Zenmastur vs Ovyron No it's not another GROB II!

[abulmo2]

I think your making a mountain out of a mole hill. On the other hand, I hope you like mole hills as you're likely to come across quite a few of them in this position!

Here is "Le Môle", a mountain in France (Haute Savoie).

[Ovyron]

Who knows, maybe you'll get lucky!

The variations I analyze don't matter. What matters is that I understand what I need to do to win. And if I do it, then luck will not be involved, as I could beat you every single time, like with 1.g4.

[SetUp "1"]
[FEN "4r3/5pk1/1q1r1p1p/1p1Pn2Q/1Pp4P/6P1/5PB1/R3R1K1 b - -"]

1... f5 2. Qd1 Rdd8 3. Re3 f4

[d]3rr3/5pk1/1q5p/1p1Pn3/1Pp2p1P/4R1P1/5PB1/R2Q2K1 w - -

[Zenmastur]

Who knows, maybe you'll get lucky!

The variations I analyze don't matter. What matters is that I understand what I need to do to win. And if I do it, then luck will not be involved, as I could beat you every single time, like with 1.g4.

[SetUp "1"]
[FEN "4r3/5pk1/1q1r1p1p/1p1Pn2Q/1Pp4P/6P1/5PB1/R3R1K1 b - -"]

1... f5 2. Qd1 Rdd8 3. Re3 f4

[d]3rr3/5pk1/1q5p/1p1Pn3/1Pp2p1P/4R1P1/5PB1/R2Q2K1 w - -

Really?

I didn't have this as a line of play. I suspect there is reason I didn't have this move in my analysis, so it shouldn't take long for me to bust this line of play...

[Ovyron]

I wanted to see what it looked like from your perspective, so I loaded 5-man syzygy and did a multi-PV 4 on the current board position to see what it would come up with if I let it run a while. I figure if you let your computer run for a couple of weeks the output will look something like this:

Code: Select all

[-2.50 ]  Depth: 86/116  Nodes: 1969180574K Hash: 100% TB hits: 2135292945 Current:(4/41)

Notice that the score hasn't increased from your original score of -2.50.

Note that I got 3...f4 to a -2.50 score in less than an hour of analysis (from the root position before 1...f5) before the game started so this post of yours basically proves high depth useless here and big hardware unnecessary

[Zenmastur]

I wanted to see what it looked like from your perspective, so I loaded 5-man syzygy and did a multi-PV 4 on the current board position to see what it would come up with if I let it run a while. I figure if you let your computer run for a couple of weeks the output will look something like this:

Code: Select all

[-2.50 ]  Depth: 86/116  Nodes: 1969180574K Hash: 100% TB hits: 2135292945 Current:(4/41)

Notice that the score hasn't increased from your original score of -2.50.

Note that I got 3...f4 to a -2.50 score in less than an hour of analysis (from the root position before 1...f5) before the game started so this post of yours basically proves high depth useless here and big hardware unnecessary

Well, I was of the opinion before we started that almost all moves win. So, the fact that this wasn't in my analysis might just mean it wins slower than the best move. I guess we'll see after I've analyzed it.

[Ovyron]

Well, I was of the opinion before we started that almost all moves win. So, the fact that this wasn't in my analysis might just mean it wins slower than the best move. I guess we'll see after I've analyzed it.

So it never crossed your mind that it may be the fastest winning move and that after it anybody would easily win with it (specially people with slow hardware), huh?

The level of prejudice is incredible, but if this stands for a "move missed by guy on big hardware while guy on slow one finds it in less than one hour" example, I'm all for it!

[zullil]

Well, I was of the opinion before we started that almost all moves win. So, the fact that this wasn't in my analysis might just mean it wins slower than the best move. I guess we'll see after I've analyzed it.

So it never crossed your mind that it may be the fastest winning move and that after it anybody would easily win with it (specially people with slow hardware), huh?

The level of prejudice is incredible, but if this stands for a "move missed by guy on big hardware while guy on slow one finds it in less than one hour" example, I'm all for it!

It looks like a fast way to throw away your win, but I've only analyzed for about an hour.

[Uri Blass]

I wanted to see what it looked like from your perspective, so I loaded 5-man syzygy and did a multi-PV 4 on the current board position to see what it would come up with if I let it run a while. I figure if you let your computer run for a couple of weeks the output will look something like this:

Code: Select all

[-2.50 ]  Depth: 86/116  Nodes: 1969180574K Hash: 100% TB hits: 2135292945 Current:(4/41)

Notice that the score hasn't increased from your original score of -2.50.

Note that I got 3...f4 to a -2.50 score in less than an hour of analysis (from the root position before 1...f5) before the game started so this post of yours basically proves high depth useless here and big hardware unnecessary

No

This -2.50 score is only with 5-man syzygy tablebases.

Good hardware means using also 6 and maybe 7-man syzygy tablebases during the search.
The fact that high depth without more tablebases does not help does not mean that big hardware does not help because big hardware is not only high depth but also using more tablebases.

[Zenmastur]

Well, I was of the opinion before we started that almost all moves win. So, the fact that this wasn't in my analysis might just mean it wins slower than the best move. I guess we'll see after I've analyzed it.

So it never crossed your mind that it may be the fastest winning move and that after it anybody would easily win with it (specially people with slow hardware), huh?

The level of prejudice is incredible, but if this stands for a "move missed by guy on big hardware while guy on slow one finds it in less than one hour" example, I'm all for it!

It looks like a fast way to throw away your win, but I've only analyzed for about an hour.

I had the position before f5 was played at a very high score for black. I've never seen a game drawn from a position with the type of evaluation unless the winning side blundered. That move dropped the score into a range where it's possible to draw. That doesn't mean it is a draw but the possibility is there. So, it's clearly an inferior move. If he makes another move like that one a draw is likely assured. In the mean time I'm trying to optimize his probability of making a similar mistake, which takes a lot of extra calculation time.

I wanted to see what it looked like from your perspective, so I loaded 5-man syzygy and did a multi-PV 4 on the current board position to see what it would come up with if I let it run a while. I figure if you let your computer run for a couple of weeks the output will look something like this:

Code: Select all

[-2.50 ]  Depth: 86/116  Nodes: 1969180574K Hash: 100% TB hits: 2135292945 Current:(4/41)

Notice that the score hasn't increased from your original score of -2.50.

Note that I got 3...f4 to a -2.50 score in less than an hour of analysis (from the root position before 1...f5) before the game started so this post of yours basically proves high depth useless here and big hardware unnecessary

No

This -2.50 score is only with 5-man syzygy tablebases.

Good hardware means using also 6 and maybe 7-man syzygy tablebases during the search.
The fact that high depth without more tablebases does not help does not mean that big hardware does not help because big hardware is not only high depth but also using more tablebases.

True enough. It also means I can look at many variations that would normally take way too much time.
e.g. the Multi-pv I did to deapth 86. I purposely crippled my hardware to make it act more like Ovyron's. Even so it took less than an hour. That's why it only showed -2.50. The “real” evaluation, using the full potential of the hardware was several times that value. Of course, this isn't good for me, but I knew that before we started, so no big surprise there.

1... f5 2. Qd1 Rdd8 3. Re3 f4 4.gxf4

[d]3rr3/5pk1/1q5p/1p1Pn3/1Pp2P1P/4R3/5PB1/R2Q2K1 b - - 0 4

Regards,

Zenmastur

[Zenmastur]

Edit:

I had the position before f5 was played at a very high score for black.

This should read "before f4 was played".