1. Why are the Chess Mentor ratings so way off reality? The ratings seem to have absolutely no meaning.
2. Why would a simple one move forking problem in the Intro to Tactics section be at a 1700 rating? I see 1300 rated players pulling forks all the time.
3. Do you find that you work better and learn more with Sequential or Adaptive?
1. We do try our best to keep them close to reality. In reality the Chess Mentor ratings should not be compared to your real ratings and should only be used as an independent source.
2. Yes, that does seem odd. If you see a lesson which needs a rating adjustment please let us know! We are always open to suggestions and improvements for the Chess Mentor program and we will be happy to consider making a rating adjustment on some of the Chess Mentor lessons if it is necessary.
3. I would recommend the Adaptive mode for everyone as it will be tuned to your specific needs. Below I have included a detailed description of each training mode:
Adaptive Mode
The Adaptive Mode uses smart logic to determine what is the next best lesson to show you. Based on your rating and past lesson scores, Chess Mentor will choose lessons within your rating range that you have not mastered. As you do more and more lessons Chess Mentor will learn more about your strengths and weaknesses and become better and serving you the best lessons.
Sequential Mode
Though the Adaptive Mode will give you the right challenges based on performance, you may want to just move sequentially through all of the Chess Mentor lessons and courses. If this is the case, you can choose Sequential Mode where you want to go through a specific sequence of lessons. The Settings page lets you switch your default setting to adaptive or solving mode.
1. Why are the Chess Mentor ratings so way off reality? The ratings seem to have absolutely no meaning.
I love good chess mentor lessons, love them. I try not to think about the ratings, they are worse than meaningless... they are sometimes meaningful, sometimes not, making them just a big distraction.
2. Why would a simple one move forking problem in the Intro to Tactics section be at a 1700 rating? I see 1300 rated players pulling forks all the time.
See the answer to your first question. Stop thinking about the ratings and concentrate on the lessons is my advice.
3. Do you find that you work better and learn more with Sequential or Adaptive?
Adaptive is dependent on ratings etc. So I haven't even bothered with it. I don't see how it could work. I very much like big, thorough courses (I haven't done one in a while, come to think of it, been working thru a book) -- pawn endings from beginner to master (something like that) is my favorite CM course, and one of the best pieces of chess instruction I've ever encountered on any subject in any medium. If you need work on your pawn endings I highly recommend it.
Actually as JG27Pyth mentioned...one of the problems with Adaptive is that the ratings are so erroneous. Both for the problems and for the user. Because my rating is falsely set at 1995, I keep getting outrageously difficult problems.
That said, when I use Sequential...and I had the ratings set at 800 to 1900 level problems...I find problems that are not accurately rated. I bounce from a problem that I think a first week chess player should have...to deep 5 move combinations that lead to a draw that Alekhine missed in his this or that WC game.
I do find the Sequential is much much easier to deal with.
I'm going to change the gap to 1200 - 1700 and see if I can get some consistency.
If you keep getting matched with difficult problems because you feel you're overrated doesn't that mean that they're likley appropriately rated? Also, if you can't solve these problems wouldn't that eventually correct your rating to the point where you're getting matched with more appropriately difficult tactics?
If you keep getting matched with difficult problems because you feel you're overrated doesn't that mean that they're likley appropriately rated? Also, if you can't solve these problems wouldn't that eventually correct your rating to the point where you're getting matched with more appropriately difficult tactics?
You'd think so, right? That part would make sense...except that I'm still earning more points rather than the system adjusting them to my level...if I make 5 wrong moves and I try again...eventually I'm going to get it right...I should not be earning more points in such a situation?? And when I attempted to adjust the setting...I got problems I was sure had to be 1000 rating...which turned out to be 1700...something is wrong with the rating on the problems, I'm sure. It's not based on a once or twice usage.
Unless I'm nuts and it's working fine for everyone else? Some people must be rated 3000 on this thing?
Right, we are open to suggestions for modifying specific Chess Mentor lesson rating adjustments if a Chess Mentor lesson has an outrageously low or outrageously high rating when it should not.
Please feel free to post the Course and Lesson name along with the lesson code and the rating adjustment you feel we should make for it and we will be happy to consider to start to improve the rating accuracy throughout Chess Mentor. Any help is always appreciated!
It's very hard to set up an accurate rating for a problem. For instance, would you believe that a hanging queen problem is rated >2500 in Tactics Trainer? Here its initial rating was 1000, but many people did not solve it and thus it reached such a huge rating.
That is pretty funny. It's probably a product of the solvers' knowledge that it is a puzzle and as such there's a tactic to be found and that it can't possibly be that simple. In a game situation, where you don't know whether there's something to look for or not, the capture of the Queen would no doubt be the obvious top choice.
If you feel like you're getting some benefit from the program, why worry about the rating? Some of the harder problems may seem too difficult, but I'll bet that if you keep looking at them you'll begin to solve some of them.
If you feel like you're not benefiting fro the tactics trainer on this site, there are others. chesstempo.com is one.
It's very hard to set up an accurate rating for a problem. For instance, would you believe that a hanging queen problem is rated >2500 in Tactics Trainer? Here its initial rating was 1000, but many people did not solve it and thus it reached such a huge rating.
I'm having a similar probelm in that I feel like I'm rated a little too high so the suggested lessons seem a bit beyond me, although I slove them by trial and error and still get some points which makes my rating even higher! In the end I've just learned to ignore the ratings, but it does make finding the right lesson a bit more challenging.
I've been doing some fo the strategy lessons and they invariably end with something like "and white is totally winning." I can see all the reansons why white may be winning, but I'm looking for the lesson that goes a little further into the game and shows me how to turn this "winning postion" into an actual win. Is there a "how to win with a winning position" lesson LOL?
Burn...I'm up to 2216 now, so I completely understand. What I have found that works for me is to look up the courses...from easiest to hardest and then to focus on one course at a time...I'm enjoying it much more this way. I'm currently slogging through Endings, Openings, a Taste of the Middle and it's really good even though I hate working on endgame stuff.
forget the rating and points though, because they keep climbing even when you solve a 900 rated problem at only 60 percent and you're already at 2200.
"The ambitious 3.f4 adds to the already considerable gains as far as central space is concerned and it also starts to claim territory on the kingside. Now Black has firm control over e5."
I think that last sentence is supposed to be "Now White has firm control..."
This is a three pronged post about Chess Mentor.
My Chess Mentor
1. Why are the Chess Mentor ratings so way off reality? The ratings seem to have absolutely no meaning.
2. Why would a simple one move forking problem in the Intro to Tactics section be at a 1700 rating? I see 1300 rated players pulling forks all the time.
3. Do you find that you work better and learn more with Sequential or Adaptive?
1. We do try our best to keep them close to reality. In reality the Chess Mentor ratings should not be compared to your real ratings and should only be used as an independent source.
2. Yes, that does seem odd. If you see a lesson which needs a rating adjustment please let us know! We are always open to suggestions and improvements for the Chess Mentor program and we will be happy to consider making a rating adjustment on some of the Chess Mentor lessons if it is necessary.
3. I would recommend the Adaptive mode for everyone as it will be tuned to your specific needs. Below I have included a detailed description of each training mode:
Adaptive Mode
The Adaptive Mode uses smart logic to determine what is the next best lesson to show you. Based on your rating and past lesson scores, Chess Mentor will choose lessons within your rating range that you have not mastered. As you do more and more lessons Chess Mentor will learn more about your strengths and weaknesses and become better and serving you the best lessons.
Sequential Mode
Though the Adaptive Mode will give you the right challenges based on performance, you may want to just move sequentially through all of the Chess Mentor lessons and courses. If this is the case, you can choose Sequential Mode where you want to go through a specific sequence of lessons. The Settings page lets you switch your default setting to adaptive or solving mode.
1. Why are the Chess Mentor ratings so way off reality? The ratings seem to have absolutely no meaning.
I love good chess mentor lessons, love them. I try not to think about the ratings, they are worse than meaningless... they are sometimes meaningful, sometimes not, making them just a big distraction.
2. Why would a simple one move forking problem in the Intro to Tactics section be at a 1700 rating? I see 1300 rated players pulling forks all the time.
See the answer to your first question. Stop thinking about the ratings and concentrate on the lessons is my advice.
3. Do you find that you work better and learn more with Sequential or Adaptive?
Adaptive is dependent on ratings etc. So I haven't even bothered with it. I don't see how it could work. I very much like big, thorough courses (I haven't done one in a while, come to think of it, been working thru a book) -- pawn endings from beginner to master (something like that) is my favorite CM course, and one of the best pieces of chess instruction I've ever encountered on any subject in any medium. If you need work on your pawn endings I highly recommend it.
Patzer, thanks for the thorough response.
Actually as JG27Pyth mentioned...one of the problems with Adaptive is that the ratings are so erroneous. Both for the problems and for the user. Because my rating is falsely set at 1995, I keep getting outrageously difficult problems.
That said, when I use Sequential...and I had the ratings set at 800 to 1900 level problems...I find problems that are not accurately rated. I bounce from a problem that I think a first week chess player should have...to deep 5 move combinations that lead to a draw that Alekhine missed in his this or that WC game.
I do find the Sequential is much much easier to deal with.
I'm going to change the gap to 1200 - 1700 and see if I can get some consistency.
If you keep getting matched with difficult problems because you feel you're overrated doesn't that mean that they're likley appropriately rated? Also, if you can't solve these problems wouldn't that eventually correct your rating to the point where you're getting matched with more appropriately difficult tactics?
If you keep getting matched with difficult problems because you feel you're overrated doesn't that mean that they're likley appropriately rated? Also, if you can't solve these problems wouldn't that eventually correct your rating to the point where you're getting matched with more appropriately difficult tactics?
You'd think so, right? That part would make sense...except that I'm still earning more points rather than the system adjusting them to my level...if I make 5 wrong moves and I try again...eventually I'm going to get it right...I should not be earning more points in such a situation?? And when I attempted to adjust the setting...I got problems I was sure had to be 1000 rating...which turned out to be 1700...something is wrong with the rating on the problems, I'm sure. It's not based on a once or twice usage.
Unless I'm nuts and it's working fine for everyone else? Some people must be rated 3000 on this thing?
Right, we are open to suggestions for modifying specific Chess Mentor lesson rating adjustments if a Chess Mentor lesson has an outrageously low or outrageously high rating when it should not.
Please feel free to post the Course and Lesson name along with the lesson code and the rating adjustment you feel we should make for it and we will be happy to consider to start to improve the rating accuracy throughout Chess Mentor. Any help is always appreciated!
Most Recently...
Omnipotent Queen
Introduction to Tactics
Author(s): IM Jeremy Silman, FM Thomas Wolski, NM
and another lesson called Another Aggressive King.
Both rated 1700. The Aggressive King lesson is probably an 800 rating lesson.
It's very hard to set up an accurate rating for a problem. For instance, would you believe that a hanging queen problem is rated >2500 in Tactics Trainer? Here its initial rating was 1000, but many people did not solve it and thus it reached such a huge rating.
http://www.chess.com/tactics/server.html?id=36709
Costelus...I wouldn't have believed that. I had to click and see it for myself. That's hilarious.
That is pretty funny. It's probably a product of the solvers' knowledge that it is a puzzle and as such there's a tactic to be found and that it can't possibly be that simple. In a game situation, where you don't know whether there's something to look for or not, the capture of the Queen would no doubt be the obvious top choice.
My chess mentor rating is almost 2300. I wish that was my real rating
If you feel like you're getting some benefit from the program, why worry about the rating? Some of the harder problems may seem too difficult, but I'll bet that if you keep looking at them you'll begin to solve some of them.
If you feel like you're not benefiting fro the tactics trainer on this site, there are others. chesstempo.com is one.
Most Recently...
Omnipotent QueenIntroduction to Tactics
Author(s): IM Jeremy Silman, FM Thomas Wolski, NM
and another lesson called Another Aggressive King.
Both rated 1700. The Aggressive King lesson is probably an 800 rating lesson.
Thanks, I have lowered both ratings for these two lessons as both of the lessons were considerably over rated.
It's very hard to set up an accurate rating for a problem. For instance, would you believe that a hanging queen problem is rated >2500 in Tactics Trainer? Here its initial rating was 1000, but many people did not solve it and thus it reached such a huge rating.
http://www.chess.com/tactics/server.html?id=36709
Thanks costelus. I have now lowered the rating for this Tactics Trainer problem.
See...
My Chess Mentor
(see full progress report)
Adaptive is impossible.
I'm having a similar probelm in that I feel like I'm rated a little too high so the suggested lessons seem a bit beyond me, although I slove them by trial and error and still get some points which makes my rating even higher! In the end I've just learned to ignore the ratings, but it does make finding the right lesson a bit more challenging.
I've been doing some fo the strategy lessons and they invariably end with something like "and white is totally winning." I can see all the reansons why white may be winning, but I'm looking for the lesson that goes a little further into the game and shows me how to turn this "winning postion" into an actual win. Is there a "how to win with a winning position" lesson LOL?
Burn...I'm up to 2216 now, so I completely understand. What I have found that works for me is to look up the courses...from easiest to hardest and then to focus on one course at a time...I'm enjoying it much more this way. I'm currently slogging through Endings, Openings, a Taste of the Middle and it's really good even though I hate working on endgame stuff.
forget the rating and points though, because they keep climbing even when you solve a 900 rated problem at only 60 percent and you're already at 2200.
Mine was around 2200 on my last account... I never really use it. Only a few of the courses seem to have any value.
Patzer...
I think there might be a small mistake...
Course: Endings, Openings, a Taste of the Middle
The Beginnings of Space
"The ambitious 3.f4 adds to the already considerable gains as far as central space is concerned and it also starts to claim territory on the kingside. Now Black has firm control over e5."
I think that last sentence is supposed to be "Now White has firm control..."