The 2017 U.S. Championship was a 12 player round-robin with Caruana, So, Nakamura, Robson, Shankland, Xiong, Kamsky, Onischuk, Naroditsky, Akobian, Zherebukh, and Shabalov. ... [more]
Player: Gata Kamsky
< Earlier Kibitzing · PAGE 27 OF 27 ·
|Apr-15-17|| ||john barleycorn: < thegoodanarchist: <<AylerKupp: <Sally Simpson> Do computers get lucky> Everything, of course, depends on your definition of "luck". ...>|
Here is what found its way into German jusrisdiction: Luck is independent of skill and memory of the players. and game of luck involves a serious disagreement of its outcome by the players and a "substantial" amount of money is involved.
|Apr-15-17|| ||Tiggler: <AylerKupp>:<I frankly don't know what happened to you>|
He stopped playing chess, is what happened. See his bio.
|Apr-15-17|| ||SirRuthless: <AylerKrupp> Trying to reason with people who don't want to be reasoned with is a waste of time. They will take all the ground you give and mock you in return.|
|Apr-15-17|| ||john barleycorn: < blunderclap: Sorry guys, I thought I had this all figured out, but a fatal flaw has emerged in my calculations. Goedel just told me that he just made that silly theorem of his up to @#$% me in the ass!>|
Take it up with <morfishine> since he pulls his *wisdom* out of the same hole.
|Apr-15-17|| ||AylerKupp: <<morfishine> Computers do not get "lucky">|
They probably don't in the general sense bit, if you remember, I limited the domain to chess engines, not computers in general, and put it in the context of non-determinism and the effect that it has on the alpha-beta algorithm. And it requires that one associates "luck" to some extent with "chance". I agree that this association is somewhat tongue-in-cheek but I don't think that it is entirely inappropriate. There's even, as I just found out, a class of nondeterministic algorithms (see https://en.wikipedia.org/wiki/Nonde... for a very superficial discussion) that even have useful applications.
If you think that having chess engines spit out different move evaluations, rankings, lines, and even different top-5 moves (as I'll discuss in my next, hopefully very brief post, purely by chance does not constitute "luck", the I would be interested in knowing your definition of "luck".
And if quantum computers ever get built, they would surely generate non-deterministic results due to the uncertainty principle operating at the quantum level. But that is still off in the future and certainly an off-topic discussion for this site.
And, BTW, I'm not hung up on the subject, I just think that it's fun to speculate.
|Apr-15-17|| ||AylerKupp: <john barleycorn> I don't necessarily agree that luck is independent of skill and memory. In many games of "chance", the best players can calculate the odds of different outcomes. They would typically play according to the best odds but sometimes they might win even though they're playing against the odds, as when filling an inside straight in poker when your opponent holds 4 aces.|
But let me put it in chess terms because <Sally Simpson> and I had a related discussion some time back. In human chess, of course, when you are faced with a clearly losing position, playing for a swindle is certainly understandable. Against a computer this would probably not work, but if a human was playing against a computer (perhaps he was given odds) and happened to find himself in a winning position, wouldn't the computer be justified (assuming that it was programmed to do so) to attempt to swindle his opponent? Chess engines would not typically do this, when faced with all possibilities being losses, they would pick the line that delays the inevitable as long as possible; i.e. the horizon effect as I believe was originally defined.
But let's assume that an engine had been programmed to attempt a swindle when it was in a lost position and its human opponent overlooked his best move and the computer won. Would you in that situation consider that the computer got "lucky"?
|Apr-15-17|| ||morfishine: <SirRuthless> Just do us all a favor and STFU|
<blunderclap> Just do us all a favor and crawl back under the rock you came from
|Apr-15-17|| ||john barleycorn: <AylerKupp: <john barleycorn> I don't necessarily agree that luck is independent of skill and memory. In many games of "chance", the best players can calculate the odds of different outcomes. ...>|
Yes, calculating all the outcomes is satisfying in a sense. the problem is just to get the *right card* and that is the factor of luck.
|Apr-15-17|| ||Jambow: <He's a typical left coast liberal, fixed in his own ideas who will blab on eternally trying to convince you he's correct. He won't acknowledge your viewpoints at all, which is typical of Liberals, and couches his condescending arguments in a veil of self-deprecation>|
<Wow! I don't know what I've done to elicit such a strong reaction from you or anyone else. Yes, I am, in general, a liberal and yes, I live in the left (west) coast and yes, I have a self-deprecating style. I would like to think that this is not a veil but evidence that I fortunately don't take myself too seriously, but perhaps>
Agree or disagree sounds like <Morfishine> garnered access to <AylerKupp's> personal profile.
No offence to <AylerKupp> since I find he has some insightful posts, but yes for someone who doesn't take themselves too serious you go on for a loooong time. Your arm is off no it's not... that too.
I find fault with <Morfishine>, your list is compelling, but to leave off the fact that liberals also run to compliment commend and bolster each other uncritically, complain about a single post they deem off topic by those outside of the pack, while allowing those of their own ilk to go on for page after page of verbose off topic nonsense without saying narry a word deserved your mention as well. I didn't notice either so shame on me.
<Best endorsement for <aylerkupp> I've ever read!> Thanks Jim nobody does it better. ;0] and right on cue too, just ring the bell.
Sounds like the Rogoff septictank is full the sewage is backing up here now.
Yes and Happy Resurection weekend to those who it matters too, the rest of you have a good weekend.
|Apr-15-17|| ||saffuna: <...your list is compelling, but to leave off the fact that liberals also run to compliment commend and bolster each other uncritically, complain about a single post they deem off topic by those outside of the pack, while allowing those of their own ilk to go on for page after page of verbose off topic nonsense without saying narry a word deserved your mention as well.>|
That describes the rightwing club on Rogoff perfectly.
|Apr-15-17|| ||zanzibar: <<AK> ... Chess engines would not typically do this, when faced with all possibilities being losses, they would pick the line that delays the inevitable as long as possible; i.e. the horizon effect as I believe was originally defined.>|
?? I don't think that's the same "horizon effect" as used in reference to engines.
|Apr-15-17|| ||Jambow: This was a very great US championship, So won a tough battle and Onischuk took it to him, as did Akobian. So just keeps racking up the accolades. I expected a day when Magnus might have someone to potentially challenge him, but So soon it is upon us. Hope you qualify... |
Frankly I thought Caruana was going to be the man, and that still might be the case. Who would think the US championship would ever have three world top ten players. Thank you Caruana for coming home, thank you So for making this your home and Nakamura for your chess. How many critics you endured, who said nothing special, not a 2700 not a 2750 and not a 2800. No way a top ten never a top 5 but alas you have broken 2800, took up residence in the top ten, been the worlds highest rated blitz player, highest rated rapid player and world #2 classical all in the very recent times.
I look forward to when your crazy tactical skills gel with you well developed positional game. Oh how the critics have went the way of Elvis.
US chess champion Wesley So!!! Not bad for a poor kid from the Philipines, with a humble attitude and thankful heart.
We are still a land of opportunity thankfully So who is so thankful.
|Apr-15-17|| ||morfishine: <Jambow> I will always support you, especially when you are being attacked|
US Navy and spec ops always stick together
|Apr-15-17|| ||thegoodanarchist: <john barleycorn: < thegoodanarchist: <<AylerKupp: <Sally Simpson> Do computers get lucky> Everything, of course, depends on your definition of "luck". ...>
Here is what found its way into German jusrisdiction: Luck is independent of skill and memory of the players. and game of luck involves a serious disagreement of its outcome by the players and a "substantial" amount of money is involved.>|
Hi Barleycorn. Of course, I was talking about people taking precautions, in case on line dictionaries are hacked by machines. I am not sure that you're replying to that.
|Apr-15-17|| ||Tiggler: <jambow>:<Sounds like the Rogoff septictank is full the sewage is backing up here now. >|
and you can ring the bell for creating this classic post ++. Boinngg
|Apr-15-17|| ||Sally Simpson: Hi AylerKupp,
In a Human v Computer game if the computer was losing and the best line from a lost position was a swindle attempt then the computer would go for it.
The fact it may have presented the human with a complicated position they blundered and lost or drew then I would call that lucky because the computer did not mean to bamboozle the human. It does not know how to. It was simply playing what it thought was the best move.
Examples are hard to find today because the computers have no need to resort to lucky trickery v humans, they never get lost positions.
But one game from the past.
D Levy vs CHESS 4.7, 1978
Levy (in Total Chess by David Spanier) says he was very surprised by the move here...
click for larger view
55...h3+ draws. 55...Nxe7 actually loses but Levy believing the computer had found a way to draw it failed to spot the win that runs:
Levy played 57.Bd6 Kf6 58.b8=Q RxQ 19.BxR Kxe7
click for larger view
A draw. Lucky in respect the computer blundered turning a draw into a loss and the human trusted it without looking for the flaw.
|Apr-16-17|| ||Sularus: congrats to GM So!|
|Apr-16-17|| ||AylerKupp: OK, this is my last post on chess engine non-determinism, I promise. And I will try to keep it short (no promises!). I ran additional Stockfish 8 analyses on the same position, including some with 2-threads and 3-threads (24 analysis in total) and thought that some would be interested in some general results. If you are not, then this is a good time to skip to the next post.|
1. Non-determinism is evident in Stockfish 8 whenever you are running multiple threads, it shows itself when running 2, 3, or 4 threads; only when running a single thread are the results deterministic.
2. The speed-up when running to a search depth = 35 plies is 1.4X for 2-threads over 1-thread, 1.7X for 3-threads over 1-thread, and 2.0X for 4-threads over 1-thread. If you want non-deterministic results, this is the running time penalty that you will be paying.
3. Non-determinism is not restricted to evaluations, move rankings, or lines. In one of the analyses with 2 threads a move not previously indicated as being one of Stockfish's top 5 moves, 14.Qb4, replaced 14.Qb6 as one of the top 5 moves. So don't be surprise if additional moves "sneak in".
4. Evaluations for the same move when running an analysis using the same number of threads can vary quite a lot; the difference between the largest evaluation and the smallest can be similar to the value of the smallest evaluation.
5. But the evaluations, while variable, not surprisingly become much more consistent when you average them. So, the average evaluation for Stockfish's top move, 14.d3, was [+0.32] for the single thread analyses, [+0.33] for the 2-thread analyses, [+0.35] for the 3-thread analyses, and [+0.37] for the 4-thread analyses. Other moves has similar results. So, if you are going to place a lot faith in a multi-core engine analysis, make sure that you run more than one and average them. And, if you are going to do that, it would probably be faster to simply run the analyses using a single thread.
So, good luck with your multi-threaded engine analyses. And use the results wisely.
|Apr-16-17|| ||AylerKupp: <<zanzibar> I don't think that's the same "horizon effect" as used in reference to engines.>|
I believe that there are two definitions (at least) of the horizon effect. I describe them here: Stockfish (Computer) (kibitz #115) and shows an example. Hopefully it clarifies my opinion. Let me know what you think.
|Apr-16-17|| ||AylerKupp: <<Sally Simpson> In a Human v Computer game if the computer was losing and the best line from a lost position was a swindle attempt then the computer would go for it.|
I think that we've discussed this before. I don't think that the computer would go for it because it is not likely to be programmed to do so. The reason is what you said, this tactic would not likely work against other computers and computers play humans so infrequently that I didn't think that it would be worthwhile for the computer's programmers to spend the time and effort to incorporate this capability. But, who knows?, some of the engines might have it.
And that's a good example from the past. I broadens the concept of computers getting "lucky". But I think that this would only happen if computers were playing humans, not other computers.
|Apr-16-17|| ||Sally Simpson: Thanks again AylerKupp for your time and your in depth replies.|
|Apr-16-17|| ||morfishine: <Sally Simpson> Very nice post on Levy vs Chess 4.7 1978|
|Jun-17-17|| ||tonsillolith: One of the best examples of chess luck I read somewhere on these forums was this: a tactical skirmish between you and your opponent has just come to a close, his turn, and -- oops! -- you didn't realize your bishop would end up under attack by his knight, but -- thank heavens! -- your own knight protects it.|
You certainly didn't intend it, nor are you displeased. But are you tempted to credit your brilliant subconscious?
I know this phenomenon has saved my haunches in many a blitz game, and probably saved many of my opponents as well, but that's not for me to know.
There have also been several times am I guarded against back rank mate because of a bishop obliquely eyeing the single entry square from afar. And I hadn't considered that until my life was on the line.
|Jun-21-17|| ||jerseybob: <madlydeeply> Geeking out over the strength of opposition in 1963 when Fischer won the US championship 11-0. Hmm. According to chessmetrics...Lombardy #34...1963 US championship wasn't exactly completely Fischer v. milquetoasts.> Agree with the point, but Lombardy wasn't in the '63/64 tourney.|
|Jun-21-17|| ||HeMateMe: WOTTA lineup! we have three top ten players in our championship. when is the last time that occurred? Back in the 40s-50s with Fine, Reshevsky and Larry Mel? I think only the USSR has had such a deep national championship.|
Go Wesley! [But not his irritating CG.com fans.]
< Earlier Kibitzing · PAGE 27 OF 27 ·
Spot an error? Please suggest your correction and help us eliminate database mistakes!
NOTE: You need to pick a username and password to post a reply.
Getting your account takes less than a minute, totally anonymous,
and 100% free--plus, it
entitles you to features otherwise unavailable.
Pick your username now and join the chessgames community!
If you already have an account, you should
Please observe our posting guidelines:
- No obscene, racist, sexist, or profane language.
- No spamming, advertising, or duplicating posts.
- No personal attacks against other members.
- Nothing in violation of United States law.
- No posting personal information of members.
See something that violates our rules? Blow the whistle and inform an administrator.
NOTE: Keep all discussion on the topic of this page.
This forum is for this specific tournament and nothing else. If you want to discuss chess in general, or
this site, you might try the Kibitzer's Café.
posted by Chessgames members do not necessarily represent the views of Chessgames.com, its employees, or sponsors.|
your profile |
Premium Membership |
Kibitzer's Café |
Biographer's Bistro |
new kibitzing |
Tournament Index |
Player Directory |
World Chess Championships |
Opening Explorer |
Guess the Move |
Game Collections |
ChessBookie Game |
Chessgames Challenge |
privacy notice |
Copyright 2001-2017, Chessgames Services LLC