Home » Forum Home » Analysis

Topic: Pente - Understand vs Memorize
Replies: 122   Views: 238,637   Pages: 9   Last Post: Nov 17, 2010, 10:26 AM by: zoeyk

Search Forum

Back to Topic List Topics: [ Previous | Next ]
Replies: 122   Views: 238,637   Pages: 9   [ Previous | 1 2 3 4 5 6 7 8 9 | Next ]
alisontate

Posts: 157
Registered: Nov 27, 2008
Age: 30
Re: Pente - Understand vs Memorize
Posted: Oct 6, 2010, 3:09 PM

Jason,

one other side to this of course is your personality. If you are disposed toward memorizing complex things then no doubt you would not have posed this question. So I guess, the subtext here is something like "I would prefer it if there was some way to avoid all this memorization but still become a master" which I am sure many people - myself included - would also ideally want.

You could look at memorization of well proven master lines as simply absorbing the understanding of the masters as expressed in their moves. So the key is that the understanding you seek is in those moves. The move is the expression (phenotype) of the ideas, concepts, analysis, understanding (genotype) of the masters.

IMO there are two levels of memorization. One is to copy the moves without the understanding (shallow copy) or copy the understanding as well (deep copy).

So the answer to your question might be that you can, as zoey implied, understand AND memorize. You can do this by memorizing the understanding.

Alison


xtraclassy

Posts: 66
Registered: Dec 13, 2008
From: New York City
Age: 58
Re: Pente - Understand vs Memorize
Posted: Oct 6, 2010, 7:31 PM

Pretty progressive post.

Don't quite get the part about memorizing "understanding". Don't know that understanding is something that can be memorized.

My simple anology would be that memorizing means you have a tool; understanding means you know how to use it; wisdom would mean that you know when to use it.
There are always new situations that routinely arise, so I'm not quite sure how memorizing can help in those situations.


For Jason, there are problems and there are solutions. Some solutions may work but they may be totally unacceptable options to some people. Memorizing a few thousand lines or paths is an unacceptable solution to some people regardless of whether it works or not. I, for one, do not believe that doing that will ever get you to Nosovs or Richard III's level.

As an education professional, I also know that not all approaches work with people with different learning styles.

It will improve your game (memorizing) but no true champion's goal is anything less than number one. (Would hope that you want to eventually become the best player in the world; not just one of the top players)

I believe more games are lost by mistakes in the middle and later stages of games than are lost in the opening moves. If a player is capable of making a mistake in the first 5 - 10 moves, I see no logical reason to believe he/she is not equally capable of making a mistake in moves 11-20 and beyond. Granted, you may not get the chance since an early mistake can mean you don't get to move 11-20.

On any move there are many questions you can ask youself. Most players ask themselves many questions which make this game seem a lot more complicated than it should be. I am not getting into that in this post, but the first and most important question should be: "Is there a play that I have to make or lose?"

I have seen players waste a lot of clock when they really had only one choice: "play here or lose."

Any other question only becomes pertinent, if and only if, the answer to that question is "no".

The moral of that story: "the order in which you ask yourself important questions matters."

NB. That is more often than not a set of three or four moves one of which must be made to prevent or postpone losing. If a player limits himself/herself to those pertinent questions, he/she will, at the very least, do well at speed pente.

zoeyk

Posts: 2,070
Registered: Mar 4, 2007
From: San Francisco
Age: 43
Home page
Re: Pente - Understand vs Memorize
Posted: Oct 6, 2010, 11:34 PM

z;
if there is a positional line that you come to understand its logic, and you like it's style, or it is the most "optimal path" you have yet to date found, then memorizing that positional line is ok. this is memorizing that which was understood and deemed as best choice for you at that moment. understood, thus certified as acceptable to copy. so yes, "memorizing the understood".


xtraclassy said;
My simple analogy would be that memorizing means you have a tool; understanding means you know how to use it; wisdom would mean that you know when to use it.


Z; hmm, there is truth to what you say classy, but its confusing the subject for this topic a little perhaps.
we are not initially talking about shape memorization.
although this topic can lead into that.
we are talking about a line beginning from move 1 at K10.
and it continues until P2 deviates from the positional line. if this line (while un-deviated) is accepted as a solid known winning beginning, then wisdom of when to use it is not aplicable. but if you talk shape knowledge from this line to apply to other games with slight similarity then your wisdom this begins to apply.
for example, the wedge line up until P2 deviates, or until the wedge lands at the 2 caps vs 2 caps point. not much wisdom is required during this transaction.
and yes, there are memorized lines that are as logical as the wedge line is, where no thought is required until the line is either strayed from by P2 early, or until it cycles to a certain point, where P2 has many choices, and P1 now awaits for the question to answer.


xtraclassy said;
Memorizing a few thousand lines or paths is an unacceptable solution to some people regardless of whether it works or not. I, for one, do not believe that doing that will ever get you to Nosovs or Richard III's level.


z;
you think those 2 guys don't have tons of P1 starting lines memorized lol? they might make many of them on the fly, or even reinvent for various reasons, but they remember too.
i don't think any one is saying, if you memorize the data base you'll be a grand master... memorizing is just one piece to the puzzle.
memorizing does not take away from your skill level if you understand that which you are memorizing. it adds to what you have. it is not intended to be a replacement for something, it is a upgrade or add on.


xtraclassy said;
I believe more games are lost by mistakes in the middle and later stages of games than are lost in the opening moves. If a player is capable of making a mistake in the first 5 - 10 moves, I see no logical reason to believe he/she is not equally capable of making a mistake in moves 11-20 and beyond. Granted, you may not get the chance since an early mistake can mean you don't get to move 11-20.


z;
well,.. depends what level of players we are talking here, but, i'll say that for the majority, most mistakes are made early. and thats a fact that many top level players will back me up on.
a sloppy 2nd adds struggle, and bad 3rd is unforgiving, and so on. and yes, most players seem to have trouble making it out of the woods of the opening in a winning form. a mistake can happen any where tho sure, but we are talking about the majority here. when you do the first 11 moves as you had stated, and if these are perfect play, and you understood what you were doing so far, the board should generally be set up in the same intuitiveness to make it easier to seal the deal. the likely hood of a error at that stage is much less if your opening and mid is already set up with perfect play.

moves are like a pebble being dropped from on top of a snowed mountain. or like a pebble being dropped into a still pond. it starts small and simple, but every small action at the beginning will be amplified showing small flaws as huge ones as it resonates down the line...

a bad 3rd is so small it is seemingly an invisible ripple at first to some,.. but get down into the mid game and it has started a title wave thats about to crash down in your direction instead of your opponents direction.


don't get me wrong classy, you did make some good points too. and if i wasn't needing to head to work now ide address those too. ill have a couple days off soon, and adding more to this is something i look forward to. good thread so far.

and don't call me on my bad spelling cus you spelled anOlogy wrong

Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
zoeyk

Posts: 2,070
Registered: Mar 4, 2007
From: San Francisco
Age: 43
Home page
Re: Pente - Understand vs Memorize
Posted: Oct 6, 2010, 11:44 PM

but hey!, what about this idea, wouldnt at the bare minimum, wouldnt it be great to memorize what not to do?
wouldnt it be great to know what P1 3rds are auto sure loss moves? and there are many of these... when i have time ill post many examples to show what i mean..

or perhaps someone might think it unethical for me to share such information in a public arena like this. if so, let me know now before i post the moves. you have 24 hours to prevent me


Message was edited by: zoeyk at Oct 6, 2010 6:13 PM


Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
xtraclassy

Posts: 66
Registered: Dec 13, 2008
From: New York City
Age: 58
Re: Pente - Understand vs Memorize
Posted: Oct 7, 2010, 1:39 AM

Great points, Z.

Only objection I have is to the spelling note.

It is "analogy" not "anology" as you suggested (lol).

Forget that, anyway, I look forward to the knowledge you are known and appreciated for imparting on the right way to play the game.

..)

alisontate

Posts: 157
Registered: Nov 27, 2008
Age: 30
Re: Pente - Understand vs Memorize
Posted: Oct 7, 2010, 8:52 AM

I would just like to put my vote in for zoey not to publish the bad 3rds.

There are about 10 good reasons I can think of why this is not desireable.

zoeyk

Posts: 2,070
Registered: Mar 4, 2007
From: San Francisco
Age: 43
Home page
Re: Pente - Understand vs Memorize
Posted: Oct 7, 2010, 10:49 AM

classy, sorry bud, but u did spell it wrong. read your upper post, it clearly says "anology"... you always give me a hard time for not using spell check, and u clearly didn't use it that time. but im just messin with ya, so never mind.

alison, i hear ya. although one stated reason in the hand would had been nice, i guess 10 reasons in the bush will do. so i wont post it. but it would had made for some great examples...

z

Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
zoeyk

Posts: 2,070
Registered: Mar 4, 2007
From: San Francisco
Age: 43
Home page
Re: Pente - Understand vs Memorize
Posted: Oct 7, 2010, 1:03 PM

jason said;
I'd be very interested to hear what your thought process is when determining your next move.



Z;
hmm,
first this question seems to ask me to go off topic. so a new thread for this perhaps?

and, a good question, but vague?
my step father jay e hoff once said, "we have all the answers, we just need to figure out what the question is".

if a vague question, i might give a vague answer.
so ide recommend refining it into meaningful sub categories of specific questions. my answers will become more meaningful perhaps if you do this.

and keep this in mind, it would be good to keep P1 and P2 separate when devising questions. the thought process answers are often different with these 2 sides of the coin. so pick a side to focus on. i always recommend starting with P1. P2 is more complex, it is the next level of mastery. P1 is the first level.


P1 just plays the board to maintain its available perfection of advantage with skill (pshycology not required),

P2 plays both the board and the human with intent to guide P1 into err with both skill and illusions (pshycology is very useful).


this difference of "just board" vs "board and human" is important to understand.



when a player switches up his P1 second move (vs me) to some random move that creates struggle for him. i ask why change your normal stronger move to this? they reply, "i felt i needed to try and trick you". this answer clearly shows me that they have failed to realize my points above about board vs board and human.

when playing as P1, not only should you realize your advantage and own up to it. but you should always pretend you are playing nosovs. even if its a 900 rank player who couldnt win if thier life depended on it. every move as P1 you should play only the best move. or a move you know can force a win. every move think that nosovs will take their chair and start placing god like moves that will exploit any error you might have made. players will look at you in the data base to see your sleepy moves, they will target you.

you play like this to stay sharp. falling asleep at the wheel just becuase the opponent is easy to beat will only make you weaker over time to the point that your amount of errors as P1 will begin to increase.

Note; of course the exeption to this is if your teaching a new player, playing at half strength is ok then, but for sure cancel those games, unless your padding the data base for devious reasons

and exception #2, its ok to play at half strength if your nosovs, becuase his 50% strength still wins regardless lol

all just in my humble opinion...

Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
zoeyk

Posts: 2,070
Registered: Mar 4, 2007
From: San Francisco
Age: 43
Home page
Re: Pente - Understand vs Memorize
Posted: Oct 7, 2010, 2:22 PM

and alison, i liked what you wrote so much im going to paste it here one more time to point out my favorite 2 paragraphs. very nice.


"You could look at memorization of well proven master lines as simply absorbing the understanding of the masters as expressed in their moves. So the key is that the understanding you seek is in those moves. The move is the expression (phenotype) of the ideas, concepts, analysis, understanding (genotype) of the masters.

IMO there are two levels of memorization. One is to copy the moves without the understanding (shallow copy) or copy the understanding as well (deep copy)."





but i'll add and ask, what if they just copy the concepts found in the moves, but disregard the positional lines? what type is this? mid copy? concept copy? or.. am i just complicating things with out good reason here..

Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
zoeyk

Posts: 2,070
Registered: Mar 4, 2007
From: San Francisco
Age: 43
Home page
Re: Pente - Understand vs Memorize
Posted: Oct 7, 2010, 4:09 PM

and to just briefly address initiative (off topic slightly, but a definition that could come in handy in this discussion)


a copy and paste from nosovs about initiative;

Nosovs;
I think that initiative and wining position different things.
Initiative and tempo is very important and show us who will attacking. In Pente it very often means that this player build his winning combination and improve his position.

But sometimes if the player can not win he just create more addition stones to his opponent and improve his position.

20 times Meijin and 2 times World Champion in Renju Shigeru Nakamura style is
"gave up initiative and haughtily wait mistake of opponent"






"gave up initiative and haughtily wait mistake of opponent",

can mean;

"I gave my opponent initiative, and now I wait for my inevitable victory"


bottom line, initiative might be a forcing move, but does in no way state who is in the winning position, unless it is initiative of a VCT sequence. but we just call that a VCT. it isnt initiative by its self, it is initiative plus continuous plus victory. so initiative only means winning position when in a combination, but not by its self as a stand alone term.
winning position is just winning position.
initiative is just forcing move(s) of the moment(s),
2 different things...

every time i see someone loosly say "initiative" to literaly mean "winning position", as opposed to correctly implying that both are two different terms simultaniously taking place as a combined function, its like nails running down a chalk board.

Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
jasonb

Posts: 105
Registered: Jan 3, 2010
Home page
Re: Pente - Understand vs Memorize
Posted: Oct 7, 2010, 4:53 PM

Wow, this is turning into a very interesting thread! Thanks to everyone for their support.

Jason said, "I'd be very interested to hear what your thought process is when determining your next move."

Zoey said, "first this question seems to ask me to go off topic. so a new thread for this perhaps? . . . and, a good question, but vague?"

Response: I believe I'm staying on topic. Zoey, I'm looking for your version of the punch lists that up2ng and I made earlier in this thread regarding the questions we evaluate prior to making the next move. This doesn't need to be a complete list, just the general questions you ask yourself, things you look for, etc. prior to committing to your next move. Please focus on P1 first and come back to P2 later when you have time.

zoeyk

Posts: 2,070
Registered: Mar 4, 2007
From: San Francisco
Age: 43
Home page
Re: Pente - Understand vs Memorize
Posted: Oct 7, 2010, 5:09 PM

I'll need some time to ponder that big can of worms.

Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
watsu

Posts: 1,234
Registered: Dec 16, 2001
Re: Pente - Understand vs Memorize
Posted: Oct 7, 2010, 9:34 PM

As a slight aside from the thrust of the thread, I'd like to take a look at the title proposition: Understand vs Memorize.. Suppose we imagine for a moment perfect understanding without memory and perfect memory without understanding.

To be more concrete, let's suppose someone in the future develops a pente program which perfectly understands the best possible move in every circumstance in the game but which has no memory and has to reinvent each perfect move from scratch each time. Given sufficient time and computational power, this program will play perfectly. However, within any fairly short time span (let's say 24 hours, to give it a cutoff) there are certain positions which it will be unable to analyse with current computational power. So, its exploitable flaw will be that people can find such positions and force it into timing out.

Now, a different computer program comes out at the same time. Using games which have already been played (from any available databases) and adding its own played games into this ever expanding database it plays strictly by percentages for each move, playing the move with the highest win percentage in every case. This program has a few obvious exploitable flaws as well. As it inches towards a perfect memory of all possible games and optimal moves, it will play a lot of junk from the db - whether intentionally or unintentionally inserted into the db doesn't matter.

So, what one seems to face in timed games based on this limited analogy is a trade off. Without understanding, memorization is only as good as the quality and quantity of the material memorized. Without memorization, understanding has to reinvent the wheel every time for every move, which often takes far too long.

Humans with a certain amount of pente experience can currently easily eliminate moves from consideration which the computer will unnecessarily spend inordinate amounts of time analysing. The human may not always be correct in this elimination, of course, but will often arrive at the exact conclusion for the strongest move in a position that a computer will - only often they will do so many hours more quickly than the computer. Learning to see further ahead in the game and to anticipate and explore the opponents possible counter moves and your counters to their counters, etc. is something which mostly comes with practice. Some are better than others at this, but I think play improves this ability in almost everyone. And it also improves one's memory of the game as well, in most cases. After one has lost several games as P1 from a certain opening position (for some people as soon as they lose once, this process happens), you'll start to try to figure out where you are going wrong if you want to advance to the next level. Usually, the opening sequence is simply a sure loss to an expert player (such as the hammer, for example).

So anyway, my advice in a nutshell is- read (pente books, forum analysis), study (database and toss stones around on a board), play (as much as possible and with opponents who are ideally just enough better than you so that you can still see how they are beating you) and try to retain as much as you can so that you can at some point easily replay the first eight or more moves of a game you which just played in order to try a different possibility in the later stages of the game. Oh- almost forgot- study your opponent's P1 openings looking for instances where after 3-4 stones black has a stronger initial shape than P1 has managed to create. This type of opening sequence usually represents a flaw which you may be able to take advantage of.


Message was edited by: watsu at Oct 7, 2010 3:35 PM


Retired from TB Pente, but still playing live games & exploring variants like D, poof and boat
zoeyk

Posts: 2,070
Registered: Mar 4, 2007
From: San Francisco
Age: 43
Home page
Re: Pente - Understand vs Memorize
Posted: Oct 8, 2010, 2:26 AM

good points watsu.

memory only of all data base moves = loss of quality in the percentage.

understanding = loss of time to process from scratch every move.

finding the middle ground is the goal. the data base is a double edged sword as a whole. the trick is to filter out the junk and retain the modern used goods. and realize that these might evolve and become replaced should logical reasons arise warranting such. then the edge towards you becomes dull, and the edge towards the opponent becomes sharper. for studying your P1 responses, the most important is to design your opening responses. 2nds 3rds are very important, and a great starting goal in this process.

studying the data base can also be good to understand what is not good for P1. this helps you to build your P2 openings, especially P1's 2nds and 3rds to exploit early flaws to either guarantee a P2 win, or to build the most solid possible early attack to give you the highest odds of winning. but remember that this is playing the board. since we are talking about P2 now, playing the human is important too, so sometimes as P2 you will do less optimal moves because you realize in a unique situation that your odds of winning are higher with this particular human to do a certain play that is less optimal, because you understand that their mind processes in a certain way.

designing the best 2nd for P1 is much easier than designing the best 3rds. it is the 3rds we spend much time developing, then going back to the drawing board on when exploits arrise.

the 2nds of P1 have been known to be redesigned due to struggle issues. usually because of a bad 3rd, but we didn't know, so we just scrap the second. but sometimes the 2nd was just too much struggle to warrant putting a Band-Aid on. and sometimes we repair the 3rd by replacement and resume using the old 2nd.

the reason developing the 2nd happens much less is because there are only 12 of them to figure out.

but when it comes the P1's 3rds, this number reaches close to 500 moves. an average of 30 to 40 X 12.

so if you want to study, ide recommend figure out the 12 best 2nds as P1.

then pick one of those moves in the 12, and look at all of P2's possible seconds. now for each of those design a custom 3rd that best answers it. then test it out. if you don't like it then try another.

once you have settled into a consistent P1's 2nd and 3rd move repertoire, you will need to examine your own games as P1.

If you win, you will need to go back and study it to see if there was a way you could had lost. because players will put your P1 wins under a microscope to find a flaw.
if you find a flaw, fix it. to fix it work backwards from the end. until you determine the bad move. remove the bad move and replace with a good move. if no good move available then go back another move, until you are able to delete a move and replace it with a good one.
also, maybe you find at move 11 the ability to replace a bad move with a good move, but you realize that it will be messy to seal the deal, do able, but a headache and un-intuitive. then you look back earlier, and earlier, until you find a easier to use move that is less headache and more intuitive to use. you now switch to this earlier move, because you realize that your goal is not just to win, your goal is to find the most optimal path. one win is better than the other. to eliminate time troubles and the chance of your own human flawed mind. keep things as simple as possible.

now that was talking about if you win, study your win for flaw.

ok now we will talk about studying your loss, and fix it.
players will also study your loss. they know that you will deviate, because you know it wasn't good last time.
this is not so much for them to get you with a positional trap that they predict. this is more over to put you into tactical territory where you are trail blazing and are unsure of the way. this gives the tactical opponent to level the playing field.

so it is good to know your changes ahead of time when fixing a P1 loss, so that when they do this, you are less lost in the new position you twisted into, because you had planned it already with some for thought. this will increase your chances, like having a partial map when trail blazing through a tactical jungle of unknowns. this will also help you from falling into time trouble.



back to designing your P1 3rds. understand that,

there will be 3rds of sure loss, value=0

there will be 3rds of great struggle that can win, value=1

there will be 3rds of medium struggle that can win, value=2

there will be 3rds of small struggle that can win, value=3

and

there will be 3rds of mini-max struggle that can win (Perfect Play) value=4


your goal should always be a perfect value of 4.

but hey, don't get me wrong, some times a 4 is goina be struggle. but just remember for these openings, less than 4 is a headache to deal with. especially if it was a game in a tourney with $5,000 prize on the line?? see what i mean? keep it as simple as you can when white, and realize that sometimes the perfect move is not obvious. to find the non-obvious moves of perfect play takes much training and many mental tools.



these value examples may not be 100% accurate, but they are very close to it.
and L8 is a 4 for sure.
Image and video hosting by TinyPic

Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
zoeyk

Posts: 2,070
Registered: Mar 4, 2007
From: San Francisco
Age: 43
Home page
Re: Pente - Understand vs Memorize
Posted: Oct 8, 2010, 4:35 AM

once we get further down the thread ide be interested to see someone try and create a summery of the valued information found in this thread from all players involved.

im not always good at summarizing, my mind tends to scatter with information. i forget who it was but i remember a player who did this and it was rather good.

perhaps you jasonb can do this. and we can see how much has soaked in, to see what is the natural learning process for you in the way we are laying out information.

perhaps we will discover that we need to adjust our methods of explaining.

Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
Replies: 122   Views: 238,637   Pages: 9   [ Previous | 1 2 3 4 5 6 7 8 9 | Next ]
Back to Topic List
Topics: [ Previous | Next ]


Powered by Jive Software