Russell Sherwood

Choosing a Chess Engine

Russell Sherwood  Monday, March 5, 2018

So what should you look for in choosing (a) Chess Engine(s)? Much depends on what you want to do with it but a few factors to consider are:

Cost – How much are you willing to pay? If the answer is nothing then there are still plenty of choices!

Advanced Parameters – Does the engine have features useful to Correspondence Chess, such as LMR, Wide Search…….

Saveable Hash File – Are you able to save the Hash file to be able to resume later?

Learning – Does the engine has a learning function implemented?

Speed – Is the engine fast enough for your analysis style? (Whilst some engines appear to be going deeper than others, in general, this means that the engine is being much more selective in the lines it is examining, which does mean it goes deeper but could also mean that it misses something!)

Frequency – How often is the engine updated? Some engines are only updated on almost a yearly basis, whilst others are updated almost weekly. How much difference this has to the strength of the engine is debatable but psychologically this can be an issue for some.

Family – Is the engine part of the same family as the other engines you use? If using a multiple engine strategy then different families of engines are a must!

Maximum Cores/Hash – Is your hardware and the engine compatible so that you will get the most out of it?

Correct Version – is there a version for your hardware or better still one you can compile yourself?

Tactical Mode – Does the engine have one?

Tablebases – Which (if any) tablebases will the engine interface with?

As easy as XY but not Z

Russell Sherwood  Friday, March 2, 2018

The majority of modern CC players realise after a while that simply allowing a Chess Engine to analyse to depth is not the most effective strategy. There are numerous reasons for this, all of which boil down to the fact that all of the mainstream engines are not designed for the needs of CC players. To be competitive they adopt a number of strategies to find a good (great?) move quickly rather than trying to find the absolute best move.

There are many “workarounds” to this but three of the most basic are:

Next Move (Y)

Here we encourage the Engine to look at the next best move (and then the next best and so on). The purpose of this is to look at a number of moves which the Engine may consider to be inferior.

Pass (Ctrl Alt 0 aka Null Move)

A fairly simple idea – “pass – Don’t make a move”. The purpose of this option is to determine the threats your opponent has.  Under normal circumstances t, e evaluation will swing massively, especially if you are playing black and any threats in the position become obvious but what is more interesting is that if the evaluation does not move much. These positions are ones where the premium on planning and creative play cannot be underestimated as we are no longer in a position where we are simply responding to the opponent’s moves.

 

Show Threat (X)

Here we show the threat in the position if our opponent did not have a move.

 

These three basic ideas allow the bedrock of starting to dissect a position and if you don’t know how there are implemented in your GUI perhaps its time to find out!

Who's been doing what, with who?

Russell Sherwood  Sunday, February 18, 2018

I'm in the process of Finalising places for WCCF 6 and 7, plus a number of other Invitationationals.

The attached file shows all the Norm events we have had in the past, who has been placed where and so on.

It does not show other events such as the Tri-Nations which do not offer Norms!

Anyway, it may be of interest to some players, although I would suggest caution in drawing any conclusions - often players were the only one qualified by rating, gender.......or we may have had different qualification requirements at the time!

 

Download

SICALM #1

Russell Sherwood  Saturday, February 17, 2018

Here we start a new monthly article – Stuff I came across last month!

Like many people these days I spend too much time wandering the internet. However, rather than watching funny cat videos I tend to follow down information paths related to my interests. The purpose of this column will be to share a few of these! If you have any interesting little side places let me know so we can enlighten all!  You will notice I don’t give a full review – my aim is to encourage you to go and have a look yourself!

Decode Chess  https://decodechess.com/

This is a new site and very much in beta stage. In a nutshell, you enter a position, its analysed for about 90 seconds and then various insights to the position are given. It’s early days yet as the analysis suffers from a few bugs (reports from a few CC players suggest missed wins) but even in its current condition, it can help a player understand the position better. One warning is that that the analysis is based on Chess engine software (it would have to be unless they had a GM chained up next to a terminal……..) and so its view of some positons (King’s Indian Defence for example) is suspect.  At the moment the site and its analysis need to be taken with a pinch of salt but if they deal with the bugs it could become a new vital tool for the ambitious player.

Tartajubow on Chess II  http://tartajubow.blogspot.co.uk/​

This is a treasure trove of interesting articles on a number of subjects – ranging from Chess History to Chess Engines! As an added bonus you can subscribe to the regular email for a morsel of chess goodness.  Do not be put off by the sites rather dated formatting – the content is excellent!

Chess Improver  http://chessimprover.com/

This is a fascinating site with a number of articles on improvement on a wide range of Chess subjects – written by numerous players of a wide range of strengths. SIM John Rhodes provides an input to British and English Correspondence Chess (Something missing here maybe!)

Going Postal!

Russell Sherwood  Saturday, February 10, 2018

Until fairly recently (the mid-1990’s) the only option for Correspondence Chess was via the Post. Various matches had been played utilising radio, Telephone and Telegraph services but these all tended to be for international team matches rather than events for the individual player.

 

Correspondence Chess then morphed at a fairly rapid rate in the last twenty years – first through email, then onto the Server chess and with Mobile device based chess being at the cutting edge.  Unlike email based CC, which is more or less defunct (A tiny number of events exist) Postal Chess still maintains a small dedicated following.

 

Why is this?

 

There are a few obvious reasons – the most important one, which is often overlooked, is that Postal does not require a computer of any sort to play, which franchises a group of players to take part who could not otherwise. Also, these players’ value receiving and sending physical moves – often accompanied by, sometimes extensive, personal notes. This personal touch should not be underestimated in terms of the attraction to the adherents.

 

So what are the downsides? The vagaries of the postal service are a major one and the potential for lost moves. These in many ways are small compared to the potential cost issue. Let’s imagine I am playing a game against an opponent and the game goes on for 50 moves. So if I send each move one at a time this would be a cost of £32.50 for the game. Let’s scale this up to six opponents with all the games being doubleheaders, our total has risen to £195!  Thinking positively let assume we have a lot of conditional moves and reduce this by 50%, giving us a total of £97.50 for an event!

 

This is for a UK based event, an International one would be significantly more.  Now, this has to be compared to the cost of playing a server event (Entry fees can be ignored as they would be the same for both). This becomes a little tricky. If you already have an electronic device and data (Be it a Phone, Laptop or Tablet and Mobile data/Broadband/Fibre) then the cost can be considered zero (as they are sunk costs to use a financial term) but if you did not then there would be a bare set up cost of around £100 and a minimum of £10 a month – so potentially another £120 a year.

 

So as a summary of sorts – if you are not on the net at all, the postal chess is probably a slightly cheaper option but if you are connected it is not.  Why all this discussion on Postal Chess? I was musing about entering the BCCA Postal Champs next season – to attempt to add it to my Trophy cabinet. Having gone through the costs (which amount to the entry fees for 2 Grandmaster Norm events) I think perhaps it’s not worth the cost of adding to my collection after all!

You have been weighed, you have been measured and you have absolutely found wanting!

Russell Sherwood  Saturday, February 10, 2018

Over Christmas, I was watching one of my most-loved films – A Knight’s Tale. Within the (admittedly lightweight film) there is one very interesting concept – that of overcoming any natural disadvantages. This is shown in the film with two concepts – “Change the Stars” and “How a man like you, hope to beat a man like me?”

https://www.youtube.com/watch?v=tdhQWkTl1PQ

Now, what has this got to do with Correspondence Chess or Chess at all? The simple answer is psychology!

 

How do you approach each opponent – thinking you have no chance? Or that you will smash them?  Both of these are faulty ways of thinking for a large number of reasons including:

 

  • The nature of the Elo system is that you always have a chance, be it ever so slight of beating or losing to an opponent.
  • All rating systems are flawed as a tool to predict an outcome, based on the simple premise that they are an indication of how the player has performed in the past, rather than how they will perform in their next game.
  • Player’s ratings are on the move – so if two meet and their published ratings are 200 elo different but their newest rating will only be 50, which gap is correct?
  • One player may be up for a certain game, whilst the other may not be.

 

The very valuable lesson to learn from this is to think that you could win every game you play (Could not will!!) and the only thing that matters is this game – not ratings, titles or past history.  This may seem romantic If your opponent is 200 or 300 elo higher than you.

 

My riposte to this is that I have opponents who, a few years ago, I would have considered hard to beat, who now I look at the people I would target for a win in a tournament. Have my skills improved over that period, yes but more importantly so has my mindset – I now respect ALL opposition but fear none and this has shown in my upward curve of results and could do for you.

 

So in that time I have “Changed the Stars” and for any opponent who might consider “How could you hope to beat me?” I have the message – I know I can beat you, maybe not in this game but at some point!

 

This brings us full circle to the title of this article – “You have been weighed, you have been measured and you have absolutely been found wanting”. Is it not the time for you to consider how to unseat you next highly rated opponent? (Just as long as it's not me!)

 

https://www.youtube.com/watch?v=tdhQWkTl1PQ

 

Hold the Front Page

Russell Sherwood  Sunday, February 4, 2018

Dear Players and Readers,

 

For 2018 we have an exciting range of articles planned for your perusal including

Annotated Games – Including the Sensible to the Silly, The Entertaining to the Instructional

Player Interviews – Ranging from the Top Players to the most committed of players

Book Reviews – Looking for something for a new read?

SICALM (Stuff I came across last month) - - Potentially interested Chess related stuff

News from the Front – Regular round-ups of performance and progress in events

Norm Update – Who has gained what and Where?

Event Boosters – What’s coming up and Where?

Congress Update – The 2018 Congress is taking place in Llandudno this year – get updates on what’s happening

Articles – Ongoing articles on many facets of Correspondence Chess

From the Forge – What’s hot and What’s not in CC

Jaeger – Development news on the CC Specific Engine – Jaeger

WBook – Need an Opening book for CC – look no further

Software Guides – Want to know how to use Chess software to its full capability – look here!

Welsh, British and ICCF News – What’s happening in CC?

All together this comes close to a post most days so call back often!

 

 

In contempt

Russell Sherwood  Thursday, February 1, 2018

Here we are not talking about annoying people but the concept of considering the gap between you and the opposition in terms of playing strength. This has been coined “Contempt” in the engine world and is worthy of consideration for the aspiring player.

 

So what is contempt in engine terms? In we consider that, in very simple terms, an engine works by finding what it thinks is the best move and then the best response to this move and then the best response to this response and so on, creating a mainline.  This makes the simple assumption that the opponent will play the best possible response. Now if we are playing someone who is demonstrably a weaker or stronger player this will probably not be the case.

 

What was being found was that top engines were overestimating the strength of response (giving them too much “respect”) so the concept of “Contempt” was born. Different engines implement Contempt in different ways but all implementations use a modifier. The main practical benefit of this is to change the path the engine may take when it is searching for the best line. In general, it will mean avoiding exchanges and simplifications unless there is a clear advantage. The opposite would be true if a negative contempt were used (Respect would be a better name)!

 

Traditionally this function has only been available for Engine v Engine matches – but versions of Stockfish and other engines now exist where Contempt can be adjusted in Analysis.

 

Many players have held the view that Contempt should be left alone but it is an area worthy of experimentation. For example for a higher rated playing in an early round Open event, it is worth examination to see if alternative lines exist.

 

So in summary – this is a fairly unexplored area and one which might generate interesting results for pioneers!

7th Chess 960 World Cup

Russell Sherwood  Monday, January 29, 2018

 

The International Correspondence Chess Federation

announcement

7th Chess 960 World Cup

 

The tournament will be played by server and will be organized in three or four stages, depending from the number of the entries. The sections of the preliminary round will have 6 players (10 games) with 2 qualified for the next round. All sections will be with double round robin (players will have the same position with Black and White but all positions will defer from one opponent to another). Time control will be 10 moves for 30 days. The event will be rated for Chess 960 rating system.

Entries

Players may enter in the usual way through their National Federations or, where eligible, via the ICCF Direct Entry system.

·  All the finalists of the 4th Chess 960 World Cup can start the 7th Chess 960 World Cup from the Semi-Finals (starting approx. in the first half of 2019), entering via National Federation up to 30.11.2018.

Entries via National Federations should contain player’s name, e-mail address and ICCF ID and should be forwarded by e-mail to the Tournament Organizer, Leonardo Madonia, (e-mail: [email protected]) to reach him before 28th February 2018. It would be appreciated if Federations send entries “as received” and not delay them until closing date.

A player can apply for more than one preliminary group.

Allocation of players to their sections will be made as random as possible; however, geographical distribution as well as an achievement of reasonable equality of the average rating will be taken into account.

 

Tournament Organization

The tournament will start on 15.03.2018.

Registration will close on 28.02.2018 to allow for pairings.

 

Prizes

Prize 1000 € for the first 3 players (500 €; 300 €, 200 €), no tie break.

 

Qualifications

The winner and the runner-up of each preliminary section will qualify for the next round.

Although the number of preliminary groups which each player may enter is unlimited, no player will qualify for more than two semifinal groups or more than one place in the final (in the case of four stages the qualification to the second round is unlimited).

In case the total number of players is not a multiple of 6, the players with the highest ELO (up to a maximum of 5, no multiple entries) will be moved directly to the next round.

 

· The winner of the 7th Chess 960 World Cup will qualify for a World Championship Candidates (tie break applies);

 

Member Federations are asked to give the tournament wide publicity to all their players.

We wish all participants many interesting games and new connections with CC friends in other countries, based on the spirit of friendship and the ICCF motto AMICI SUMUS.

 

The Blueprint

Russell Sherwood  Friday, January 19, 2018

So you want to progress to the pursuit of higher titles? There a number of areas to consider (and action) if you want to make progress. Most of these I have covered in some depth in previous articles but bringing these together is of some interest.

 

Access

You need to gain access to events which will generate the necessary norms.  It is fairly simple to consider which events will give IM norms (Category 1 upwards) but this leads us to  two sub questions

 

                Availability

 

                Once you reach 2300  up to Category 3-4 become openly available, Category 7 once 2400 is reached. Getting into a higher Category event is an aim of many players but this is not often available.

 

                Relative Difficulty

 

Not All Norm events are born equal, even those of the same Category. This may seem odd but the nature of an event, the nationality of the players and the prestige of the event can all affect the level of difficulty of achieving a norm. In addition, even and oddly numbered categories tend to be slightly different in strength.  Here one area to consider is the spread of ratings of the players as the average of the ratings drives Norms but spread can have a massive impact.

 

Norms vs Rating

 

This is a difficult pill for many players but once you are chasing Norms, ratings become secondary. You do need to avoid dropping any significant points but your aim is to secure the norm and this may mean taking negative elo-draws!

 

Winning

At lower rating levels, progress can be made by simply waiting for your opponent to make a mistake and blindly following an engine’s output. Once around 2200 is reached this is no longer the case and if you want to secure Higher norms, you need to win games. I covered this in “USP” but there are many ways to try and win games and you need to consider what your method(s) will be

 

Gameplan

At the start of the event you need to consider your game plan – what do you need to achieve for the Norm – 3 wins from 12 games? Whatever is necessary you need to consider how you intend to achieve this? Aim to draw with Black or play to win? Target specific players? Identify other players winning style?  All of these go into determining your game plan.

 

Time Management

Most players will have a portfolio of games. Are these Norm games you’re most important? If they are then you should give them the most time and attention.

 

Work it

With all the clever ideas in the world, one thing that is necessary to win games at the higher level is hard work and a lot of analysis. One thing is for certain, if you are playing mainly the Engines first or second choice then you will probably not win games at the higher level. The reason for this is quite simple – your opponent can see you coming!

 

So go for it!

Welsh Correspondence Chess FederationBritish Correspondence Chess AssociationSchemingMind Internet Correspondence Chess ClubSocial Correspondence Chess AssociationNational Correspondence Chess ClubWelsh Chess UnionInternational Correspondence Chess Association