Cheating! Or not!

Posté par northlecale le 30 mars 2018

Heh! That’s not a nice topic to blog about Mister. Yes, you’re right on that, but it is a topical one. There is an accusation that some cheating has gone on in the PGETC. An observer from one team, who was using Leela to analyze all the games, noticed that one guy in the other team was always playing Leela’s move. It is one of the inevitable things about having really good software that you can easily use for live game analysis, and indeed for live cheating.

  1. Take the new Lizzie program, still in 0.x development stage but very powerful given that you can live sync to LeelaZero.
  2. Then you can just as well use classic Leela 11
  3. Or indeed hook up any other AI to Sabaki, GoGUI, or whatever.

In short, if you fancy cheating, you’ve got everything going for you.

In Chess a guy called Ken Regan came up with a powerful detection model. It’s not a very easy task to prove that somebody cheated. You have to show (obviously) that somebody is playing at a different level to what they normally would. You cannot simply say that so-and-so played 5 moves in a row like Zen. Ken specifically warns us against the belief that a move that is given a clear standout evaluation by a program is much more likely to be found by a strong human playerSEE HERE . Where he talks about how certain forced games can have as their characteristic a higher level of AI choices.

Using the GoR calculator on the EGD website, I did find that the player in question had an increase of about 50 points (half a rank) if it were a level A event – which is half a stone. Looking a bit further, I found that if I moved his rating up to 2600 (+ 2 ranks), I got roughly that his performance was as it was expected to be. That might look like a good piece of evidence. Well actually it alone doesn’t prove anything other than that he played well on the internet for that PGETC season.

In his defense you have to note two things:

  1. The player is clearly improving in over the board play.
  2. They openly admit to studying with Leela (Classic). Not just studying with, extensively studying with for 2 years, with witnesses to back them up.

Whilst we wait for the statistical analysis to play out, I repeat what I’ve said earlier on here. If you want to spend time studying Go with a computer, then there’s never been a better time to do it.

Publié dans Baduk, gossip | Pas de Commentaire »

TD Tools

Posté par northlecale le 19 février 2018

If helping out with the preparations for the Confucius Cup (Irish Chess and Go Congress) has given me anything, it has given me an interest in tournament draw utilities.  Well actually that’s something of a lie. I’ve always had a vague interest in that direction. I have a bit of a train spotty love of reading over old tournament result tables to see who finished in what position, and who beat who. It started with Chess tournaments, but it has continued into Go tournaments as well. Should I be seeking medical help over this? There are some who might consider it necessary, but I tend towards regarding it as a harmless little obsession that has the beneficial side effect of producing good documentation of the tournament activities of the various Go organisations I might try to assist.

For that end, I put together this little tool h92html, which although it isn’t very clever, does let me quickly put an end to missing tournament results tables on the IGA website. There are increasingly a lot of them. After running registrations for the Open and the Rapid through GoDraw and OpenGotha, I’ve also realised that there are other utilities that can be useful. Wouldn’t it be nice to have a xml2gdi to dispense with the need for that, and indeed to run the same draw in both. Hmm, perhaps not very nice, but it can still be a funky little utility if you don’t want to deal with GoDraw having bad to super-bad support for non UK players. Beyond that I’ve also got the mild itch to reboot Ire, luckily I am too busy.

Publié dans Go | Pas de Commentaire »

Coin Flips

Posté par northlecale le 27 décembre 2017

Chess scandals are normally sidelined into the mundane digital media of the twitlogscape, but the last few weeks have seen one hit the broadsheets. What happened exactly? FIDE, that well known bastion of impropriety and honour, awarded a World Championship event to Saudi Arabia which eventually rejected any participation from Israel and grudgingly admitted Qatar and Iran. This is totally against the statuates of FIDE which all its nations are also supposed to be abiding by. Protests are naturally thin on the ground, despite the scandal the flipside is the hefty coinage. Money talks, and here it probably said something about there not being a problem for this and that, but it forgot to mention Israel. So the end result is pretty terrible. How can you have a World Championship when you actively exclude countries which are entitled to compete in that championship? Is the perpetual religious strife of the Middle East really an excuse to allow this to happen? Okay, Chess isn’t exactly a theatre for morality, but the idea that some members can be forgotten about whenever it is (in)convenient is a bit much. What about Go? Well in Go you have the routine exclusion of North Korea on a pretty regular basis for its ongoing diplomatic situation. Visas are refused. I suppose the ongoing state of near war and lack of diplomatic contact is somewhat similar. Does it mean that the IGF is any better than FIDE? Well I think for 2 reasons that it probably does. Firstly, there is usually little choice as to where events are actually going to be hosted – Japan / China / South Korea are always going to be the primary source of sponsor money. Furthermore, from my memory, the IGF issues an apology each time this happens. Indeed, I was actually quite taken aback that the Japanese chair expressed some displeasure about his government’s decision. I wonder what FIDE will say? In any case, it only illustrates once again the complete folly of man, particularly when poisoned by quasi religious beliefs.

Anyway, going back to coins. The highlight of the year for me was probably the DeepMind feat of learning from zero. Go, Chess and Shogi were totally taken apart with no prior knowledge, except the various techniques which they effectively suggest in the neural network layers – ladders is an example I believe. I am particularly interested in the progress of the Leela version of AlphaZero – http://zero.sjeng.org/ – which is running away at the minute. I believe the initial mode at the minute is trial run, but regardless, today it is roughly dan level, despite what looks like immense lapses of reason when you observe it playing on KGS. It’s kind of mind blowing just how successful this project can be, even it does take a whole shit load of processing power and distributed resources to train it to that level. (Other reproductions exist – Viking and Pachi are also having Zero versions) I think it’s going to make an interesting new opponent/analyst to add to my current choices of https://github.com/ymgaq/AQ and Leela 11.

All of which brings me back to that classic but lazy little excuse for making a blog post, the New Year’s Resolutions! What is my plan for 2018? Well, play less and study more is one thing. I have been falling too often into my bad habit of playing to pass the time – which is now getting more and more uncomfortable, since it’s in a darkened room waiting for a child to go to sleep, and with a child protection taped over my keyboard. So, 10 second blitz has to say goodbye, slightly longer time limits for me. Analysis of games is an absolute must, but how to actually put that into practice I am not sure. The ICC-2018 is my little tournament challenge for 2018, try to play something vaguely interesting enough that my brain has to depart to pastures new. Real life tournaments – well that depends on babysitting, but I’m hopeful for at least some kind of visit to Dublin and the Confucius Cup. Go Tourism, rather than Go Playing is more my thing right now. Translation projects seem out of the question right now, nothing else useful in Romanian or French seems to be lying around to try. Maybe some other kind of writing might be in order? Anyway, 2018 here I come.

Publié dans gossip, Weiqi | Pas de Commentaire »

Competitive Drive

Posté par northlecale le 1 décembre 2017

The competition is perhaps the focal point through which many Go players meet and indeed exist, I shall contend that it is even more important as a driver than the club. We meet, we play, we fight, we talk, we live, we hunger. To keep us all active, there are, or should be, many layers of competition. We don’t want the same people to win all the time. Therefore a smart organisation organises a wide range of events – team, youth, club, state, province, senior, and indeed women’s. I used to argue with a friend that there belief that women’s only tournaments were sexist, whilst pair-go was to be extolled was simplistic and probably wrong. They are now playing in the IAPGC. The first Sotetsu (?) cup gave out hairdryers, cooking utensils, hoovers, makeup kit … that kind of thing, to the women taking part. This did raise a nerve for some. Yet, the competitions don’t have to be that way inclined. Pampering the community is a good thing, but it does rather ought to be done so diplomatically? Having experienced the IAPGC first hand, it did raise my anti-establishment haggles somewhat. A smartly dressed hetrosexual world free from all forms of non conformist behaviour, an iron fisted schedule with some crazy fancy dress urges hurled in, a Dutch theme park extroadinaire; I could entertain all these images up to a point. Yet the whole event is simply to promote Go, or rather Pair Go, which is a vehicle for women to play Go. Creating an event were countries are pampered, women are pampered, weak players are pampered, it is strategic genius after all. So I do feel it is an event which deserves praise rather than poo-pooing. Raised on the international stage we feel special. Women’s events are to me not at all disimilar, so long as there is the proviso that, or the understanding that, they do not exist because women are not able to compete with men, in the same way that junior events do not exist because juniors are not able to compete with men and women. Having been to both, I think they are enjoyed equally without degrading anyone involved in them.

Don’t throw the keys out of the pram.

Don’t throw the keys out of the car.

Publié dans gossip | Pas de Commentaire »

A dilemna of bears

Posté par northlecale le 4 octobre 2017

Bears are supposed to be grouchy creatures. Scientists put that down to all that honey giving them a diabetic profile, but the lack of wifi in the woods is probably equally frustrating. Pandas are supposed to be grumpier than your normal brown bear, and after being disconnected  from IGS 10 times in the space of a couple of hours, I really feel for them. The problems of these creatures are just not fully appreciated. Let’s stop with the Alexei Sayle influenced monologue for a moment though – what I’m trying to say is that the PGETC has, and probably always will, managed to annoy me in some way. Out of pride, pleasure, or ambition – whichever – I want to play for Ireland. Yet there I am in an arena that is about as attractive as the smell at the back of Paris Go Club. Disconnections, lag, people not turning up on time, people not turning up, a client from hell, glued to the screen for 3 hours to eat up everything else you could have done with your evening – it’s a combination of frustrations for me. There isn’t even a reward for a podium place in the minor divisions, you’re just an afterthought, as are the moves played perhaps. Anyway, Ireland’s campaign kicked off again last night with a draw to Kazakhstan and I am sitting on the bench watching, because I just don’t have time to go back in there. Maybe I’m better off that way. The scoreline suggests so. 2-2. A draw is not a loss.

Publié dans Baduk, Go, gossip, Weiqi | Pas de Commentaire »

Can KGS come back to life?

Posté par northlecale le 4 mai 2017

For years now KGS has been suffering from a dripping death in popularity. New players generally prefer OGS because it is easy to set up and use. KGS lost players with each new update of Java because they couldn’t work out how to connect to the server any longer. The top echelon of players departed to new servers. Even Go Federations started to switch over to using OGS – look at France, Sweden, UK, hey – even Ireland. Now though, it looks like it will finally have an HTML client. I don’t remember when wms started working on one, or when I realised that he would never finish one. New clients are coming. GoUniverse or ShinKGS are 2 decent examples right now. Can it bring the players back – perhaps not, but it will at least even out the share of beginners who choose between KGS/OGS to learn to play on.

Publié dans Baduk, Go, gossip, Weiqi | Pas de Commentaire »

Death of Brian Gallagher

Posté par northlecale le 16 février 2017

I learnt tonight of the death of Brian Gallagher. He had actually passed away some time ago, right at the end of 2016. A notice was already on the Irish Chess Union website. It seems trite to say it, but it certainly wasn’t the nicest thing to discover. He was only 17 years older than myself, and one of the strong kyu players I had trouble battling against on my visits to Dublin back in the day. Brian managed rather well the Collegians Go Club website back in the days of Geocities, and was twice our representative at the World Amateur Go Championships. Technically he lacked a certain amount of information, for he was want not to learn joseki, but to delve by himself into the game. He was still sharp though, picking up 3rd place in the Top-8. Although detached from Go in recent years (effectively he quit in 2009), he still kept on with Chess. A nice fellow, who left us too soon.

Publié dans Go | Pas de Commentaire »

Death of Tony Goddard

Posté par northlecale le 1 février 2017

I learnt of the death of Tony Goddard through a posting to the BGA’s GoTalk list. The email was from Jon Diamond, an old sparring partner of Tony’s. They contested the British Championship together a number of times. A few days after reading the email I am still sad to learn of his passing. Not knowing him very well, I am in no place to write a decent obituary, but I would note a few things about him. As well as the results listed here on Senseis Library he won the Irish Open in 1994 and a mini-tournament in Belfast in 1992. He was living in Belfast around that time, working at QUB by all accounts, and made a small club there. Paul Donnelly learnt a lot from him. They were two ghosts that I chased after for some time. I was pleased to have been able to meet and talk with Tony at the London Open many years ago.

From his own website, I would recommend reading this, this, this, and this. Reading those will take you a while, but its worth it if you are, or were, interested in Tony.

Publié dans Weiqi | Pas de Commentaire »

Resolutions

Posté par northlecale le 2 janvier 2017

As if a birthday wasn’t enough to remind you that you are getting older, the arrival of the new year serves to shout the fact loudly at you through a hangover. Time then to celebrate that there is little to do today, and to pretend to wisely invest my time in recording what I am likely to be doing over the next year. Naturally we leave out unrelated nonsense like developing a taste in contemporary Irish history, or trying to speak French without sounding like a complete cretin. No, here I reveal, almost exclusively, what I intend to do over the next year in regards to that game we love to call Go.

  1. For a bit of fun I started an exposition on a different kind of ;San Ren Sei
  2. Then there is the joy of satire in EuroGoHebdo
  3. Continuing to play Go, and not only through ICC but maybe even a real tournament(?)
  4. Continuing proofreading for the EGF newsfeed and translating for the RFG/AGA-EJournal

I think that’s more than enough to be getting on with, don’t you?

Publié dans Baduk, Go, Weiqi | Pas de Commentaire »

Breaking Tiebreakers

Posté par northlecale le 14 novembre 2016

Tiebreakers are not as unimportant as some people might wish them to be. Unless you’re using something like GoDraw, your pairing will be influenced by the tiebreakers which have been chosen by the tournament director. In other words, it is not just at the close of the event that the tiebreakers come into play. The EGF has a page which describes its current policy on tiebreakers - http://eurogofed.org/egf/toursysrules.htm - these date from 2007. This page is not as clear as it might be, and at times is offering some bad advice, but it is relatively useful all the same.

In a recent competition the following criteria were used

Swiss Score, Direct Comparison, SOS-2 – SOS-1, and SOS.

This is an unusual choice, but we can see where it came from. On the EGF’s page we have

If only one tiebreaker is used, then these tiebreakers are recommended:

  • Number of Board Wins.
  • Direct Comparison.
  • SOS-2.
  • SOS-1.
  • SOS.
  • Rating.
  • Previous Order.
  • Lottery.

What the organiser missed is

Recommended Order of Equal Players or Tiebreakers

  • This general order of priority is recommended:
  • Playing more rounds.
  • Playing playoff rounds, possibly with short thinking times.
  • Having equal players.
  • Using tiebreakers.
  • If only one tiebreaker is used, then the recommended order of priority is given by the list of tiebreakers that may be used.
  • Only one of SOS-2, SOS-1, or SOS may be used.
  • If more than one tiebreaker is used in a relative order of priority, then this is given by the relative order in the list of tiebreakers that may be used.

What he might have been further mislead by

Recommended Usage of Tiebreakers

  • Number of Board Wins: It can be applied only in a team tournament. There it is highly meaningful and should be the first tiebreaker.
  • Direct Comparison: Provided it can be applied at all, it is very meaningful because it might be interpreted as an already performed knockout playoff among the tied players. So, for the final results, generally it should be the first or even the only tiebreaker.
  • SOS-2, SOS-1, SOS: They should be used only in McMahon or Swiss. They express a mixture of opponents’ strength, statistical noise, and pairing luck. Their apparent numerical precision is greater than their true significance. Therefore they must be used with care. Application for the final results is doubtful while application for making pairings is reasonable. SOS-2 filters more noise than SOS-1 than SOS; more noise can be filtered in more rounds more easily.

Direct Comparison is a total devil if we try to use it to break ties between more than 1 player, in which case it is not at all a play-off game but an imagination which has run wild. SOS is much better suited to Swiss and McMahon pairings as it is considering the actual tournament performance, and not just extrapolating from 1 game.

An advice he was not given is that Direct Comparison should not be used in pairings, it should only be used at the end of the event in tiebreaking. It should never be used in making the pairings. It was then all rather unfortunate that this system was chosen. We have to consider why it was chosen.

  1. It was chosen because the TD studied the 2007 page and tried to apply it.
  2. It was chosen because the EGF does not validate or rubber stamp championship tournament systems
  3. It was chosen because nobody was brusque enough, or competent enough, to dispute it

Publié dans Baduk, Go, Weiqi | Pas de Commentaire »

12
 

Dylansiloret44 |
Mistymil |
Godemon Games |
Unblog.fr | Créer un blog | Annuaire | Signaler un abus | Freshbad
| Pokémon Rush
| Mechanicat