Posts:
2,220
Registered:
Mar 4, 2007
From:
San Francisco
Age:
45 Home page
Re: Can Nosovs be beaten?
Posted:
Dec 13, 2009, 12:55 AM
why would forcing lines along the way (or not) at certain moves have a contingent impact on a foreseeable inevitable unstoppable cap win in all scenarios as a set rule? a forced cap win is not defined by consecutive forcing "lines" every move. a cap win's initiative transcends this commonly thought dependency of continual line force every turn. momentum can be more invisible than that but still retain 100% of the same driving force towards the final victory stone.
in that game there were forcing lines, and not, caps and not, but when white's mistake occurred on the 14th, the cap win was unavoidable.
so perhaps consecutive, and continuous, and words of these meanings should be changed for this term. as cap wins can be slow and subtle, yet inevitable and forcing.
VIC = Victory by Inevitable Captures
z
Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
Re: Can Nosovs be beaten?
Posted:
Dec 14, 2009, 2:40 PM
zoey, at this point I have to ask what the point is of all this terminology?
For me, it's the convenience of expressing a concept in shorthand, knowing that the reader will know what I mean. I think we can already see that there is widespread confusion about VCT and VCF let alone the multitude of other Acronyms you have proposed (I realize not all of them were serious).
Although I think there is a place in the pente playbook for some kind of shorthand term for a string of captures leading to an inevitable win, I think it's value diminishes as you broaden it's definition. The broader the definition the lower the precision and the higher the confusion. You will not have the certainty that a reader will know what you mean without the requirement of further clarification.
My advice to you is to keep things as simple and straightforward as possible. So I suggest confining this term to mean consecutive moves that are captures which lead to a 5 capture win.
Lastly, in reference to a previous question you asked - no I did not edit that line in my post, you must have misread it.
Posts:
2,220
Registered:
Mar 4, 2007
From:
San Francisco
Age:
45 Home page
Re: Can Nosovs be beaten?
Posted:
Dec 14, 2009, 11:34 PM
ive been working on a pente book. i use the forum to discus terminologies and theory for the sake of finding better accuracy. by taking it very slow and searching for new ideas on old things i find my self constantly re-questioning my original thought process on things. this is a good thing
this thread does not belong in the general forum. not sure if that is my fault or not, but it seems to be the situation in general.
any how, i appreciate every ones input, wrong or not. we pente people are a great "think group" when given the right subject matter kicked off in the right tones. good stuff.
any how thanks, and yes im sure i must had misread you.
zoey
Message was edited by: zoeyk at Dec 15, 2009 4:08 PM
Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
Re: Can Nosovs be beaten?
Posted:
Dec 15, 2009, 2:09 PM
I look forward to that book zoey!
Anyway, where were we? This thread is about beating nosovs and how it could be achieved. So far we have heard from a few people, and I heard from a number of people who haven't posted into this thread.
To summarize the views expressed so far:
1. That there are problems with the rating calculations that make it difficult to surpass nosovs. I think I have countered this point and I think successfully established that this cannot be so much of a problem.
2. That frequency of rated play by nosovs and careful opponent selection play a part in protecting his position and discouraging challengers. My response to this, if this is true, is firstly that this is a smart approach if you want to stay on top. I remember Pete Sampras just played the grand slams for the last 3 years of his career and still managed to stay at or near No1. In the end to get to No.1 you didn't necessarily have to play Pete, but you had to beat players that could beat him. In pente you don't even have to do that, all you need to do is accrue enough points from playing others. Secondly, anything that can be done in this regard by nosovs can be done by anyone else. So I don't see that this would prevent someone from surpassing him, and using these tactics to stay there.
3. There is a whole world of learning to be done by anyone wishing to play at his level and or surpass him. We have learned this in part from zoey's posting of part of nosovs' work on Renju. Nosovs is clearly a very fine player and it would take a very great effort from an extremely devoted player to surpass him in real-time play over an extended period. I think this is the key. I know there are players who are fairly close to nosovs in standard at real time play, and even closer in turn-based. So it is not beyond imaging that someone could displace him at the top of the real-time tree. But can such a player stay there over an extended period? And, will doing so require the kind of commitment and learning discussed in nosovs' writing?
Is there anyone out there who can or will do it? Is there someone out there who will publicly throw down the gauntlet and challenge the top dog? Will anyone step forward and draw a line in the sand and say "I, player XYZ, this day shall no more till the soils of tyranny, and plough the fields of shame, I will no longer feel the brute whip of failure on my brow nor the dishonour of my family in my heart. No. I shall this day stand and fight. I will meet my oppressor and have victory for my prize!" (Or the words to that effect)
Posts:
2,220
Registered:
Mar 4, 2007
From:
San Francisco
Age:
45 Home page
Re: Can Nosovs be beaten?
Posted:
Dec 15, 2009, 9:44 PM
yes he can possibly be surpassed. no, i don't think that person cares enough to do it. watsu already explained this.
and nosovs writing does not explain all the steps needed to get to his level in pente.
1- it was written for renju
2- even if it was written for pente, it is still just the tip of the ice burg, the instructions are far from completed
and, is the question can he be beat? or is it "is someone goina beat him"? i think this is important.
can and will are 2 different things. you have already received an answer to the first question from watsu.
as for your second question, i think it will be a cold day in hell before a player seriously say (as if OCD), "Yes! it is now my life long goal till im old n grey and dieing that i will surpass nosovs. you'll need to pry my keyboard and mouse from my cold dead fingers first before ill stop my compulsive life long quest of such."
well, something along those lines hehe.
it is more likely that a player really will surpass him quietly, than to make a public announcement to you here that said player is going to. this does not count the jokesters that may chime in saying such for a laugh.
even if my goal is to take him in one set. which by the way it is one of my goals, this in no way means i aim to surpass him. he is just too good. in other words he hasn't left much space above him to squeeze into. hes at 98.5% Perfect Play, leaving only 1.5% to fit into above him, which is like gods level. matching his level would be hard enough.
and alison,.. why are we so concerned to surpass him? i thought we would had been content that a renju master has graced us with his presence here in the first place. if he didn't come here, we would be saying, 'can any one surpass melanie?' lol hehe
ok im done,.. for now
z
Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
Posts:
542
Registered:
May 9, 2002
From:
Northeast USA
Re: Can Nosovs be beaten?
Posted:
Dec 15, 2009, 9:52 PM
Hi Alison,
You gave a great summary of the important points. I'll tackle them in reverse order:
First, nosovs is the best player. He has frequently won sets against other expert players and has been able to do so more often and more consistantly than anyone else. His mastery and understanding of the game is so complete that it will not be possible for another player to become significantly better.
It is true that he does not play a large amount of rated games and he pays close attention to the particular opponents that he faces. However, I do not believe that he purposely dodges games or is so overly concerned with his rating that it prevents him from playing. I think his frequency of play has more to do with his personal schedule and the time zone in which he lives. When he is available, the game room is often empty or there is no one near his level available for him to play. So, he takes matches where he can get them, even if the players might not be as strong and he knows he is likely to win from both sides. When he does play against the strongest opponents, he will often go on streaks where he will win several matches against these players also, sometimes he has found a new opening that the other expert players are not yet familiar with, for example. These streaks of wins, especially the most recent ones, contribute heavily to an increased rating.
Lastly, there are serious problems with the rating system here which make it even more difficult to pass nosovs by a significant margin -- you could not have countered this point because this point is extremely clear. It is indeed a serious problem. Besides the fact that nosovs is the best player and it is not possible to become significantly better than he is (see above), if one were to try and maintain a rating of say, 100 - 200 points above nosovs' rating (say 2500), the rating system is massively detrimental to being able to do this.
The reason why it is such a problem, in short, is because ratings changes here are based on individual games of pente and not based on SETS of pente as they should be.
If you examine the rating change formula (see General FAQ), you will see that the only variables in play are the current rating of each player and whether the current GAME (not SET) ends in a win or a loss. This formula would actually work extremely well if the game itself were fair. But it is not. A single game of pente has a player 1 advantage. (A set of pente, on the other hand, presents a perfectly level playing field.) There is no variable in the formula which accounts for the game itself having a certain expectation to yield a player 1 victory. But it's not that simple to fix either. That's because a simple variable would not accurately reflect what happens in practice. Although the game's P1 advantage remains constant in all games when looked at by itself, it actually has quite different effects based on the skills of the players. In other words, if a 1200 player is playing against an 1100 player, the P1 advantage plays almost no role because neither player's skill is strong enough to take advantage of it. But, when a 2200 player plays a 2100 player, the P1 advantage actually dominates the game. It is extremely rare for either player to be able to overcome it when playing as P2, so the expectation is to split the set most of the time. Two matches between players rated just 100 points apart -- vastly different expectations due to the varying degree to which the P1 advantage plays its part. Because the rating system does not handle this correctly, the result is that expert player end up facing a glass ceiling imposed by the rating system itself. It has been thought that this ceiling is around 2150 and that players would not be able to maintain ratings above this for prolonged periods of time over the course of dozens of games. So far, nosovs might be the only one who has been able to do so, and not by much. In fact, nosovs is probably vastly better than all other players and yet often only has 25 - 50 points higher rating than the next "tier" of players. This is entirely due to the problem with the rating system. If ratings were based on sets you might see many of the red players spread out quite a bit and nosovs might go way out in front, maybe to 2600 - 2700 for example.
A quick example, if nosovs (rated approx. 2300) decided to play several rated sets against me (approx. 1900), it would be a terrible decision for him. Even though he is the stronger player, I should reasonably expect to win at least 20% of games against him (and it might be more like 40%) because the P1 advantage is a very significant advantage for an expert player like myself. In this matchup, he would only gain 2.91 points for a victory and would lose 29.1 points for a defeat. Thus, he would have to win 10 times for every time he lost, just to maintain his rating. This might be accurate on a level playing field (such as if we played SETS) but is certainly not accurate for individual games.
The point is, for someone to maintain a rating of 100 points more than nosovs, it's just not going to happen with the current rating system no matter how good the player might be.
Posts:
2,220
Registered:
Mar 4, 2007
From:
San Francisco
Age:
45 Home page
Re: Can Nosovs be beaten?
Posted:
Dec 15, 2009, 10:06 PM
so, your basically saying in a nut shell if i may, nosovs is near to perfect play in all his moves, thus leaving little to no room to surpass him, and that yes the rating system is flawed, and set based would be better for all of us.
and i agree.
z
Scire hostis animum - Intelligere ludum - Nosce te ipsum - Prima moventur conciliat - Nolite errare
I'm curious why anyone would wanna dissect and rally to the cause of beating Nosovs in particular.......Generally I can appreciate the fact that he's one of the top 2 (that we know of) online, but if the goal is consistent and constant improvement in one's game then the opponent's role diminishes significantly in the equation(doesn't it?). I dunno just strikes me as a bit of a disarray......
Its been a few months now with the new rating system. What is the verdict?
I have three questions: 1. Is the rating system meeting expectations? 2. Are people happier with this system? 3. Has there been any meaningful change to the top few ranking positions as a result of the new rating system?
Re: Can Nosovs be beaten?
Posted:
May 26, 2010, 10:35 PM
Not sure why this is in this thread, but... 1. yes, it meets my expectations 2. yes, I'm happier with it 3. no, there haven't but it seems to be by people's choice of who and how often they wish to play rather than a factor of the new system
Message was edited by: watsu at May 26, 2010 4:35 PM
Retired from TB Pente, but still playing live games & exploring variants like D, poof and boat
Re: Can Nosovs be beaten?
Posted:
May 28, 2010, 5:04 PM
Well it is in this thread because you and up2ng and zoey all made the argument that nosovs can't be surpassed in large part due to the rating system not being set based. I am just following up on this. Thanks for responding.
Re: Can Nosovs be beaten?
Posted:
May 28, 2010, 9:08 PM
Okay, got it. So, yes now I do think that it would be possible for an excellent player to surpass nosov's rating- if s/he played a lot of games. Does this actually mean they would actually have to be significantly better at pente than nosovs is? Of course not. I suppose it boils down to no high caliber player feeling it to be worthwhile to pad their stats and/or play enough games to arrive at/surpass nosovs' level (so far at least).
Retired from TB Pente, but still playing live games & exploring variants like D, poof and boat
