Re: First Move Advantage in Pente
Posted:
Jul 14, 2011, 6:32 AM
To Karl.
I hope I am not coming across too strong, sometimes my style gets a bit assertive. I do respect you as a player . I will try to be more concise here also.
Karl said: "not novel positions, just positions where white's best response isn't absolutely clear"
In response to this comment and your previous comment that P1 can counter every conceivable attack. Sorry to be pedantic, but either they can counter it or they can't, so your posts lead me to question you further. I agree with your point about the cap situation with the wedge opening, I was going to make that same point myself. But, however exceedingly difficult that situation becomes, if it is mathematically true that P1 still has the advantage at that point, then the solution while difficult, will be found if it is not too obscure. Or, if P1 does not have the mathematical advantage at that point, then there was some previous point at which the way forward was too obscure to solve for P1. I therefore posit that P2s success must ultimately come from creating a successfully obscure position at some stage in that game.
Re: First Move Advantage in Pente
Posted:
Jul 14, 2011, 7:05 AM
I don't have the answer to the question of whether P1 has the advantage in the standard wedge opening. I think at this point our disagreement is mostly or entirely semantic: you claim that P2 can gain an advantage over P1 in the wedge using an "obscure" line, while my understanding of the wedge line (which I hope you believe is sound) is that the wedge opening at this point in its history represents the a concept distinct from the obscurity you discuss.
Compare a forest where visibility is reduced to 20 feet because of the density of the trees to a forest where visibility is reduced to 10 feet because of a thick fog. The wedge is analogous to the first case -- P1 could very well lose a line that they have studied extensively because it is very had to see a clear victory even after 12 moves of excellent play. P1 can see the layout of the forest quite well but still can have a hard time finding the exit. I consider a truly "obscure" line to be the analogous to the second case: a line that might have a much weaker mathematical probability of giving P2 an advantage down the line, but one where P1 might nevertheless head down the wrong path and quickly get lost due to the thick fog.
I think the best openings for P2 to play are in one of these two categories, and the wedge is firmly in the first, and truly "obscure" variants are in the second.
It's no use going back to yesterday, because I was a different person then.
Re: First Move Advantage in Pente
Posted:
Jul 14, 2011, 8:11 AM
Yes it seems a definition of obscure needs to be agreed. I provided one in an earlier post and am therefore expecting that you are working to that definition given that your point seemed intended to counter mine rather than specifically introduce a new concept of obscurity.
Of course I believe your knowledge of the wedge is sound. I am in no way questioning that.
Out of interest: you gave an example to match your dense forest analogy, but what kind of position constitutes a fog?
I see the forest density example as a corollary of up2ng's complexity, and as I said to up2ng, I see complexity as one way in which the solution can be obscured. Another way is through lack of information, or the search breadth or depth is too great at that point for a human player to get enough certainty in real time. For me, obscurity does not imply a low odds line for P2 but is a general term that is a property of the position and that relates to the ability of a skilled human player to identify a solution to that position, regardless of what causes that obscurity.
Posts:
2,220
Registered:
Mar 4, 2007
From:
San Francisco
Age:
45 Home page
Re: First Move Advantage in Pente
Posted:
Jul 14, 2011, 11:23 AM
to karl, and others too..
please define your usage of the term "novel" and or "novelty".
(note that; in the next post below this one i have made a very short summed up version of this post. so if you cant handle reading all this then you can read the next post instead . and if that lower post is confusing then read this bigger post for more explanation.)
for me there are two basic usages of novelty as i have discussed in a previous thread. with in those two i find that they can be subdivided even more;
here is my take on it, usage #1; if live rated play with no data base access, then a novelty move...
(resulting in a novelty position it goes with out saying)
...is simply a move that the opponent has not yet seen or has no recollection of.
in this situation we can think of it one of two ways;
we can say its novel if the data base shows they did not yet face it from either seat. but this does not mean they didn't see it while browsing the data base, or by viewing it while spectating a game played by others.
or;
in this situation we can say that its only novel if they really truly have never seen it before, and at that point only the opponent knows if its a novelty.
how ever and on to the next usage #2;
if turn based with access to the data base (and we assume the opponent is using the data base) then a novelty is now a move that has not yet been recorded to the data base (by any one, ever).
how ever here again we might reconsider, and think of the inclusion of some (or even all?) viewable games not in the data base. this means brainking.com games pente.net and others not listed.
we can not assume the opponent has seen and memorized every game from every where recorded (although improbable its possible?). in this situation we can only assume they just view the data base, view popular games at brainking tourneys or by top players in regular games, and turn based games of their own at such sites. so here once again, only the opponent knows if the move is actually novel.
i seem to find this important in this topic when bringing up the wedge and using this term.
any one agree?
and to over simplify all these variables, we can perhaps say either one of these 3;
#1 novel to opponent (NTO)
or
#2 novel to world, by way of recorded games(NTW)
or
#3 novel to the existence of all observers of pente whom ever lived, recorded games or not all included. (this one is over kill perhaps in basic theory discussions but still has its place in truth) (NTE)
now, to change this stuff slightly, and sorry karl for this being long, my bad man, but;
what if the term is not considered limited to moves not yet tried. what if it is used for a move that has been tried and recorded before, and in this case its being considered as still remaining a novel move in the future when repeated in a new games, because it has had a very low usage, of lets just say one game one time. and in that game the player whom made the novel move ended up winning that game. in this situation one may have a low understanding of the move in terms of the solution to it. and since it has won that game, the human perception of it might be that its strong, whether or not this be the case. so in this case, is the second time, third time, etc, of doing the move, is it still considered being called "playing a novelty move"?
since we are going down the road of humouring me about calling a move "that has been done before and won that game" a novelty and that we play that exact same move once again and still we are calling this a novelty in game #2 or #3 ext... (how many games of it till its no longer novel? until its beaten? or a certain number of games of it have been recorded?)
we can now subdivide this to look at. meaning two kinds where one type of novelty is sure win, and the other is not sure win.
does sure win or not sure win change the usage of the term novelty in certain discussions?
and yes even after all this i have yet another idea of how the term might be used, but will need to post a viewable board for that one to explain.
any how again sorry for being long written, thanks for reading.
~z
Message was edited by: zoeyk at Jul 14, 2011 5:39 AM
Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
Posts:
2,220
Registered:
Mar 4, 2007
From:
San Francisco
Age:
45 Home page
Re: First Move Advantage in Pente
Posted:
Jul 14, 2011, 12:35 PM
here are some of the above points and questions in short.
novelty is? new to; 1 person's own memory?, recorded sites that are publicly viewable? , to all of existence ever in regards to pente? (even OTB is included in this one)
does a novelty have to be a new move? can be a move the opponent knows but it had low history and was successful in not showing its solution yet?
does it make a difference if the novelty is sure win or not sure win?
Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
Re: First Move Advantage in Pente
Posted:
Jul 15, 2011, 2:53 PM
I agree with the distinctions you make regarding the degree of novelty.
In the situations I have been describing with only real-time games without access to the db, I think it doesn't matter whether anyone else has seen the move, or knows an answer to it, or doesn't know an answer to it etc. The only thing that matters to me is if it is new to my opponent, and they cannot see a solution even though mathematically P1 might still have the advantage.
When I said P2 is forced to find a novelty, it was meant in the context of master level play where masters are studying the db and spending a lot of time privately working on variations around particular openings. To counter this P2 is forced to find a novelty that is obscure, since there is every chance that other moves, although confusing to people who might not know them, would not be obscure to someone who has seen them before via private practice and research.
At the point where P2 chooses to play such a move she has no certainty that this is really an absolute novelty since, as you pointed out Zoey, the opponent may well have covered this in private research. But the more novel a position is with regards the db and known lines played before, then the better the chance that your P1 opponent is not familiar with it.
But IMO novelty is a necessary but not sufficient condition for success. Something that is new is not necessarily obscure, and it may be straightforward to find a successful reply in real time. I argue that it is sufficient to play a move that is obscure enough that the opponent cannot resolve for certain which move is better in the amount of time available. The attribute of novelty for the move simply increases the chance that the opponent hasn't spent time previously on solving it.
Again, I don't mean that a move is obscure because it is not known, but because its solution is not clear.
Posts:
542
Registered:
May 9, 2002
From:
Northeast USA
Re: First Move Advantage in Pente
Posted:
Jul 15, 2011, 5:46 PM
Good stuff Alison. I think you've done a fine job of describing P2 strategy. Note that this line of thinking wouldn't even be necessary unless P2 acknowledges from the start that he is at a disadvantage. Knowing this, your description of finding novel and obscure moves in order to thwart your opponent is accurate.
"In the situations I have been describing with only real-time games without access to the db, I think it doesn't whether anyone else has seen the move, or knows an answer to it, or doesn't know an answer to it etc. The only thing that matters to me is if it is new to my opponent, and they cannot see a solution even though mathematically P1 might still have the advantage."
I particularly like how this is worded. This was my main point earlier. The "advantage" in pente is what it is, it's a mathematical certainty for one player or the other in a given position. Whether or not the player with the advantage possesses the skillset necessary to capitalize upon it is irrelavent to whether or not the advantage exists. If two 1100 rated players face off, there may be plenty of situations where the advantage is quite clear and strong to more experienced players and yet this player makes mistakes in this situation often. Just because the mistake is made, it is not correct to claim that the advantage did not exist simply because the player was unable to find a correct solution in real time. This is still true at all levels, right up through the best masters of the game. The advantage is there, and then each player's skillset comes into play so that one player may "test" the other who has the advantage, and that player must use his skills to try to maintain it.
---------------
A quick side note -- when we speak in terms of search depths and finding moves that are unclear (obscure) so far into the future that a player cannot search deep enough to see the answer ... this sort of rationelle really only applies to computers. In my opinion, it's important to understand that humans do not think about the game in these ways. That's why the debate of human vs. computer is always so interesting. Human players, even the best ones, really do not search very deep (or very broad) when determining their next move. I would guess that in typical situations, a top player might try to look ahead about 4 to 8 ply and that's about it. In some rare cases perhaps they will try to see slightly further ahead, but generally it is nowhere near as far as a reasonably powerful computer. A strong human player thinks in terms of concepts and not (just) search depth. These concepts are numerous, especially during the opening, including forcing captures, draw moves, taking an inside line, the strengths of vertical/horizontal vs. diagonal play, keystone captures, false threats, and on and on. Even with only a couple of stones on the board, a top player is thinking very conceptually in trying to set up the position a certain way that he feels will be advantageous -- not because he has looked ahead 20 moves, but because there are certain concepts and principles at work which make some moves stronger than others.
From this, it would seem unlikely that a human player will always choose the "perfect" play, but it is amazing how often a top player can do just that by just thinking conceptually -- some would say it's both a skill and an art.
Re: First Move Advantage in Pente
Posted:
Jul 16, 2011, 8:06 AM
Thanks Dean, I appreciate your feedback. I don't claim to have originated any of these ideas, I only seek to summarize them in terms I can understand and this will hopefully assist some other lesser beings like myself to wrap their heads around it.
I hope to one day write as clearly as you.
Your example using the two 1100 players is well made.
On your side note: As you will remember from a while back, I have a good understanding of the computer play side of things so I am familiar with the human v computer comparison in game analysis. Note though that there are well know situations in which human players indulge in search. This is very true in chess. A large part of chess is concept based and memory based and involves a lot of heuristics and pattern recognition. But within this, such as planning mating sequences and endgame moves there is quite a bit of deep search required. Study some of the games of Morphy or Steiner and you will find several examples of deep sequences of 15 moves per side to mate. I am not saying this is the default or that it is frequent, but it is not altogether uncommon either. I only list search as being something that may also not yield a successful result if the situation is too obscure.
. I will try to be more concise here also.