chess.emrald.net landing
wew. i was just warming into the first 250 problems of ct-art when i hit a wall. since i am not focused on doing all 7 circles in a flash, certainly not the first circle or two, and my rating there hit 2030 then fell like a stone, THEN cts resumed operation, i went back in full confidence. the thing is, in my trip east to see my aging dad, then my mom, when i did 1500 exercises in eight days, with an RD of about 13 or 14 near the end, about 87 to 88% success in that spate, and inches from 1500 again at 1491 or so, i had EVERY expectation of 'just' getting back on, warming up a bit, then restoring +1500 as i feel and felt that i belong...
but literally on the first exercise, after the outage, when i had an rd of 38 or so, i literally couldnt move the mouse. i forgot how, just like many do when starting--you dont grab it like ICC or yahoo or playchess, you dont slide it like ct-art, but you touch the piece moved to the square intended...
ok. so one bad problem? but one after another. then my 11,500 shown as 11,340 and my rating as it was at that point, like 1474 the day before the outage. mind you, not to complain about CTS, but my failure! and frustration at rustyness... so there i was, ready to hit 1500 in a flash, but lower, then fall all the way to 1438. yikes.
not so bad. but since im trying to get my percentage success to 85% from 83.2, i fell to 83.05, so had to fight and fight, got back to the outage peak of 83.2, and now tonight 83.3449 (ive reversed engineered the break points [i.e. notice at what integers of solved problems i go from, say, 83.1 to 83.2 etc, then converge the data like a simple calculus] to track my success one single problem at a time, noting on a pad each success/fail).
but to go from 83.05 to 83.35 and soon to be printed as '83.4' (=>83.350~ 83.351 etc), with 13,000 problems, you cannot just 'whipp it out'. you to crawl one problem at a time--two to three hours a day. my settings stop at each problem as i need those few seconds to center myself...
which gets me to now. temposhlucker mentioned peak and floor ratings. right on. now that im 1500, unlike last time when i let the rd raise to do fewer or lesser problems per day and got to 1524 quite rapidly, i plan to hold the 15 level, but use extra activity NOT to push it up fast or try to, but rather see at 85 to 90% how much i can hold that. some days i can do 90%. then i crash and burn, then must 'get it back'. it really averages now 85.5% to 86.5% pretty steadily. there is a certain exchange between rating, speed, and success. so now i will see how far i can push my % success, while holding the rating. rating falls, i must speed up. rating raises, i can slow down and extract greater % success.
at 12,999 at 83.3449, ive done 10,834 correct, 2,165 failed. IF i can do the next 7,001 at 86%, that would be 6,021 success. thus id be at 16,855 or 84.28% for 20,000. i aim for more like 85%+ for 30 or 40,000 but my own jubilee is a LONG ways off--but not that far off!
deminishing returns... as resistence.
lastly, i really need to get back to cta, but i did NOT like cta when im learning to calculate better but cannot do it fast! so my plan is to start each day, 20 cta3.0 exercises, THEN 40 cts, then 20 cta at pm, then maybe 60 cts at pm. i dont really need or want 150 cts per day--takes a lot of time, and, again, to use tempoS term, a LOT of stress--and ensuent exhaustion...
thus, with the dualness of this, run both horses: the horse of accuracy; the horse of rapidity and automating pattern recognition. like tempS, ive seen the same cts exercise again and again, only to fail, then others a slam dunk EVERY time--like the one with P a7, Ke7 is it, and R at a8, goes over to h7 to check Blk K. how many times must i do this?????
so deminishing returns on new patterns, yet use enough time to NOT this time loose the skill while encorporating it into rapid (3 min/8 sec increm), then standard (20 min--or so/12 sec increm).
i will finish reading Averbakh's: Endings, Essential Knowledge, and am much enjoying Alburt's Chess Training Pocket Manual at bedtime or at the beach (laying down in swim shorts with earplugs {children!}), and am at problem 185 or so, an excellent counter point to CTS and CTA3.0.
thank you to this extended community--CTS and blogger. you all are the best.
dk
but literally on the first exercise, after the outage, when i had an rd of 38 or so, i literally couldnt move the mouse. i forgot how, just like many do when starting--you dont grab it like ICC or yahoo or playchess, you dont slide it like ct-art, but you touch the piece moved to the square intended...
ok. so one bad problem? but one after another. then my 11,500 shown as 11,340 and my rating as it was at that point, like 1474 the day before the outage. mind you, not to complain about CTS, but my failure! and frustration at rustyness... so there i was, ready to hit 1500 in a flash, but lower, then fall all the way to 1438. yikes.
not so bad. but since im trying to get my percentage success to 85% from 83.2, i fell to 83.05, so had to fight and fight, got back to the outage peak of 83.2, and now tonight 83.3449 (ive reversed engineered the break points [i.e. notice at what integers of solved problems i go from, say, 83.1 to 83.2 etc, then converge the data like a simple calculus] to track my success one single problem at a time, noting on a pad each success/fail).
but to go from 83.05 to 83.35 and soon to be printed as '83.4' (=>83.350~ 83.351 etc), with 13,000 problems, you cannot just 'whipp it out'. you to crawl one problem at a time--two to three hours a day. my settings stop at each problem as i need those few seconds to center myself...
which gets me to now. temposhlucker mentioned peak and floor ratings. right on. now that im 1500, unlike last time when i let the rd raise to do fewer or lesser problems per day and got to 1524 quite rapidly, i plan to hold the 15 level, but use extra activity NOT to push it up fast or try to, but rather see at 85 to 90% how much i can hold that. some days i can do 90%. then i crash and burn, then must 'get it back'. it really averages now 85.5% to 86.5% pretty steadily. there is a certain exchange between rating, speed, and success. so now i will see how far i can push my % success, while holding the rating. rating falls, i must speed up. rating raises, i can slow down and extract greater % success.
at 12,999 at 83.3449, ive done 10,834 correct, 2,165 failed. IF i can do the next 7,001 at 86%, that would be 6,021 success. thus id be at 16,855 or 84.28% for 20,000. i aim for more like 85%+ for 30 or 40,000 but my own jubilee is a LONG ways off--but not that far off!
deminishing returns... as resistence.
lastly, i really need to get back to cta, but i did NOT like cta when im learning to calculate better but cannot do it fast! so my plan is to start each day, 20 cta3.0 exercises, THEN 40 cts, then 20 cta at pm, then maybe 60 cts at pm. i dont really need or want 150 cts per day--takes a lot of time, and, again, to use tempoS term, a LOT of stress--and ensuent exhaustion...
thus, with the dualness of this, run both horses: the horse of accuracy; the horse of rapidity and automating pattern recognition. like tempS, ive seen the same cts exercise again and again, only to fail, then others a slam dunk EVERY time--like the one with P a7, Ke7 is it, and R at a8, goes over to h7 to check Blk K. how many times must i do this?????
so deminishing returns on new patterns, yet use enough time to NOT this time loose the skill while encorporating it into rapid (3 min/8 sec increm), then standard (20 min--or so/12 sec increm).
i will finish reading Averbakh's: Endings, Essential Knowledge, and am much enjoying Alburt's Chess Training Pocket Manual at bedtime or at the beach (laying down in swim shorts with earplugs {children!}), and am at problem 185 or so, an excellent counter point to CTS and CTA3.0.
thank you to this extended community--CTS and blogger. you all are the best.
dk
posted by transformation at 11:46 PM
1 Comments:
Have a look at an old post of mine http://temposchlucker.blogspot.com/2005/12/halfway-math.html
I expressed successrate as #problems correct. Only I thought by then that the random picks of a problem from the problemwindow was a linear function, which it isn't.
Post a Comment
<< Home