Russell Sherwood

To Draw or not to Draw?

Russell Sherwood  Friday, December 22, 2017

When should a player consider accepting/offering a draw in CC?  The answer to this question is very personal to the player, based on a number of beliefs, which we will now try and untangle!

The first area to consider is the player's approach to draws. Depending on the event player’s will adopt a number of different strategies, ranging from “Win with White, Draw with Black” – at the higher levels against peer level opposition to the frantic “Win at all costs with either colour” seen in some of the large ICCF Open events. (Interesting side note, often high rated players underperform in Open events as they struggle to move from the first strategy to the second!)

This strategy (or even the middle ground of taking it as it comes) will all introduce a level of Bias into our acceptance/rejection decisions.

So what does a draw offer actually entail? Effectively we are saying that “I don’t think I can win this game” and by implication “I don’t think you can either”. So how do we reach that conclusion? This varies between players but from discussions with a number of players a few (not mutually exclusive) approaches include

 

  • #1 Statistics for this position indicate it likely to be a draw – e.g. the last 100 games played from this line all ended in draws.
  • #2 My engines(s) evaluation of the position as 0
  • #3 My Chess sense tells me that the position is dead
  • #4 The Rating (or more precisely rating difference) of my opponent

 

We also have a number of other reasons for draw offers, not related to the position (or opponent)

 

  • #5 The position of the tournament – If a draw is only needed to win/qualify/…….
  • #6 Title Norm Requirements – I only need a draw
  • #7 Rating reasons – Either the player wants to “leech” rating points or wants to take the gain/or hit to their rating before the end of a rating period

 

Then to add to this we have psychological reasons

 

  • #8 They are  in an inferior position but want to cast doubt on your evaluation
  • #9 They want to “buy” their way of a bad position
  • #10 They are bluffing, having seen a lost position (and hoping you have not yet)

 

Mixing all these together we then see a range of behaviours ranging from: Offering a draw around move 20 or earlier – these tend to be related to players who rely on #1,#2 or #7 to grinding out a position to very end – they tend to be players who focus on #7 and on a factor not yet discussed – Fudge Factor: How likely is your opponent to make a mistake in a draw position.

 

So how should the intelligent player approach draw offers?

 

First, forget any nonsense about the higher-rated player having to offer the draw. Ratings are rather more fluid now than they were 20 or more years ago, so this just does not make sense!

 

In my opinion, the criteria should be considered:

 

  1. What are my goals for this event and this game? This should have been done at the start of the event but may have been modified as the game/event progressed. So what was the expected result at the start of the game (determined by your preparation) and does a draw meet the needs of this event?
  2. What is the state of play on the board (#3)? This is not what an engine or opening books tell you but what you actually think. The key here is, to be honest with yourself, don’t delude yourself that your position is better or worse than it really is. If we do come to the conclusion that our position is much worse, then a tactical draw offer may be on the cards
  3. What do the Engines say about this position? Remember that an evaluation of 0 does not mean it is drawn, simply that the engine considers the position to be balanced. A few helpful things to consider here (and in #2) are the imbalances in the position (based on Silman’s theory of imbalances) – what is different about black and white – pieces, mobility, pawn structure.
  4. What do the Databases say about the position?  There are a number of factors to consider – the performance level, the percentage of draws, how recent the games where.
  5. Am I content with a draw (and is my opponent likely to be content? This is the least analytical criterion to consider but, even with recent rules changes, draw offers are still limited.
  6. What type of Guy/Girl is my opponent? This information should have been gleaned in your preparation but how often do they win/lose? Do they take early/late draws? What is their recent form? (Always difficult to assess as games in the public domain tend to have been won/lost a year or more before!) An example of this was an opponent I kept bumping into Regional events. He would play like a 2400 rated player for about 30 moves, then 90% of the time blunder in the middlegame – so I knew against him the approach was to never take early draws.

 

Unless you can find support for a draw from most of the  6 of these criteria then a draw offer should not be made or accepted (#2 is the exception if all the others are green as Engines often have favourable scores in what are in reality drawn positions).  There is a 7th criterion which is related to your time – both in an individual game and for CC in general. It used to be the case that you could recover your time in almost any event – even if down to less than 1 day on the clock but with the event of Tri-Bi your relative clock positions becomes something you should consider on a regular basis.

 

 

Anyway, I hope these thoughts are helpful!

What’s your UWP?

Russell Sherwood  Thursday, December 21, 2017

Within Sales and Marketing (S&M) there is a concept of a Unique Sales Proposition (USP). This is a concept that basically says “This is why you should buy from me”. If you consider, say Volkswagen and Ferrari both have different propositions on why you should buy a vehicle from them. Anyway, what has this got to do with CC? Using similar thinking we can come up with the idea of the Unique Winning Proposition. Let’s simplify this more – How do you win?

 

If we look at almost any sport, different methods are utilised by different players and/or Teams – All out attack, “Park the Bus”, Counter Attack, the list is endless…….

 

You might think that I just play the moves and the win comes (or not!) If this is the case then you are probably not getting the results you deserve! So let us consider a few of the approaches I have observed:

 

#1 Out-Hardware your opponent. It’s sad but this is an approach used by some – basically, I have the bigger hardware, so can search deeper. It does have flaws, one of which is the concept of Minimum effective dose (or search in this case) and another is that of diminishing returns. Both of these can be summed up as once you get to a certain depth of search, going beyond it tends not to gain you much!

 

#2 Out-Preparing your opponent – this is probably one of the most effective methods. Examining the games of your opponent and finding weaknesses to exploit. On the surface, this is the Opening’s play but we can go much deeper, how do they handle certain types of positions? How do they win? How do they lose? Early Draws?  Here we can generate strategies such as knowing your opponent does much worse in closed openings – so guess where we go!

 

#3 Out-Preparing your opening – This is similar to the previous method but here we focus on the opening much more – do we actually know the concepts of the Opening, the plans, the ideas, Why this move is played/not played rather than simply looking at percentages in Opening books?

 

#4 Playing Chess – This approach involves very early deviation from the mainlines and playing based on our understanding of the game (with an engine acting as a tactical checker). The advantage of this approach, especially when combined with Out-preparing the opening is that we are able to ignore the engines protestations of a position being 0.2 pawns worse and play the long-term game.

 

#5 Middlegame Knowledge – Engines are tactical beasts but have fairly poor middlegame conceptual knowledge. If we specialise here we are able to see that we want our Knight in that outpost and thus generate a plan which will get it there. Engines will (almost certainly) not see this sort of manoeuvre. The same goes with the concept of translation (a core anti-engine technique) which is the idea of moving your entire position up the board.

 

#6 Endgame Knowledge – Engines are pretty poor in endgames and a good human player, (with an engine peeking over their shoulder for tactical blunders) can outmanoeuvre the engine

 

#7 Psychological methods – This is a bag of methods utilised (none of which I would recommend!) by some players – including attempting to upset their opponent (whilst saying just within the rules) in a number of ways. These methods tend not to work as much these days and most TD’s take a dim view of them if discovered.

 

#8 Communal Analysis – These are a number of methods utilised by some players which are in some cases, breaking the rules and in others are on the very edge and untested in terms of the legality. A couple of examples of this are (a) Team Rooms. In very simple terms the details and moves of the game are shared between the teammates in a private area, with analysis, comments and suggestions being discussed. This is against the rules and risks severe sanction if discovered (b) Shared Hash Files.  Imagine a number of players decide to play a certain opening and all have very large Ram on their PC’s. If one has a very fast PC, they can let this run for a period of time and fill the hash table. This hash table is then saved and shared between the different players. A similar idea is running Monte Carlo or Gauntlet analysis on their opening positions and sharing the results.  In my mind, this is on the edge of the rules, although probably legal as the sharing of opening books is legal

 

#9 Out Software your opponents – Here the basic idea is to gain a winning edge by always having the cutting-edge software. This is generally the latest version of engines, databases or GUI’s. This is, in general, a bit of a fool’s errand as the difference between, for example, Komodo 11 and Komodo 11.2.2 is not going to make that much difference to your results. However, it is worth noting that saying reasonably up to date is a good idea.

 

 

So dear reader? How do you win your games? If you don’t already know then specialisation in one or more (can you specialise in more than one thing?) could possibly give you improved results and somewhere to focus your limited study time!

 

Lean Chess

Russell Sherwood  Thursday, December 21, 2017

I am not missing a letter A in the Title! This article came about after a work colleague read my article on Industrial Chess (He is not a Chess player but like myself a Change Management professional). How and why he read the article is another story but it did lead to an interesting discussion on the subject of process optimisation and one which led to me getting some friendly stick for not having applied Lean methods to my CC practice.

 

So what are Lean methods? These came out of Toyota (the Car manufacturers) approach production and are utilised in one form or another by most large organisations, both in Manufacturing and Services. In its most simple form, it is applied to process to standardise the time it takes and then eliminate waste within the steps.

 

Anyway over a drink, we outlined the steps of general CC practice (and my take on it) and some very interesting ideas came out of this. Putting these together meant, based on some basic calculations, around a 40% reduction in the time spent on each game, despite increasing the time spent analysing.

 

I won’t reveal more details on this until I have tried the approach out over the next few months but to give the interested reader a leg up. Consider the following:

 

Do you have a routine?

What are the stages of a CC Tournament and Game? (Think through a game, it's not just Opening, Middlegame and Endgame)

Do you approach the stages differently?

Do you approach each move the same way?

Are you consistent with this approach?

 

Exciting Times!

Chess on the Fringe

Russell Sherwood  Monday, December 18, 2017

Over the years a number of books have been published which purport to be general theories and methods.  These books are often great resources for the learner or ambitious student, even if, on occasion, some of the ideas in the books are not well understood until many years later (Nimzovitch) or date with the passage of time.  

These days these books to tend to be lambasted when published, for a number of reasons: sometimes the player is not that strong, sometimes the ideas seem just strange and often they are just plain wrong! However, there are a few of these around which do warrant a 2nd glance as there may be good ideas hidden within them. I would suggest, dear reader, if you are tempted to examine them in more depth, then take the claims made with a pinch, indeed an entire bag of salt.

A.R.B Chess System: 2/10

If you google this method you will find many you-tube videos of the author using the “system” to score many victories against Chess Engines. The method in itself is reminiscent of a number of Mike Basman’s ideas from the 1970 – The St Geroge, The Grob and the similar. Practically the methods described can work reasonably well against engines and some examination is worthwhile for this reason, however against either a strong human or human-engine combination the method’s merits are questionable at best.

The System – Hans Berliner 5/10

This is a different animal altogether. For younger readers, Hans Berliner was a CC World Champion and Top class player for many years and involved in the development of early chess engines. The System is an attempt to formalise these ideas into a workable thinking system. The problem is that many of the ideas whilst very interesting, don’t quite mesh together into a viable system. In addition, the book is written in a rather self-indulgent style. There are some excellent nuggets within the book – for example, the basic concept of if I have a choice of moves I want to play the one which keeps my options as wide as possible but limits my opponents – a very useful concept indeed!

Best Play: A new method for discovering the strongest move  - AlexanderShashin 6/10

The Sashin method is a combination of ideas which in summary says based on certain position characterises you should play in a specific manner. These methods are named after World Champions – Tal for example for highly attacking positions. Not well publicised this does have the potential to significantly improve a players results – simply through thinking about the game differently

The Secret of Chess Lyudmil Tsvetkov 6/10

This is the most modern of the four books here and is a child of the chess engine age. The author has spent a significant amount of time playing against and analysis the methods of Chess Engine Evaluation. His first book is mainly aimed at Chess Engine Authors and includes a large number of evaluation criteria and what he believes the values should be.  These evaluation criteria include a lot of hitherto unconsidered ideas, which could give benefits if exploited. The major problem with the book is the writing style – it has the feel of a maths textbook and is a very heavy read, with no real indication of how to put the ideas into practice. The authors 2nd and 3rd books are examples of his victories against Chess Engines. Whilst very interesting I believe a synthesis of the concepts would have helped the reader somewhat who has to rely on working through all the games.   What is the value of this body of work, personally I believe for the more advanced CC player this could be a useful book to work through but it is not a page-turner!

This concludes our short survey of some of the books/ ideas on the fringe of chess thinking.   I believe they are worthy of some attention, as, even if the reader rejects the ideas, at that point they are considering what they believe the “right” concept to be! I would suggest the reader does not pay much for them though!

 

Pushing the Limit!

Russell Sherwood  Sunday, December 17, 2017

Pushing the limit

 

Serious CC players are a little like Petrol-heads, these always want a bigger, more powerful machine. Often this is not possible for simple financial reasons! So how can the aspiring player improve their technical lot without spending (too) much money? For the purpose of this article, we will assume a laptop in use, so the processor cannot be upgraded.

 

#1 Current Version of Engine

                For any engine, the latest version will give the best results. For the Free engines, it is a no-brainer, for the commercial ones it is something of a tougher decision. In addition to this consider if there is a faster version of the same engine – for example ASMFish for Stockfish!

 

#2 Version of Engine for your processor

                Many engines these days come in a variety of versions. For example POPCNT, BMI……  If you have access to all of these, try them all note the results and this will determine which is the most suitable for your processor. Typically there can be around 10-20% difference between the fastest and slowest versions

 

#3 More Ram

                The amount of Ram utilised determines the HashTable, which is one major determinant in the effective deep searching of the position. Pound for Pound maxing out your RAM is probably the most effective way to improve performance.

 

#4 Tablebases

                If you don’t have them then this is a simple performance improvement. Whilst they probably won’t win you many games they will steer you away from draws and losses

 

#5 External SSD

                Following the use of Tablebases, an external SSD drive is an excellent way to improve performance. This much faster storage means that your Tablebase (and Opening book) access will be significantly faster

 

#6 Compile your own

                Some of the Freeware engines can be compiled on your own machine – all of the Stockfish family can be done so using the Automated software. When compiled this was the compiler is taking note of your hardware and the engine will run that 10% faster than an “off the shelf” compile from another person.

 

#7 Don’t run other software at the same time

                A no-brainer really. Ideally, you should not be running other software at the same time but many people still do. If you need to run a browser to enter your moves, look for one will a small “footprint” which does take too much away from the engine.

 

#8 Engine Settings

                Check your engine parameter file settings. Is your Ram increased to 50% of your total? Is Large Pages enabled? Is the number of threads correct?

 

#9 Use a Cooling Pad

                Multi-Core engine use creates a lot of heat, more than a typical laptop is designed to handle. The use of a cooling pad can help somewhat, both in cooling the laptop but making it more comfortable, especially if you do have it on your lap!

 

#10 Determine which GUI/Engine combination runs best on your machine

                As shown in past research, certain combinations of GUI and Engine show significant performance swings. Experimentation here can reap rich rewards.     

 

#11 Keep your OS up to date/Consider a Windows reinstall

                Apply the OS and BIOS updates that come along, they will keep your machine running. Also consider re-installing Windows once in a while BUT make sure you back up everything first!

 

#12 Ensure Large Pages are enabled

                We mentioned enabling Large Pages in the Engines parameter file but we need to ensure this is allowed on the machine. To do this Google “Enable Large Pages” and work from there! What are large pages you may ask? It's simply the way that windows stores things, this setting encourages Windows to keep items close together, which gives a small speed increase.

 

#13 Memory Boost

                I discovered this one by accident. Certain high-quality USB sticks can have a “Windows Boost” feature enabled. This allows effectively extra Ram to be utilised to help out the Operating system.

 

‘#14 Reboot

                Once in a while, reboot your machine. Why? Even the best Operating system in the world leaves bits of junk in memory and over time this clogs up the system. By rebooting this is cleaned out. How often? That depends on you but I tend to do it at least daily, more often depending on circumstances.

 

#15 Horses for Courses

                I noticed a strange phenomenon with AsmFish/CFish and Stockfish. Normally we would expect Asmfish > CFish > Stockfish in terms of speed but in Endgame analysis this is often not the case. Trial your own engines to determine the truth for you!

 

So what do you get from all of these – potentially something like 40-50% improved performance if you were badly set up beforehand. This won't mean you go 50% deeper but will mean you get to the same Depth 50% faster!

Industrial Chess

Russell Sherwood  Saturday, December 16, 2017

I was pleasantly surprised to see myself 4th on the list in terms of games played (>4500) on the new Chessbase Correspondence database. Whilst I don’t play anywhere near as many games as I used to in the past I still play greater than most. This does occasionally lead to the question of how I do/did play so many games at a time. My normal response is badly but unlike most of the players at the top of this list, I don’t have a rating in the sub 1000 region!

So some thoughts if we you do want to play a lot of games at the same time?

#1 Be clear on why you want to do it! As it can soak up a lot of your free time. When at my highest activity levels as I was in a professional management role which had me spending probably 3-4 nights in a hotel most weeks, so I (a) Had the time and (b) Was on my own much of the time (no meeting up with Colleagues for food or drink)

#2 Have a clear time management system. Out of the games, I had played I have only lost a tiny percentage on time – and almost all of them were related to holidays. What I settled on (in the days of server play) is the 10 and 10 method. If I have either been thinking for 10 days or have less than 10 days on my clock then these games are treated as a priority.

#3 Have a clear Opening Repertoire – I tend to play the first 10 moves VERY quickly to put some time on the clock.

#4 Use conditional moves (in server play) where possible – If I have a line I want to play – enter it – if generally takes only an extra minute or so to enter a number of moves and the majority of the time my opponent follows the line, giving me a net benefit!

#5 Save Analysis – any analysis I do is saved in the game file, be it engine lines or my own verbal commentary. This does also help in both annotating games but less obvious in terms of being able to play moves quickly as I can look at continuing down the line I have already started to play. In addition to this, if I undertake a post-mortem after the game the process can be much easier

#6 Unattended Analysis. Both Chessbase and Aquarium allow you to schedule analysis whilst away from the machine. This means I can set up the engine to look at a few candidate moves whilst I am at work or in bed!! This method, when combined with #5, is very powerful

#7 Unintended Tools.  Do you know how to make your software do, what you want it to do? So its vital to learn how to use all its functions but also to figure out what those functions can be used for (which is not always what they are designed for!  An example of this. Reinforcement learning is in fashion at the moment due to AlphaGo but let’s say you have a version of your favourite engine with a learning file. You could (and would) teach it by backsliding within games you are analysis but a method could be to using the Game analysis function of Chessbase/Aquarium both of which perform a kind of backsliding. Let your engine run on key games of your opponent and quickly you have just gained a small edge!

#8 Voodoo.  If you play a lot of games you will start to play certain people repeatedly. Often, although players can have similar rating results are skewed against one player. It is vital to spend to considering how to get an edge in this situations as it will pay dividends in terms of results.

#9 Horses for Courses. Consider why type of analysis you are performing and choose your weapons accordingly. If you want to check a lot of moves quickly you want to use an engine with low TTD (time to depth). If you watch engines for a while you will see that some race to quite a deep position quickly, whilst others take a lot longer to get there – yet the ratings of the engines are similar.  Generally, you want to use Stockfish or AsmFish for fast analysis.

#10 Blunder Check – You are going to be playing a lot of moves and need a method to minimise this – I’ve covered this before but making the move on your analysis screen before making it in the game is a good start

#11 Pragmatic results – Once a game is “drawn” offer the draw and move onto the next game.

#12 Be on the lookout for prepared novelties. Once you start to play a lot of games, people have a large pool of games to prepare against, so keep your eyes open for odd novelties, especially in team events. One antidote to this (and boredom) is to vary your openings.  Anyone looking at me will see I have played almost everything as black and white over the years and now throw them all in on occasion, including recently a King’s Gambit – which I know shocked my opponent, so much so he refused the Gambit and transposed elsewhere!

#13 Hardware – You don’t need a high-end PC or laptop to play Industrial Chess but it helps!  For the majority of my CC time, I have used slightly above average specification hardware. The reality is that for most of us it does not make a difference unless you are competing at the highest levels.

#14 Move Alignment. Consider the games you to analyse and the methods you use. If for example, you use different books or databases for black and white, then it makes sense to analyse all black games then all-white games to reduce the time loading various databases.

#15 Move Now!  If the move is obvious or forced - make it, don't let your time run, especially "sulking" if your position is inferior!

 Till the next time!

Size Matters?

Russell Sherwood  Thursday, December 14, 2017

I was looking through Correspondence Database 2018 and was intrigued to look at the most prolific players.....

 

and there I am sat at No 4 on the list on 4364 games played

Bob Venturas 5433 710 LSS

Maria Vonita 5017 749 LSS

Kazimeriez Bester 4877 858 LSS

Russell Sherwood 4364 2354 ICCF

Sometimes persistence pays of!

 

WCCF 6 & 7 Applications

Russell Sherwood  Wednesday, December 13, 2017

 

In April 2018 we will be starting the next two International Invitational Title Tournaments and invite applications from suitable Welsh Players

 

WCCF 6 is planned to be have an average rating of around 2250, so applications are welcomed from players with a rating of 2100+ on rating list 2018/1 (or players who can show progress towards achieving this level for rating list 2018/2)

 

WCCF 7 is planned to be a Category 6 event (average rating 2376-2400), so applications are welcomed from players with a rating of 2300+ on rating list 2018/1 (or players who can show progress towards achieving this level for rating list 2018/2)

 

In addition to these criteria entry to WCCF 6 will require a £5 payment to the WCCF and all players will be required to confirm acceptance to the WCCF Tournament Conduct Rules (Now being introduced following some players not understanding the privilege or representing Wales in events and the necessary obligations this brings). Details of this will be provided to players at the time of a place being offered.

 

To explain the rating limits used – firstly players below these levels are unlikely to be able to compete successfully and more importantly the key target is the general average rating of the event and every significantly lower rated player drags down this average making the event less attractive for the rest of the participants

 

Examination of the attached shows invitations have gone before and where offers have been made for WCCF 6 & 7

A Correspondence Engine

Russell Sherwood  Wednesday, December 13, 2017

I get asked with tiresome regularity “Which engine is best for CC?” My general answer to this is “none of them” as all the main engines are written and tested with fairly short time controls in mind. As I have written in the past some are better than others and most can be bent into playing better for CC.

 

These are a number of engines with modifications to aid CC players – CorChess and ThinksFish come to mind but these only tend to look at modifying the search rather than the many other areas which could be improved for CC players.

 

Recently I was working on my publication on Engine Analysis techniques (over 30 now!) and what struck me is that most of these methods have to be used to overcome the “Blitz” nature of Chess Engine Settings. So the next thought that popped into my head was “Why not write one for CC”? I mulled this one over for a while and considered the pro and cons.

 

Most things fell into place but one issue remains – which is testing. If I want an engine for CC, then it needs to be tested in CC Conditions as the engine will almost certainly be weaker in Blitz Conditions (as want a CC engine to find a better move in a slower way, rather than a good move quickly). This makes testing rather difficult unless a few people are involved.

 

My first objective is to create an ASMFish/ CorChess Hybrid prior to the development of a CC specialist engine.

 

Any player interested in getting involved (not just Welsh players for once!) please get in touch. You don’t need to be a programmer, Ideas and testing are just as important!

 

ICCF World Championship Preliminaries

Russell Sherwood  Wednesday, December 13, 2017

World Championship Preliminaries

 

The ICCF World Championship Cycle starts again in 2018 with the Preliminaries. It costs £25 to enter (if paid via the WCCF), if qualified and there are two routes to qualification.

 

  1. Winning a Master Class Promotion Tournament. Not is first place is shared this leads to a fractional qualification.
  2. Federation Nomination. We have a very limited number of nominations and so invite applications from interested Welsh Players with a rating above 2250 on 2018/1 rating list. If interested please get in touch indicating why you think you should obtain one of the nominations.

 

 

To explain why the rating limit of 2250 is used. These events a highly competitive and it is unlikely that any player with a rating below this level will have a reasonable chance of Advancement (which is the primary objective of the entry, not Norms or Rating points)

 

Welsh Correspondence Chess FederationBritish Correspondence Chess AssociationSchemingMind Internet Correspondence Chess ClubSocial Correspondence Chess AssociationWelsh Chess UnionInternational Correspondence Chess Association