![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Pushing wood...
Aug. 24th, 2014 07:41 pmSo from time to time, I enjoy going over to lichess.org to do a little "training," which is to say the site challenges me with a chess position where I must find the best sequence of moves.
When I first started at the site (free to sign up, free to play, all suitable disclaimers uttered), my default rating was 1500 for all forms of play (slow, fast, blitz, bullet, and training), and pretty much every training position required me to find a mate. The only problem with this approach is that if you know you're supposed to find a mate, then you won't give up—or at least I won't—until it's found.
Gradually, however, as my training rating inched up, a couple of things happened. First, the positions don't all require you to find a mate, which is heaps more realistic. Second, I began to notice subtle—and some not-so-subtle—issues with the all-or-nothing way the site scores one's solutions. Here's a case in point.
It's Black to move in the following position, shown "naturally" with Black at the bottom of the board. The square a1 is in the upper right-hand corner of the graphic; h6 is in the bottom left-hand corner.
Looking for White's potential threats In this position, I don't see anything major. It's not as if Black is facing a mate in one or anything like that. Still, it wouldn't do for Black to waste time.
There is the obvious candidate move 1...Qxh2+, which forces 2. Kg1. Black can then play 2...Qh1+, which forces the White King to emerge from behind the pawns with 3.Ke2, but now the Black Queen is under attack by the rook on d1, so Black needs to keep the ball rolling with an incessant attack against the White King. This can be done with 3...Rhe8+, which forces 4. Kd2. (And while this may look like I'm thinking several moves ahead, it's not, because so far, every one of White's responses has been forced.)
It's really a pretty position to consider, and if this kind of thing is of interest to you, clicking on the position should take you to the site. (FWIW, the rest of my solution is at the end of the post.)
The thing that bugged me about this position—and which caused me to a full lose 10 points while solving it—was something that happens all too often with me at the site: I get to the point where I see a clear mate in a couple of moves or so, but because my sequence is longer than the optimum solution (e.g., I see a mate in two where there's a mate in one), the algorithm decides my solution has "failed"
This isn't something that will keep me from using the site, but it is annoying as all getout, from time to time.
Cheers...
My solution(s), starting with 1...Qxh2+ 2. Kf1 Qh1+ 3. Ke2 Rhe8+:
4...Be5+ 5. Bd5 Rxd5+ 6. Ke2 Bg3+ (which I thought was pretty) 7. Kf3 Qh5+ 8. Kxg3, whereupon I saw 8...Qh4+ and 9...Qf4 mate, but 8...Ne4 mate wins quicker!
I liked 6...Bg3+ because it kept White's g pawn from moving, but it turns out that 6...Bd6+ also works: 7. Kf3 Qh5+ 8. g4 Qh3 mate!
When I first started at the site (free to sign up, free to play, all suitable disclaimers uttered), my default rating was 1500 for all forms of play (slow, fast, blitz, bullet, and training), and pretty much every training position required me to find a mate. The only problem with this approach is that if you know you're supposed to find a mate, then you won't give up—or at least I won't—until it's found.
Gradually, however, as my training rating inched up, a couple of things happened. First, the positions don't all require you to find a mate, which is heaps more realistic. Second, I began to notice subtle—and some not-so-subtle—issues with the all-or-nothing way the site scores one's solutions. Here's a case in point.
It's Black to move in the following position, shown "naturally" with Black at the bottom of the board. The square a1 is in the upper right-hand corner of the graphic; h6 is in the bottom left-hand corner.
Looking for White's potential threats In this position, I don't see anything major. It's not as if Black is facing a mate in one or anything like that. Still, it wouldn't do for Black to waste time.
There is the obvious candidate move 1...Qxh2+, which forces 2. Kg1. Black can then play 2...Qh1+, which forces the White King to emerge from behind the pawns with 3.Ke2, but now the Black Queen is under attack by the rook on d1, so Black needs to keep the ball rolling with an incessant attack against the White King. This can be done with 3...Rhe8+, which forces 4. Kd2. (And while this may look like I'm thinking several moves ahead, it's not, because so far, every one of White's responses has been forced.)
It's really a pretty position to consider, and if this kind of thing is of interest to you, clicking on the position should take you to the site. (FWIW, the rest of my solution is at the end of the post.)
The thing that bugged me about this position—and which caused me to a full lose 10 points while solving it—was something that happens all too often with me at the site: I get to the point where I see a clear mate in a couple of moves or so, but because my sequence is longer than the optimum solution (e.g., I see a mate in two where there's a mate in one), the algorithm decides my solution has "failed"
This isn't something that will keep me from using the site, but it is annoying as all getout, from time to time.
Cheers...
My solution(s), starting with 1...Qxh2+ 2. Kf1 Qh1+ 3. Ke2 Rhe8+:
4...Be5+ 5. Bd5 Rxd5+ 6. Ke2 Bg3+ (which I thought was pretty) 7. Kf3 Qh5+ 8. Kxg3, whereupon I saw 8...Qh4+ and 9...Qf4 mate, but 8...Ne4 mate wins quicker!
I liked 6...Bg3+ because it kept White's g pawn from moving, but it turns out that 6...Bd6+ also works: 7. Kf3 Qh5+ 8. g4 Qh3 mate!
