On Sep 16, 10:56 am, Taylor Kingston wrote:

Nonsense. Yesterday afternoon I started a game between your program
and Fritz8. GetClub moved so slowly I could not finish the game until
this morning, but essentially the game was over by the 7th move, when
GetClub threw away a piece for almost nothing, and then played so
idiotically that it lost any chance for compensation. Here is the
game:

White: GetClub
Black: Fritz 8

note: time used, in minutes and seconds, given after each move:

1.Nf3 10:00 d5 :00 2.d4 6:10 Nf6 :00 3.Bg5 18:39

Notice the long think. Sanny's program repeatedly overstepped the
self-announced time limit of 5-10 minutes per move.

A nice diversion to avoid recognition of the fact that
here, GetClub stands well due to her solid opening play.



3...Ne4 0:00 4.Qd3? 6:03 Nxg5 0:19 5.Nxg5 8:40 e5 0:00 6.h4 13:31 h6
0:23 7.Nxf7?? 13:32 Kxf7 0:04

Basically, the game is already decided. White might have had some
slight chance if he'd followed up the knight sac with 8.dxe5, but
Sanny's program proceeds from bad to worse.

8.Nc3? 13:52 e4 0:10 9.Qb5? 5:58 c6 0:00 10.Qb3 9:47 Qb6 0:00

Good, straightforward chess technique. When you're ahead in
material, seek to exchange queens. White must oblige or lose the d-
pawn.

What about 11.e3, guarding the d-pawn and
stubbornly refusing to trade Queens? If the
opponent initiates the trade, it is a psychological
coup; in effect, Fritz must admit that he can't
handle the tactics with Queens on the board.


11.Qxb6 6:27 axb6

An excellent move! GetClub is a tactical
genius, sometimes selecting the same move
as Bobby Fischer almost certainly would.



0:02 12.Nd1 14:05

A ludicrous spectacle. Not only is White a piece down, but he's lost
control of the center and does not have a single piece developed.

Read Nimzowitch; one need not always occupy
the center, or rush development if the center is
closed.


12...c5 0:17 13.Ne3 10:32 cxd4 0:10 14.Nxd5 14:59 Be6 0:00 15.c4 13:07
Bd6 0:00 16.g4? 16:39 b5 0:05 17.e3 7:18 bxc4 0:00 18.Bxc4 6:02 Rc8
0:00 19.b3??

Another classic example of anti-GetClub annotation.

Objectively, this is just about as good a way to lose as any.
While most humans would prefer to "attack something" with
....Nb6, there is little objective reason to prefer one losing
technique to another. In fact, many grandmasters are known
for being poor at losing brilliantly. You take GM Kramnik, for
instance; in one game he allowed a mate-in-one on himself,
when there were myriad better ways to lose!


9:49 b5 0:00

Most human players would probably have gone for more simplification,
but Fritz8 has enough calculating ability, even at very quick time
limits, to handle this more complicated line.

20.Nb6 11:27 bxc4 0:00 21.Nxa8 9:15 dxe3

Look at this: no query for inferior moves, if *Fritz*
is the one making them. My computer shows that
....Bb4+ is quite superior: 22. Kf1 Nd7, 23. ed Rxa8
(+5 for Black since the Knight does not escape).


0:00 22.O-O

An error. Now the Knight is trapped for sure.


11:26 Nd7 0:04
23.fxe3+

Another brilliant tactical blow by GetClub! Too bad
the attack will be made without any army left.


23:15 Ke7 0:00 24.Rf2 9:11 Rxa8 0:04 25.g5

A serious error, as now the pin-- a8-a1 -- will allow
Fritz a quick and smooth finish, like a Pinot Grigio,
1948.


7:38 cxb3 0:02
26.a4 8:23 Bg3 0:00 27.Rd2 7:51 Bxh4 0:04 28.gxh6 7:45 Ne5 0:00

Most human players again would probably just chop wood, but Fritz is
weaving a mating net.

Well, perhaps it is, but Fritz obviously lost the thread
when it gave up on promoting the b-pawn and switched
to grabbing irrelevant pawns on the far wing. IMO, White
should already be down to one Rook and completely
defenseless about here.


29.Kh1 13:16 Nf3 0:03 30.Re2 14:52 Rh8 0:00 31.hxg7 16:23 Rh6 0:04
32.Kg2 8:30

At this point Fritz announced mate in at most 8.

Yes, well, that may be because White has defended
poorly; my computer sees no forced easy mate after
better defensive moves. In other words, was Fritz
really "weaving a net" or did GetClub just trip and
catch itself?


32...Rg6+ 0:00 33.Kh1 15:56 Bf6 0:06

Here, my computer shows a forced mate via ...Be1.
If Fritz misses shallow mates, how can he have been
"weaving a net" many moves back?


Here I shut down for the night, and restarted the game this morning,
hence the "0:00" time for White's next move

34.g8=R 0:00

Seems odd to promote to a rook, but that's what Sanny's program did.
Doesn't matter anyway. The way Fritz now completely ignores the newly
promoted piece is amusing.

34...Rh6+

Amusing perhaps, but after this lemon my computer
shows the quickest forced mates have evaporated into
thin air. In fact, Fritz seems to almost be struggling
not to *lose ground* here, in terms of "mate-in-x".


0:00 35.Kg2 6:18 Rh2+ 0:00 36.Kf1 6:23 Bc4

Finally, a good move my program did not foresee.
We're back to the quickie mates again.


0:00 37.Rae1 9:42
Bh4 0:00 38.Rg7+ 9:41 Kf8 0:00

And here Sanny's program resigned; after the spite check 39.Rf7+ Kxf7
it's mate next move.


I trust this satisfies Sanny's curiosity about the relative merits
of his program versus Fritz8, and puts paid to the notion that his
program is "improving." I will waste no more time on this issue.

Sometimes, Sanny's program does really stupid
things like the Knight sac' in this game. But other
times, it plays rather well for a long time. The
problem is that it only takes one such blunder for
the program to lose the game, and this is why I
keep telling Sanny to get his tactics up to speed.

I believe it was Sam Sloan who recommended the
book "Pawn Power in Chess" to Sanny, but if you
play the program you will see that he can have all
of that programmed in and it would still lose, as in
the above game, on elementary tactics.


-- help bot