Giving up on Tactics

BobbyTalparov
MickinMD wrote:

Another thought: Look at my rating at chesstempo over the last 17 months. See how there are rough plateaus and valleys?  Every time I hit a valley I wondered if the site was making the problems harder or if I wasn't concentrating as well, etc.  Don't let times like that frustrate you - keep trying.  I think the valleys are mostly bad luck in getting a higher percentage of problems involving tactics I'm weaker at (I wish chess com would provide a list giving the %age of problems of each type that were correct instead of just the %age of max. points achieved - which are greatly affected by time, not correctness).  I've spent time exploring tactics I do less well on and, eventually, I hit a new high:

 

The following is fairly accurate on both sites (note:  the rating is for the tactic, not your rating!):

 

Tactics rated <1000 are all very simple tactics:  pins, skewers, mate in 1 or 2, double attacks, forks, etc.

Tactics rated 1000-1500 start to involve 1 move prior to the elementary tactics (i.e. "you make this move, which forces X reply, leading to Y tactic ..."

Tactics rated 1500+ start to chain tactical themes together and make use of more advanced tactical motifs.

 

So, if you are having problems with puzzles below 1000, you need to focus on your basic tactical principles.  If you are having problems with tactics in the 1000-1500 range, you need to work on visualizing what tactics are threatened on the board and how you might make use of them (i.e.  "If I could force his king to this square, I could pin his queen to his king and win the queen ... so, what move could help accomplish that?")  After that, it is working on calculation skills and identifying the most forcing combination in the correct move order.

 

The issue I have with the chess.com tactics trainer (since the change about a year ago) is the range you are with is too large.  When I break 2000, you start getting tactics in the 1600-2400 range.  The difference between a 1600 level tactic and a 2400 level tactic is massive!  And when you are bouncing back and forth between them, I've routinely missed low level tactics (even ones that I've done before) because I just had 3-4 2300-2400 level tactics in a row and then get a 1600 and over-think it.  The range really should be +/- 150, not +/- 400!

ghost_of_pushwood
cottonsock wrote:
You’re all wrong ! .........just saying.

grin.png

drmrboss

Just to give you an idea how hardworking  others are.

This guy did 20k tactics. 🤔

null

greghunt

I think the core problem is the lack of directive feedback from the tactics puzzles. I want to know "what was it that I was just supposed to recognise?".  There is an assumption in some of the comments above that I should be seeing tactical patterns a move or two in advance, but I don't know the patterns well enough to have the catalogue in my head that enables that.   Sometimes it appears that the objective is to find the least-worst response in a situation or the "right" response leaves me wondering why on earth you bother doing that and understanding what the scenario was supposed to illustrate would help a lot.  Some of the time the analysis tells me why the right answer is clearly better (but not what it was), but some of the time its just that the analysis engine scores the resulting board position highly without any particular resolution and when it does that I'm still in the dark.  

Hamur
greghunt wrote:

I think the core problem is the lack of directive feedback from the tactics puzzles. I want to know "what was it that I was just supposed to recognise?".  There is an assumption in some of the comments above that I should be seeing tactical patterns a move or two in advance, but I don't know the patterns well enough to have the catalogue in my head that enables that.   Sometimes it appears that the objective is to find the least-worst response in a situation or the "right" response leaves me wondering why on earth you bother doing that and understanding what the scenario was supposed to illustrate would help a lot.  Some of the time the analysis tells me why the right answer is clearly better (but not what it was), but some of the time its just that the analysis engine scores the resulting board position highly without any particular resolution and when it does that I'm still in the dark.  

I see where you're coming from on that.  Sometimes I don't know until after I get it wrong and I look at the tags for it.  Sometimes that's not too much help.  Had one puzzle that had the tags hanging piece and mate in 3.  Hanging piece was the correct one, but I wonder if mate in 3 was possible on that tactic?

 

Maybe for the lower level ones, they could be titled to help us just getting started, then ween us off of the later.

IMBacon

Take a look at the last tactic you did.

https://www.chess.com/tactics/89974/practice?autostart=1

You spent 4 seconds on it, and got it wrong.  Now go to it again, and use this to find the tactic.

First:  Ask yourself: "How did my opponents last move, change the position?"

Then look for Forcing Moves:

Checks.

Captures.

Threats,

greghunt

 I think a blunder is not a good example of the problem I'm referring to.  Sometimes (often enough) I simply screw up and those cases are uninteresting for this discussion.  Sometimes I can see how things need to turn out and simply don't calculate well enough (moves out of order/insufficiently forcing) and practice will eventually help with that.  Sometimes however, like exchanging a queen for two rooks a few days ago, I wonder why on earth its a good idea.  The tactics exercises just present those as something to accept.  Now you could find that one and explain it to me, but that doesn't deal with the underlying problem - there will be another like that soon enough.

drmrboss
greghunt wrote:

 I think a blunder is not a good example of the problem I'm referring to.  Sometimes (often enough) I simply screw up and those cases are uninteresting for this discussion.  Sometimes I can see how things need to turn out and simply don't calculate well enough (moves out of order/insufficiently forcing) and practice will eventually help with that.  Sometimes however, like exchanging a queen for two rooks a few days ago, I wonder why on earth its a good idea.  The tactics exercises just present those as something to accept.  Now you could find that one and explain it to me, but that doesn't deal with the underlying problem - there will be another like that soon enough.

I think you are talking about this game. The puzzle must start after black move Ke7.

 

Then your moves should be. 1. Re1+ Re6 (or 1....Qe6 2. Rxe6+ 3. Rxe6, or 1. ..Kf6 2. Re6+  Rxe6 3. Qxe6+ checkmate ) 2. Rxe6+ Qxe6 3. Qxe6+ Kf8(3... Ke7  4. Qxe7+mate ) 4. Nh7+mate   should be the solution. If you do alternative moves, you wont get mate in 4 anymore.null

nthompson

I question whether the tactics server is "honest"  I have runs of great performance where I am in the 1300's for a week and runs where I fall back into the 1100's.  The scoring just seems to get tougher and the puzzles more ... intricate.  It seems unlikely that my performance is varying that much.  I don't have bursts of genius and moronic plunges in any other domain of my life.  I wonder if perhaps we might start a discussion under the heading, Is the Tactics Server Honest?, and put our heads together to see what is going on.   By the way, MICKINMD, where did you get that splendid performance summary you posted?

 

N

 

 

FloaterGame

Guys, I'd like to make a comment....I've been working on a few tactics consistently daily, and my rating has shot up over 700 rating points... My rating (USCF) increased over 400+ points to around 1200-1300-ish, and my rating is still rising. I liked Chess Tempo initially, and still think it is a better option, but the fact that I can analyze my mistake using the computer helps me understand the tactical pattern, and I get better. (FYI I am 13)

FloaterGame

So yeah, it doesn't have a mistake....I'd recommend Jeremy Silman or James Eade's Chess for Dummies @Hamur. Turn to the tactics part, and it'll really help your tactical thinking process out, and it will help you become a better chess player, too. (My rating has shot up from around a 1100 to an 1845 in tactics in just a few months by reading these books/improving my chess)

472rating

I think the Tactics is okay, but the problem is sometimes there are two pieces you can use to accomplish the same end, but Tactics only ever has one solution, so then you need to try and guess which one will be correct.

472rating
[COMMENT DELETED]
BobbyTalparov
nthompson wrote:

I question whether the tactics server is "honest"  I have runs of great performance where I am in the 1300's for a week and runs where I fall back into the 1100's.  The scoring just seems to get tougher and the puzzles more ... intricate.  It seems unlikely that my performance is varying that much.  I don't have bursts of genius and moronic plunges in any other domain of my life.  I wonder if perhaps we might start a discussion under the heading, Is the Tactics Server Honest?, and put our heads together to see what is going on.   By the way, MICKINMD, where did you get that splendid performance summary you posted?

 

N

 

 

What you are seeing is the result of the range they are currently using.

 

Prior to ~1 year ago, the range of tactics you could get was roughly +/- 150 of your current rating.  So, if you were 1300, you could get some 1150 and some 1450 (and everything in between).  Now, it uses +/- 400.  So, at 1300, you could get some 900 and some 1700.  If you are shooting up to 1300 and then dropping several puzzles, it is likely due to this problem, when you end up getting several higher level problems in a row and you are not ready for them (or you miss a couple lower level problems because you over-think them after having a few higher level problems in a row).

greghunt
drmrboss wrote:
greghunt wrote:

 ... snip..

Sometimes however, like exchanging a queen for two rooks a few days ago, I wonder why on earth its a good idea.  The tactics exercises just present those as something to accept.  Now you could find that one and explain it to me, but that doesn't deal with the underlying problem - there will be another like that soon enough.

I think you are talking about this game. The puzzle must start after black move Ke7.

 

...snip...

The one "a few days ago" is not the third last tactics exercise, which I did on the day of that comment, and I'm not going digging to find it.  It involved the exchange of my queen for two rooks, leaving black with a queen but no rooks and me with rooks but no queen.  Thats why i said "you could find that one and explain it to me, but that doesn't deal with the underlying problem" , which was a statement intended to discourage what you did.  I think I got the one I was talking about right or mostly right,  based on thinking "this algorithm likes exchanging a queen for other pieces" but didn't understand why it would leave me in a better position and the analysis didn't help me understand afterwards.  There was no mate in some small number of moves after that, I couldn't see more material advantage accruing, the analysis rated the position positively but i couldn't SEE the decisive advantage.  Thinking "the computer is likely to want me to do this" is a terrible reason for doing anything, thats teaching me about an algorithm, not about the game.  

IAMBBW
Hamur wrote:

It is just getting more frustrating to see the completely ridiculous scenarios they put up. 

 

I have seen a mate in 2 be wrong because they thought I should grab a hanging piece. 

I've had my king in check on b2, I can move my king to either a2 or b1, and then my bishop on a1 is taken and I take their the bishop with my king.  I choose a2, and it wanted b1, neither of which made a difference to the outcome, king still ends up taking the bishop on a1, and there are no other pieces in play,  Also, no one in their right mind would have gone for that particular bishop swap!!

 I'm just finding too many ridiculous scenarios, making it useless as a learning tool.  I can't focus on the best chess move, I instead have to try and guess what the scenario is, which at this point, is only about 50% chance of it being the actual best chess move.

Good job

FatBertha

" It involved the exchange of my queen for two rooks, leaving black with a queen but no rooks and me with rooks but no queen."

 

I looked at the problem (0149221). You shouldn't only focus on the end result of the puzzle. You should also see what are the problems in the starting position. The puzzle solution was the only way to avoid losing a skewered rook. So you didn't gain any advantage, but you avoided immediate loss.

 

Generally speaking I haven't seen any incorrect puzzles here. If two moves appear equal then since any position can have only one right move (the puzzle stops where there are multiple ways to continue) this is a huge hint: either there is a particular reason why only one of these moves is strong, or neither of these moves is any good and you need to look for something else.

greghunt

Bertha,

Thanks for the explanation, its the kind of background that I would have liked to have access to to understand that problem after I've done it and I'm looking at it going "hmmm, I'm not sure I like the resulting position".  I don't doubt that the technical quality of the problems is good, they would be being checked mechanically, by the site staff and by the users, its my ability to understand the material thats the issue and some fairly brief clues for the post mortem would be useful.  

IAMBBW
greghunt wrote:

Bertha,

Thanks for the explanation, its the kind of background that I would have liked to have access to to understand that problem after I've done it and I'm looking at it going "hmmm, I'm not sure I like the resulting position".  I don't doubt that the technical quality of the problems is good, they would be being checked mechanically, by the site staff and by the users, its my ability to understand the material thats the issue and some fairly brief clues for the post mortem would be useful.  

Nice hair

evilstef

@BobbyTalparov , thats why i m now doing ,TT unrated and customized ,you can choose a tactic theme and the ratingrange ,also you can get hints in the unrated mode