< Earlier Kibitzing · PAGE 14 OF 31 ·
|May-05-09|| ||rogge: Are you showing off your hardcore, <zarg>???|
My Stephen Hawking is next to Carl Sagan, Erik Damman, Fritjof Capra, Ken Follett and Wilbur Smith in my shelf :)
|May-05-09|| ||zarg: Of course <rogge>! :) Any man with respect for himself, must claim a bookshelf section in the living room, free of romance, cooking recipes and such!! |
I bet, my hardcore section would even turn on <frog>.. foreplay would be TAOCP by Knuth and 4 compiler books. ;)
|May-05-09|| ||frogbert: hm... we have all our technical litterature in the basement - and in my wife's office at work!|
it's really unfair - she gets lots of computer books directly sent to her for free, by the publishers. if she decides that some title looks interesting, it's often enough to send an email and ask some question about it, and they ship it to her, with an extra title or two put in for good measure. (they know she teaches computer science.)
filthy benefits... ;o) btw - she even beat me in the master level compiler class we both took (a couple of years apart), i was crushed 1.0 against 1.1 - i mistakenly thought i'd have a chance for once. :o)
|May-05-09|| ||angslo: sorry, <alex> , for abusing your forum - i mean it, friend and regret it when i look back (i mean my posts perhaps served the purpose i intended, nonetheless it was unfair to abuse your forum for that. I have always prided myself in having higher values than vultures, and i have no intention to stoop to the level of vultures ever).|
thanks and have a nice day, <alex> :)
|May-06-09|| ||frogbert: alexmagnus, did you answer the question i asked a couple of days ago? you said:|
<My deflation argument was based on an assumption that all players perform according to the table. Then the K=15 group loses mo points than K=10 group gains.>
and i wasn't completely sure if i understood why:
let's create a single player of all the 2300s and similarly of all the 2500s. the 2300 player then scores in total
a) 0,2% more than
according to the tables. why does <any> player in case b lose or gain rating points?>
do you have any other assumptions, for instance on people going up or down first, or anything else?
i'm just curious if there's something basic that i've misunderstood...
|May-06-09|| ||alexmagnus: <frogbert> I found a mistake in my reasoning myself, which I actually corrected in my e-mail to Chessbase (which was posted there). Of course, if the do exactly as expected, the ratings do not change at all. But if those duels occur too rare to have a representative sample of games, deflation is actually more likely than inflation since a single game is more likely to be lost by the lower rated player than won by him. |
So the quesion is - do such games occur often enough to make a representative sample? If yes, diffetent K-factors (assuming the expect values are correct) don't change ratings in any way. If no, deflation is more likely.
|May-06-09|| ||frogbert: <But if those duels occur too rare to have a representative sample of games, deflation is actually more likely than inflation since a single game is more likely to be lost by the lower rated player than won by him.>|
are you sure about this?
of course, between a single pair of players, that might be true, but if we consider all players with k=15 as one player, and pretend all his/her (their) games against players with k=10 as games against one player, shouldn't that still be equal on average? (assuming all games considered to be played during one and the same rating period)
i have a feeling one needs a more rigorous argument to assume/expect either of the two cases...
|May-06-09|| ||Magnusch: <frogbert> Sorry for expressing my opionion here. If I understand what you are talking about here (the same question I asked earlier to alexmagnus) then I agree 100 % with you here. (Expected value has to be zero). The number of games has no effect on the Exp. Value (n*0 = 0). But this depends on, if my understanding of what you are talking about is what I think you are talking about. But I think it is. Nevermind.|
Now I'm to tired for expressing what I think. I'm just a patzer with low knowledge about the rating system (andd math and english too).
|May-06-09|| ||zarg: <frog>
I don't think <alex> is right on this, and I regretted accepting his argument w.r.t. global deflation.
The rating change is K times the diff between actual and expected score, so the relevant question appears to be which group is overrated/underrated. This can and should, be calculated by simulations.
Will provide a followup on your latest book post, in my forum.
|May-06-09|| ||alexmagnus: <of course, between a single pair of players, that might be true, but if we consider all players with k=15 as one player, and pretend all his/her (their) games against players with k=10 as games against one player, shouldn't that still be equal on average?>
In which sense equal? What should be equal to what? The result of "global players" agaist each other to the expected result? Yes. But how many games are that? If these are thousands, no problem. If less, the sample is simply too small to make some probabilistic conclusions...|
|May-08-09|| ||alexmagnus: <frogbert> So, what should be equal?|
|May-08-09|| ||frogbert: i was referring to the win and loss of points. i agree that how i used "equal" in my previous post wasn't enitrely clear. :o)|
|May-10-09|| ||metatron2: Doesn’t seem like too many constructive discussions have been formed lately, so maybe someone is still interested in the old time-travel discussion.|
<zarg> < <meta: We can look at the universe as a whole [..] I think we can say we know this for sure or call this an "axiom", if we want to be able to have any kind of discussion about such time travelling issue> The problem is <"in fields such as quantum physics and relativity theory, existing assumptions about reality have been shown to break down, this has usually been dealt with by <changing our understanding of reality> to a new one which remains self-consistent in the presence of the new evidence">>
Is there any problem with my "what we know for sure" part, regarding the relativity theory? The entire universe can't have a Frame of Reference? Doesn't relativity theory itself say it is impossible going back in time because it is impossible breaking the light speed?
As for quantum physics, u said it yourself: It has fundamental differences from macrophysics, with the second law of thermodynamics as the most noticeable one.
Whether in the future we might "change our understanding of reality" in view of self-consistent and <new> evidence, is not really relevant for our discussion now if such evidence doesn't exist yet, right? I mean, what is the purpose of arguing "we might change our view in the future"? We discuss based on what we have <now>.
I was trying to base my claims here on things we do know. I am not impressed with unsubstantial theories whether they are coming from great scientist or not. I think that it is all about the initial assumptions and the path chosen there. For example, even Newton, who I consider to be far-far above the rest of all time geniuses, was getting nowhere with his alchemy work, because (in that case) he took the wrong path to begin with.
And that is basically what I think about the wormhole theories, specifically:
1. As I said, no matter how you turn this around, a single object getting back in time, <has to mean> that there exist (at least) 2 versions of the universe, one in time T and the other in time (T-s).
2. Based on anything we know, I don’t see <any> reason to believe this is the case. There hasn’t been a shred of evidence for that, and everything we do know about the universe suggests the opposite. It is the burden of those presenting such theory to prove otherwise, and <not> the other way around.
3. Wormholes being a “valid solution in general relativity” doesn’t say much, now does it? I mean, having a theory that there are 20 aliens flying around my head and reading my thoughts right this moment, is probably a “valid solution” as well, yet, based on what I do know, I tend not to assume this is the case.
Note that I don’t think causality is the reason it is impossible going back in time (I explained my reasoning for this impossibility). Causality is just another (strong) evidence for its impossibility, and trying to assume that “our thought experiment is invalid” here, is another (very much) unsubstantial assumption.
Will be nice to get your educated opinion about those _specific_ points, based on your real (substantial) _knowledge_ and understanding in (physics and other).
|May-12-09|| ||Alexandru Z: <Apr-11-09 alexmagnus: <Good luck, what field of math do you work with?>|
As I said,I still didn't complete my study but... Number theory it is/ will be.>
<alexmagnus>, I noticed you have participated in a number theory seminar organized by Gavril Farkas. It seems to me that, at least as far as the algebraic side of number theory is concerned, Berlin is very well represented by its three major universities, and so a good place to study number theory.
I do not know how the system works there. In case you are/will be writing a thesis in number theory, I was wondering whether you have considered, as an option, writing something that is chess related. Perhaps this is just a stupid idea, and I apologize for mentioning it, but, taking into account your contributions to these chess pages, I think you would enjoy working on topics which are directly or indirectly related to chess.
|May-12-09|| ||alexmagnus: Connecting chess and number theory? I see no way at the moment but never really thought about it.|
|May-13-09|| ||Ziggurat: I used to have a book on chess and mathematics, and it had some material on number theory and algebra. It was written by Evgeny Gik; see references at the end of this Wikipedia article http://en.wikipedia.org/wiki/Rook_p...|
|May-13-09|| ||alexmagnus: Heh, previously I knew Gik only as co-author of Karpov's chess books. I knew he was a mathematician but never heard of that "Chess and mathematics" series before.|
|May-13-09|| ||whiteshark: The book (original: Russian, published 1983) is also available in German:
Occasionally you'll find it for sale on ebay.
|May-13-09|| ||alexmagnus: Actually, Russian original is even better to me, it's always better to read books in original :).|
|May-13-09|| ||whiteshark: If you have a djvu viewer... http://problem64.beda.cz/silo/gik_s...|
|May-14-09|| ||alexmagnus: I don't have it... I'll check it in my uni library maybe there is the book there.|
|May-14-09|| ||whiteshark: You can download the viewer here: http://www.celartem.com/en/download...|
|May-15-09|| ||alexmagnus: <visayanbraindoctor> OK, let me coment on your profile:|
1.<1. Regarding the Chess World Championship: The true Chess World Champions are the holders of the Traditional Title that originated with Steinitz and passed on in faithful succession to Lasker, Capablanca, Alekhine, Euwe, Botvinnik, Smyslov, Tal, Petrosian, Spassky, Fischer, Karpov, Kasparov, Kramnik, and Anand.>
Dpends on what makes a true world champion". One can be the best player without beating the previous best player (if active for a too short period; f.x. Icaimagine Morphy would not get the title if it existed back then - he just disappeared too quickly. On the other hand, world championship is not always succssive, see Botvinnik regaining his title several times, so that Smyslov, Tal and Petrosian didn't beat each other but all beat Botvinnik. Succession???
<No other field of human competition can lay claim to such a Title>
What do you mean "such"? Successive? Actually all versions of checkers/drghts have a sucssive title. Boxing had one until it got split into 4 federaions, but even now it is successive within each federation.
<he sacredness of this Title is what makes it so valuable and the Chess World so unique in the World of Sports; and in my opinion all serious chessplayers and fans should defend its legitimacy and continuity>
Sacredness? It's chess, not temple.
Unique? See my previous comment,there is nothing unique in it.
Valuable? For whom? For chess? Tell me, for which purpose? Even the answer "tradition" doesn't hold much since the tradition is relatively young. To define a best player? It can be done simply by analyzing game (see below). If I were in power, I would maintain this tradition, <but> I wouldn't regard it as something sacred. We are chess players and not religious fanatics.
<And how does one become the true Chess World Champion? In general and by Tradition, by beating the previous Title holder one on one in a Match!>
Matches are preferred over Tournaments because of the Tradition of the World Championship Succession and because it would preempt the possibility of connivance among the participants of a Tournament to pre-select an eventual Winner by throwing away games and pre-planning grandmaster draws.>
None of the losers would profit from a pre-selected winner. In a Candidates tournament, "conspiracy" is possible, but in a World Championship tounament it makes no sense. Everybody who made it to the final tornament wants to become WC, it would be irrational for him to be a part of a conspiracy.
<The only exceptions to this rule:
A. In case where the Candidates and World Champion participate in an event that all the participants agree to be a World Championship event because of extraordinary circumstances. Thus, the 1948 World Championship Tournament was justifiable because of the death of the Title holder Alekhine, and the Winner Botvinnik was the true Successor to Alekhine.>
Why not to orgaize a match between two potential Alekhine's challengers instead?
<Likewise, the 2007 World Championship Tournament was justifiable under the extraordinary circumstances of the Chessworld trying to heal its internal rift over the 1993 Kasparov Schism>
I don't blame Kasparov for the schism, Short was the "criminal" here. Kasparov was only a "complice".
< Anand himself became the true World Champion in this 2007 Tournament and not in 2000 when he won a knock-out Tournament that FIDE deemed as a 'world championship'. Karpov lost his Title to Successor Kasparov in 1985, and never regained it in the 1990s events that FIDE labeled as 'world championships' but held outside the Tradition of World Championship Succession. All FIDE Champions that emerged outside World Championship Traditional Succession elaborated on above, strong as they were, were not true World Champions (eg., Bogolyubov in 1928, Khalifman in 1999, Ponomariov in 2002, Kasimdzhanov in 2004, Topalov in 2005). >
As I said, what is "true" lies in the eye of the observer. To me th world champion title matters nothing. AsI said, if in power would maintain it, but I see no significant difference between the world champion and any other top player.
...To be continued
|May-15-09|| ||alexmagnus: <B. In case the previous Titleholder defaults an event that by and large the Chessworld deems as a World Championship event in the Tradition of the World Championship Succession. Thus, Karpov was the true successor to Fischer who defaulted their World Championship Match in 1975.>|
What would happen if Fischer suddenly reappeared in 1978 and you were in power? Would you force Karpov to a match vs Fischer or would you force Fischer to go through a qualification cycle?
< . How to choose a Challenger? The credible, fair, tried and tested Zonals - Interzonals - Candidates>
The old format was never fair (forgot the "anti-Russian rule", forbidding more than 3 players from the same country to qualify?). And how do you know the results were really optimal?Note, Candidates have a similar random component as World Cup - they are KO, so that one has to be lucky to get the "right" opponents in the quarter- and semifinals.
<over the random World Cup>
I already posted on several pages proves it's not as random as many think (keyword "multiple finalists").
<and the elitist Grand Prix>
Why elitist? It features 21 players, but those 21 have to <qualify>. And for qulification, "everyone" is eligible. So I see no elitism here.
<4. The strongest players in chess history? In my view, we have to take this question in the context of the limits of the human Anatomy and Physiology. A concrete example would be the one hundred meter dash. The human body is designed such that the limit it can run is about 9 seconds for this event, and this forms a Stonewall for runners. In order for a human being to run faster, we would have to redesign the human anatomy into that of say a cheetah. One can rev up the human Anatomy and Physiology, say with steroids, and obtain slightly faster runners, but this regimen would hit an eventual Stonewall too; the same way that we could rev up the human chessplaying ability with computer assistance. 'Worse' in chess, any computer assistance ends once the opening is over. After the typical computer-assisted opening prep, every GM today has to play the game the way Lasker did a hundred years ago, relying on himself alone, with the same fundamental chess rules and chess clock.>
So yo thin the theory of middle- and endgame didn't progress for centuries? Chess theory is more than just opening, <vbd>, and you as apparently a strong player should now it.
<So who was the strongest player in history? I have reasons to believe that the 1919 version of Capablanca and the 1971 version of Fischer are it; however, I would guess that the likes of Lasker, Alekhine, Karpov, and Kasparov would all be close to the limiting Stonewall of human chessplaying ability.>
IMO the answer about th strongest player ever is <impossible to answer> Why, see my next comment.
<5. The best way to evaluate a chessplayer's strength is to analyze his games.>
I would be slightly stricter - to analyze his games <agaist his equals>. Against worse players one always plays better than against one's equal (and against an equal better than against the better one). The best way to judge player's chess ability is too look at is games against his equals - there he does not better and not worse than he actually is. That's why I think the answer about the strongest player is impossible to answer - the mentioned players had no equals in their time, therefore their games give no information about their true chess strength.
<Ratings are simply fallible predictors of an active player's near-future performance against other active rated players.>
.. To be continued
|May-15-09|| ||alexmagnus: <<6... Increasing the mass base of human chess players thus expanding the normal curve of chess players simply creates more possibilities of players playing like a Fischer in his prime, but will not create a mental superman who plays chess at computer levels. There will always be a sudden cut-off at the right side of the normal curve. As a corollary, the level of strength of a quadruple round robin 8-rounder featuring the 1921 Lasker, Capablanca, and Alekhine who all played close to the limit of human chess-playing ability would surpass any tournament nowadays except a similar one featuring the very best active players, say Anand, Kramnik, and Topalov.>|
Disagree. While the players "number one" of different times may be equal (on the right end of the curve), the players "number two" are already different - with fewer players around, the density is smaller therefore the gap between #1 and #x is bigger therefore a modern tournament of #1,#2,#3,#4 and #5 is better in terms of average level than the same tournament from the past. The #1 participants are about equal, but the rest is worse.
< Earlier Kibitzing · PAGE 14 OF 31 ·
from the Chessgames Store