Since the weather has turned a bit warmer I’ve spent the last month doing some different circles - circles of the local countryside on my bike that is! It’s nice and flat around here as I live by the Severn Estuary (the strong winds compensate for the lack of hills however) and I’ve been putting in about 3 or 4 hours on the bike each week. The gain in fitness has been a great help to my chronic fatigue (at least physically - the mental side is still lacking) and I can now cycle up hills without getting out of breath, I have strength in the legs which I’ve not had since my teens. Cycling is like an anti-depressant, just it’s a bit more effort than popping a pill in your mouth!

There are some great views around here too which helps with the motivation, low traffic means I can concentrate on having fun - here is the view of the estuary at high tide from the M48 cycle path.

Now I’m back with renewed energy, a greater motivation and a hunger to make some more progress. I’ve got one month before the new season starts and I’m looking forward to it again.

If you feel it’s becoming an effort to get the chess board out, take a break!

http://grading.bcfservices.org.uk/getref.php?ref=261619B

Not too bad an improvement over the year.

From what they are saying over at the Streatham & Brixton blog, next season I will have two grades, the official one (probably about 95) and an unofficial corrected one (about 125). The season after that the corrected grade will be the one used.

Confused?

Apparently the grades have been suffering from deflation over several years. Having no idea what a 100 ECF grade player was like 20 years ago I can’t personally give an opinion as how now and then would compare, but some seem to think it’s the ECF handing out an ego boost to those whose grades have slipped rather than a deflation correction.

It makes you wonder what the ECF would have done if the grades had been inflated - I seriously doubt they would have lowered them as that would have demoralised many players, causing them to abandon league chess in their droves. However, as they are in the process of messing with the grades I can’t see why they didn’t just take the opportunity to switch over to the Elo system.

Whatever happens, looks like I’ll make 100 ECF by Summer 2009!

I propose that the new formula should be:

Elo = (ECF x 6.14) + 1050

This gives an R2 fit of 0.946 against the plotted data. The graph below shows my model (black line) vs the ECF model (red line).

I’ll give more details in a future post.

Just over a month in things are going pretty well, although I’ve made some adaptations to the Circles - the first rather fundamental change being that there are now ten circles, not seven!

Using CT-Art as my source of tactics, while the level 10 problems were pretty easy for me, the level 30s were starting to provide a stiff challenge and I was a little worried about slogging my way through the 40s, 50s and 60s without getting to grips with the earlier problems. It would have been nearly two months before coming back to the initial problems if I even made it that far without losing motivation.

I could have done a smaller set of problems, but Don on his blog who has experienced the same issues came up with the Ten Circles idea. The total time is the same, but the longest circle is 32 days (not 64), goes over Levels 10 - 30 three times before going over the 40 - 60s three times and the whole set for the final 4 circles.

Fortunately I was at a point where I could easily switch to this new idea and it should help me to absorb the more basic material before tackling the harder levels.

Already I’m feeling sharper, the solutions are coming quicker and the patterns are starting to transfer themselves to new problems. Down Under Knight gives some good ideas on how to guarantee finishing the circles.

I’m currently on the Level 20s in my 2nd circle. Here is my progress with CT-Art - don’t laugh!

Level 10 : 90% / 93%
Level 20 : 76% / 82% (in progress)
Level 30 : 60% / —

Looking around at how others have gone about studying tactics, I’ve decided to give the ‘Seven Circles‘ method a go. It works on the basis of repetition to repeatedly hammer the concepts into your head and many have had excellent results using this method. There is nothing to say it will work wonders for me, but it looks as good a method as any.

The idea is to work through a set tactical problems slowly, a few each day. Once completed, repeat them, but do twice as many each day in half the time. By the 7th circle they will all be done really quickly in one day.

The first few circles teaches you to understand the tactical concepts in each problem by spending plenty of time working them out. In the last few circles a very short solving time is used and relies largely on your pattern recognition to solve them.

I’ve actually been working on the circles for nearly 4 weeks now but wanted to see if I could stick them before posting about them here. I’m using CT-Art Levels 10 through to 60 to give me a set of 1039 problems with the first circle of 64 days. I’m averaging 18 problems a day, slightly ahead of the required 16/day for the first circle.

The last 2 or 3 circles are certain to be hell, but if I can get through them all I should have completed them just before the new season starts. Whether I can flex my newly toned tactical muscle in OTB play remains to be seen.

Something of a Minor KO specialist, the narrow 3.5 - 2.5 win over Clevedon on 29th April gave Horfield their 3rd Minor KO trophy in 4 years!

> View the Minor Knockout Match Report

However, the big game was a week later when Horfield took on Clifton in the Knockout Final. Clifton had dropped only two points in the league all season and looked favourites for the cup, but Horfield held their nerve to win 4.5 - 3.5!

> View the Knockout Match Report
> Download PGN or CBV of the games

The Knockout Trophy has eluded Horfield since 2000 so it was great to get our hands on it again, but the one we really want is the Division 1 championship which we haven’t won since 2001 - maybe next season!

It’s been a good season - 13 wins, 3 draws and 2 losses. I played up in Division 2 a few times when the B and C teams were short, but I spent most of my time playing for the D team in Division 5 on top board. Having joined the league as a late entry, I was required to organise all the D team fixtures (which was easier with some clubs than with others), and it was very satisfying to win Division 5 in the first season.

Even though the season is over, I’ve got a summer of study and practice planned - tactics, lots of games, strategy, looking over some master games, etc. It looks like my rating will head up towards mid-90s (ECF) when the new grades come out in August and I’m going to need to put some effort in if I want to hit triple figures next year as I only played a couple of games against players rated in the 100s this season and I lost both of them!

For a dedicated chess computer, it’s beautiful - none of that cheap looking plastic that lets down so many other chess computers. The wooden pieces, although nice enough, are a little small for the board and are nothing special.

The auto sensory board works well and allows fairly natural play through the use of magnetic pieces operating reed switches under each square which removes the need to press down to register your move as is common with other computers. 4 LEDs surrounding each square and a 6 digit detachable LCD screen provide a simple but effective system for feedback and menu operation.

The Citrine engine is strong (Novag claim 2330 USCF, but in reality is closer to 2140 USCF) and will provide a tough challenge for even strong club players. For the few who are stronger than this (roughly the top 2% of chess players), there is the option of connecting to a PC via the serial (RS232) cable and with an interface like Arena you can play against UCI or Winboard Engines.

Detach the LCD screen and power adaptor and you have a very reasonable stand-alone chess set offering an excellent compromise for player vs player if a luxury wooden chess set is beyond your budget or you simply don’t have space.

History

Released in 2006, the Novag Citrine is an evolution of the Novag Universal Chess Board (UCB) which acted as an interface for separate engine units (like the Sapphire, Diamond or Super VIP), or to a PC.

Novag then developed the UCB with a new self-contained chess engine and this became the Citrine. Early versions of the Citrine still had the UCB branding, and were plagued with a few bugs and poor design, but now the 3rd generation models are much better made and no longer have the bugs or UCB branding.

Technical / Features

Program Size : 56K Byte ROM
RAM : 3K Byte RAM
System Clock Speed : 20MHz
RISC Processor : H8/300H
Power Supply : 50mA max
Power Consumption : 450mW max
Adaptor : 9V DC
Opening Book : Approx 24,000 ply
Playing levels : 64
Tack back up : 112 ply
Solves Mate up to : Mate in 8
Depth Search : 18 ply
Wooden Board : 370 x 370 x 35 mm

The Board

Novag have provided a generous 37 cm x 37 cm of laminated wooden board to do battle on, with a square size of 3.7 cm (1.5 in). Standing at a height of 3.5 cm to contain the electronics, the board is raised a little higher than other boards, but doesn’t take much getting used to. It’s not in the same league in terms of quality as a luxury solid wood board, but as far as dedicated chess computers go it’s pretty impressive.

With virtually all chess computers manufactured out of some sort of plastic there isn’t much to compare the Citrine to in terms of appearance. The more expensive auto-sensory wooden DGT board looks fantastic but only operates as an interface for PC programs - there is no internal chess engine. The Excalibur has auto-sensor input, but has a rather cheap looking plastic finish and no PC connectivity.

Pieces

The pieces which come with the board are probably the most disappointing aspect of this computer. They are dwarfed by the size of the squares and the pawns in particular look rather puny. The Rook and Bishop look malnourished, only the King and Queen make any impression. The finish lacks detail and no weighting makes the pieces feel very light. They are much better suited to my £10 chess/snakes and ladders wooden board from Argos. However the felt bases feel very nice!

Fortunately, ChessBaron have provided a solution to this shortcoming by creating a set especially for the Citrine. You can purchase them with the board, or separately if you wish. I wasn’t sure whether it was worth the extra money, but they give you the option to return them if you aren’t happy with them. I needn’t have worried! The pieces are thousand times better, they are the right size for the squares and the triple weighting feels wonderful. They give the Novag Citrine an air of class that it deserves. The only drawback to the pieces is that the heavier weight makes it easier to mark the board if they are dropped.

Novag Citrine King - Diameter : 1.0 in ; Height : 2.5 in ; Weight : 10 g
ChessBaron Citrine King - Diameter : 1.25 in ; Height : 3.0 in ; Weight : 24 g

Move Input

Movement of the pieces is as simple as picking up a piece and placing it down, just as you would on a real board. Well, it’s not quite that simple but not far off. Once you understand how the board works you will realise there are certain techniques that have to be learnt (or avoided) to give faultless operation.

There are magnetically operated reed switches under each square and as a piece (with a magnetic base) is placed on a square the reed switch is closed and the board recognises there is a piece on the square. Lift the piece off the square, the magnetic field is removed and the reed switch opens - the board recognises the piece has been removed. Although the pieces are magnetic, there is no sense of ‘sticking’ like with magnetic travel sets (because the reed switches are so small) and the pieces will slide off if the board is shaken or tilted - but on the other hand they do stick quite nicely to radiators!

The method of operation gives rise to some potential problems. If you are a ‘pusher’ - that is you like to slide your pieces to the destination square, then you will have problems. This should be obvious if you consider the common opening move 1. e4. By pushing the pawn off e2, it opens the e2 switch, as it passes over e3 it closes the switch and opens again as it moves off e3 and onto e4, closing the e4 switch. The computer sees the piece leave e2 and arrive on e3 (completing the move)- which it registers as the move e2-e3. It is now black to move and moving the pawn from e3 to e4 is illegal since it’s no longer white’s turn.

In short, the piece must be picked up cleanly and placed on the destination square without activating any intervening squares.

It also has to be realised that the reed switches don’t open until the piece is lifted about ¼ - ½ inch above the square, so ‘hovering’ low over the squares causes problems and also the common capture technique of ‘kicking’ one piece out of the way with the capturing piece doesn’t work. The reason for this is that the computer has a specific sequence that has to be followed when capturing: the captor must leave the original square first, the captured piece then has to leave its square and then the captor must be placed on the vacated square. If you ‘swoop’ in with the captor to ‘kick’ out the other piece, then the reed switch is never opened between removing the captured piece and placing the piece on the vacated square - the computer is none the wiser and doesn’t realise that anything has moved as the switch hasn’t changed state at any point.

Using reed switches creates some limitations in how freely the pieces can be moved, and to some extent you are forced into a certain style of moving the pieces. Whether this suits you and how quickly you adapt to it will depend largely on what your style of moving the pieces is and how ingrained your style is should you need to modify it - I wouldn’t recommend playing much Blitz until the necessary adaptation has occurred!

The Engine

The brains behinds the Citrine engine is a H8/300H-tiny RISC processor (more specifically H8/3687) running at 20MHz - but as my knowledge of processors and electronics is rather limited I can’t tell you much about the processor - have a look at the manufacturer’s page instead if you are interested!

Link: Renesas H8/3867

Novag advertise the Citrine with an estimated rating of 2330 USCF (about 2230 Elo), but Selective Search place it 12th in their list of dedicated computers with a rating of 2032 Elo. Most manufacturers will overstate the true strength, or estimate it on impractical parameters such as infinite thinking time.

There are 64 playing levels to choose from, but rather than being a sequential progression in strength from 1 to 64, there are 8 categories - though the mate-in levels are not really suitable for playing against. There are choices between tournament type games, quickplay games, fixed depth, fixed time and some beginner levels. There is also an ‘easy’ option which can be applied to any level which turns off pondering (thinking on the opponent’s time) and a ‘random’ feature which changes the level of randomness in choosing between similarly good moves.

The Citrine is supported by an opening book of 24,000 half moves, which gives a huge range of opening potential. The only disappointment is that there are no configuration options to go with it like ‘tournament’ to strengthen it up or ‘random’ to give it a bit more variability.

I thought there might have been a bit more flexibility on the levels, but as I’m down the lower end of the levels there is plenty to keep me occupied.

Usability

The usability of the Novag Citrine is actually very good once you get used to it - doesn’t take long! A new game is started by simply placing all the pieces on their ‘home’ squares. The moves are displayed using LEDs (1 on each corner of the origin / destination square) with optional sound alerts and makes for very easy move recognition and input.

The menu system is activated via the menu button on the side which turns the board into a big 64 button console (there is a spare function activation piece for this purpose). The menu options flash round in the LCD display, though it’s much easier to refer to the manual - they really should have supplied a quick reference sheet so I’ll make my own at some point.

The board is powered by a 9V DC adaptor and there is no alternative battery option - which means to use the board you have to be permanently plugged into the wall and the memory is lost once it is unplugged.

LCD screen

The game information is displayed through a small plastic box with a six-digit LCD display. The main display shows information such as the move input, time taken/remaining, computer’s reply, computer’s evaluation, computer’s search depth, variations left to compute, game result and menu selection/options. There are other icons on the display to show side to move, captures, en passant and piece position verification.

As most of my time is spent concentrating on the board during the computer’s move I don’t look at the LCD all that often so it’s not all that useful to me, but it is useful when a move doesn’t register properly or you want to take back a series of moves. Display of the engine feedback is useful in analysis mode, but then I’d just use Fritz instead to analyse a game. In games with a time control, your remaining time is shown during your move, but during the computer’s move it rotates through a whole series of displays so the time remaining for the computer is easy to miss.

PC connection

Who cares if there is a lack of levels to play against when you can connect up your Novag Citrine to a PC and play directly against any UCI or Winboard engine you like?! With a bit of technical know-how you can turn your Citrine from a 2000 Elo grandmaster wannabe into a 3000 Elo world beating monster!!

I know that the majority of my games will be spent hooked up to a computer playing against various UCI engines which have been weakened to an appropriate strength (to give me a chance!) and at a time control of my choosing.

The board is supplied with a RS232 (serial) connector which will connect to the COM port on your PC - however, not many computers (laptops in particular) are fitted with these ports nowadays due to the superiority of USB ports. Not to worry though, for a just few pounds it is easy to pick up a USB-to-Serial converter which allows you to use a USB port instead.

Using a software interface such as Arena, the incorporated Citrine driver will allow the board to control move input. It will also receive and display moves from the computer, though moves with captures don’t quite display properly using the LEDs and it may be necessary to refer to the computer screen if move announcement isn’t turned on. None of the game information is displayed in the LCD either, but doesn’t need to be attached in any case.

Summary

The experience of playing against a computer by physically moving wooden pieces on a wooden board is infinitely better than trying to play using a computer screen, and further enhanced by the pieces from ChessBaron. The auto-sensory mechanism makes play feel almost natural and an innovative menu system and detachable LCD means the board isn’t cluttered with buttons and displays.

The board is very usable though perhaps the manual could use a little expansion to make things a bit clearer. The internal engine is not as strong as other chess computers and the level selection is a little limited, but to me that doesn’t matter because with the PC connection it becomes the Novag Citrine Unleashed!

Pros: Auto-sensor, PC connectivity, LEDs, mains powered
Cons: Chess piece quality, limited levels, no USB connection (serial only), no battery option

If you are looking to buy a new chess computer I’d urge you to go with an auto-sensor. Having owned a press-sensor for many years, I can tell you that being able to move the pieces naturally improves the experience no end - it allows you to completely immerse yourself in the game of chess rather than having to think about how to input the moves all the time. For this reason alone the Novag Citrine does not disappoint.

I failed to listen to this great piece of advice in my match on Tuesday evening. Having seen a line which was favourable (winning the exchange) I played through the line not considering that something much better might be lurking elsewhere on the board.

Had my mind been more open to other possibilites I’d probably have seen the better move. Had the possibility to win the exchange not been there I’d probably have seen the better move.

My previous move had intentionally threatened to win the exchange and when my opponent’s reply didn’t address this threat I seized the material on offer. What I didn’t notice was his move now allowed me to force his King to a square where I could subsequently fork his Queen and King - winning the Q for a piece.

Unless you are under time pressure, after finding a good move it often pays to spend a few extra moments to look for something even better.