantime some 1000 extra moves of analysis
beginning with Mephist II and then Exclusive, MMIV and Fritz1 until Deep
Yellow. And as I told you in the post here FRIT4 still needed 1 hour and
plus for Nh5. I think such informations are also interesting like the
pure autoplayer tests, NO?
Just a question to you. I don't understand you with the FEN. If I have
such a game with bursting analyses. I cut it until the position, bring
the interesting line in the front and then store it with Crtl-C.
In my newsreader I paste it with Crtl-V. Now tell me when should I get
the FEN and how?
Thanks anyway for your nice reaction.
Rolf Tueschen
<7er9eu$9nr$1...@news00.btx.dtag.de>:
>Carl Tillotson <l...@lancashirechess.demon.co.uk> wrote in
><VA.00000045.00b5b78b@carl>:
>>Anyway, the FEN I have know is :
>>3r4/1b1q1p1k/p2p1bp1/1p3N2/4P1Q1/1P2R1N1/1Pr2PPP/R5K1 w - - 0 1
>>Anyway.....
Carl, I made a big mistake. I answered you in general while I was still
online. As I said I wanted to examin Kg8 myself.
But when I had the game in front of me I was shocked.
Carl, you are really a nasty chessplayer. Your FEN is totally wrong.
PLEASE RE-READ MY FIRST ARTICLE AGAIN!
There I described the position very exactly. What you made out of it is
a scandal.
More tomorrow morning.
Thomas F. Mooney, III
Rolf Tueschen wrote in message <7erlk7$jhc$1...@news02.btx.dtag.de>...
Rolf,
I am responding as a neutral observer.
I set up the position you described using WinBoard, played the move you
described ...Rc2 and achieved the identical position posted in Carls' FEN.
3r4/1b1q1p1k/p2p1bp1/1p3N2/4P1Q1/1P2R1N1/1Pr2PPP/R5K1 w - - 0 2
FWIW,
TFM3
Note: spam-resistant e-mail address
Carl Tillotson
> In my newsreader I paste it with Crtl-V. Now tell me when should I get
> the FEN and how?
When you have an interesting position on the board (in Chessbase) click on
the SETUP postion and then click on COPY_FEN which is along the COPY_ASCII.
Thought you would have known that trick. Believe me it is much easier for
the NG to follow since many readers can read FEN no problem.
The problem with the old style of writing down all the squares occupied by
a piece is that it has to be transposed back by the end user. Hence
mistakes can occur. Like in my first attempt I was switching from my
newsreader to chessbase and back again setting up the position. In the
course of doing it I put the pawn on e6 instead of d6. Looked at it and
thought your analysis was cocked up :-)
Anyway FEN is easier for us lazy sods who need to just cut&paste :-)
Carl Tillotson
> But when I had the game in front of me I was shocked.
>
> Carl, you are really a nasty chessplayer. Your FEN is totally wrong.
Whoops ! Well how was I to know I cut and pasted after I played Rc2 :-)
Your fault you know :-)
Anyway the new one should be......
3r4/1b1q1p1k/p2p1bp1/1p3N2/4P1Q1/1P2R1N1/1Pr2PPP/R5K1 w - - 0 1
--
Rolf Tueschen
>Rolf,
>> But when I had the game in front of me I was shocked.
>>
>> Carl, you are really a nasty chessplayer. Your FEN is totally wrong.
>Whoops ! Well how was I to know I cut and pasted after I played Rc2 :-)
>Your fault you know :-)
>Anyway the new one should be......
>3r4/1b1q1p1k/p2p1bp1/1p3N2/4P1Q1/1P2R1N1/1Pr2PPP/R5K1 w - - 0 1
Carl, actually I don't have my CB ready. Would you please correct your
false FEN as follows. Black is on the move! The last white move was Re3.
Black has two moves, at least these I analysed. They are Rc2. That was
the move my opponent played. And Bxb2. Then you gave me the idea of Kg8.
Which in fact is played by the new Crafty 16.6 version. But I had it
only yesterday late in the night. 16.5 doesn't play Kg8.
This is in fact a fantastic move. Black tries to run away and IF he
succeeds he has survived the worst. But look for yourself. You will find
interesting things with Crafty. But I warn you. You must rely on your
own chess instinct. If you follow Crafty like a jackass, he will lead
you in delusions of desperation. It will be up and down. And only if you
KNOW who really should have advantage in such positions you will reach
the end. Little hint. Doesn't there ring a bell that Rolf had the
Whites????!!!! You see, I won't deceive you in the middle of the night,
Carl ...
You won't find Crafty 16.6 playing Bxb2. But anyway try to find the
refutation. It'as interesting too. THat was the purpose of my post. The
played Rc2 is answered with the almost genial Nh5. But I am not
imposturing. The fraky machines all play that today in under 1 minute.
While I still needed >15'. For Re1-e3 I needed much less.
I repeat. After Bxb2 Crafty always preefers for a long time Qh4 check.
FRITZ5.32 plays a tempo Rd1.
My question again for all savys, WHY the difference between the two? Is
Crafty more fixed on the King? Has Fritz more knowledge for a
tactico-positional two-sided attack by the nippers? [...]
Enough for you to research, Carl!
Rolf Tueschen
<7ete5c$ucu$1...@news.minn.net>:
>I am responding as a neutral observer.
>I set up the position you described using WinBoard, played the move you
>described ...Rc2 and achieved the identical position posted in Carls' FEN.
Yes, but the move Rc2 was only the game continuation. I wanted to
analyse if Black had missed a defense. I had the typical computer "move"
Bxb2 which is also not bad. And Carl showed Kg8 directly, although he
had it in the wrong row. But directly Kg8 is a fantastic try to run away
over Ke8 --- etc. However, White has a strong remedy with his Rooks.
And the KNight against d6.
Carl is right, next time I'll give you directly the FEN position and
afterwards I can still paste the variations. Shit, I'm so much fixated
on the content that I'm almost phobi-ing against technical new tools.
You can never know if they function. IF NOT you lose all the precious
time ... Can you understand my sorrow?
The same with my ancient FreeAgent. I once had a new version on a CD. I
installed it and had to enter all the proxies and stuff manually. In the
end it didn't bring me online. Always a request that I couldn't give.
And finito. Fortunately I had chosen a different directory so that I
could come back to my version as quickly as possible. Tell your techno
freinds that they should make automatical installation with questions
like: Sir, oh all too famous guru! What do you want shall I do next? Do
you want me to impute the proxies for you NOW??
I would give my answer like this: Cool man, just feel free, but let me a
copy of the old version on the board, right?!
He answers: Ay ay Sir. From now on you have the MacIntosh desktop.
OH NO YOU BLOODY FOOL! Click. Push on the red button and leave the
system alone. AC/DC OUT! Mom ---, please help me .........!
You get the idea, how _fragile_ my whole virtual reality is ... :)
You must know I'm more a philosopher. I was born with two left hands! If
you knew how many clocks I already tried to repair --- and they are
still in a big cupboard because I never had the _time_ to put them
together again. But I promise if I have my next flu I will do my best
and reconstruct the little machines.
Carl Tillotson
> Enough for you to research, Carl!
When I have the time, at the moment I have none at all. Too busy.