11 November 2018

'The World Is Watching'

Today is the rest day after the first two games of the 2018 Carlsen - Caruana title match. For the blog's November video let's look at what's available from YouTube's World Chess channel, the organizer of the event. So far the channel offers:-

  • Opening Press Conference
  • Opening Gala
  • Game 1
  • Game 2

Here's the start of the first game, although you would never know it from the title or the description of the video.


World Chess Championship London 2018 (34:01) • 'Streamed live on Nov 9, 2018'

The description of the clip says only,

First 30 minutes of the World Chess Championship

There is no game number, no link to the official site, no explanation of why only the first 30 minutes are shown or how to watch the rest of the game. C'mon World Chess! You are supposed to be the experts in marketing chess to the entire world. You can do better than this. The banner in front of the players promises, 'The World Is Watching'. Make it happen!

If you've ever wondered about GM Caruana's nationality, he explained it during the 'Opening Press Conference'. In response to the question,

You have changed nationality a couple of times. Do you feel American or Italian or what do you feel?

he said,

I never changed nationalities. I had dual nationality from birth. I changed federations when I was 13. I was a kid, I was living in Europe, and this was mainly to get back to my roots -- my mother is Italian. When I changed back to the U.S. I was in my 20s. This was a personal decision and I feel connected to both countries. I would like to represent both, but only one is possible. I live in the United States and represent the U.S. now. I feel very much an American, but I cherish my Italian roots.

In case you're wondering, the official site is london2018.worldchess.com. The first two games both ended in draws. Good luck to both players!

09 November 2018

GPU Pricing

Let's say I want to buy one of those machines featured in my previous post, GPU Configurations. How much money will I have to shake out of the piggy bank?

The 'Lambda Blade' link I referenced leads to another page Blade : Customize. There are three configurations, shown below, where the model in the previous post is the 'Basic' version.

The pricing for the three models is:-

Basic • $20,788.00
Premium • $25,852.00
Max • $51,958.00
also: Customize

In other words, I can buy a new car or I can buy a 'Deep Learning' rig. I'm sure my wife would prefer a new car.

The 'Customize' option starts with the 'Premium' model and allows to add/subtract options to determine a final price. The numbers are consistent. If I decrease 'GPUs' and 'Memory' to arrive at the 'Basic' model, the price is the same for that model.

What exactly are the technical tradeoffs for the different prices? I'll look at that in another post.

08 November 2018

Match Tales from the Twittersphere

The much anticipated 2018 Carlsen - Caruana World Championship match starts tomorrow and journalistic protocol requires at least a perfunctory nod. What is Twitter saying about it?

On top of writing about the match itself, the Twitter angle gives me the chance to try another of the Algorithmia (AI?) services that I first used in From Black & White to Brown & Blue (October 2018). The particular service is Retrieve Tweets With Keyword (algorithmia.com). I fed it the three keywords 'chess carlsen caruana' and it spit out text examples of the 500 most recent tweets/retweets on the subject, in reverse chronological order, the earliest from yesterday. The 'retweets' aspect means that the same tweets are repeated over-and-over-and-over, most of which are a link to a source other than Twitter.

Of all the standalone tweets that I looked at, the most interesting is shown below. I initially tried to use the tweet embedded directly into this post, but the dimensions of the photo broke the blog container. I couldn't see a way to fix this quickly, so I resorted to a screen capture and shrank it.


https://twitter.com/seaningle/status/1060533865259954177

As for linked articles, the Twittersphere gave me three from Chess.com:-

Of the non-chess sources flagged by Twitter, another handful appeared over-and-over-and-over:-

As for the oldest of the 500 tweets/retweets, the distinction goes to another article from Chess.com:-

As interesting as the Algorithmia exercise was, I could have saved a lot of time with a straight Twitter query:-

twitter.com/search?q=carlsen+caruana.

Next time I'll just do that.

06 November 2018

Chess in Antarctica

That's the South Pole in the crosshairs on the map of Antarctica. The word 'McMurdo' is to its left.


May 1964, Source unknown

The accompanying newspaper clipping said,

CHESS ON ICE • Static Stops Play

A chess tournament between Christchurch [New Zealand] players and Americans at McMurdo Sound, Antarctica, had to be adjourned with 10 games only half-way through when radio conditions deteriorated yesterday afternoon.

The play was tapped out on teleprinters in the communications room at the American base at Harewood and on teleprinters at McMurdo Station. But as the sun sank lower the radio frequency had to be changed, and difficulties at the Antarctic end began to slow up the games. The Americans suggested an adjournment to Sunday, June 14.

"It's a bit hard to tell," said Mr. A. S. Hollander, president of the Canterbury Chess Club, "but it looks as if we are leading." Chess is one of the very few ways in which the winter population of McMurdo Sound can participate in the life of the outside world, and Mr. Hollander thinks the Americans would like to play a lot more chess in future. "I'm sure there'll be a return match at least," he said.

My guess is that the photo was taken on the New Zealand side of the radio match. Why are there two boards in front of each player?

05 November 2018

So Many Ideas, So Little Time

In last week's post, Happy 10th Anniversary, Stockfish!, I noted,

The name of choice for the commemorative release was 'Stockfish X', but with only a few days to go, it doesn't look like anything special is going to happen.

And nothing did happen. Let's continue with the Stockfish forum, FishCooking -- first seen on this blog three years ago in Chess Engines : FishCooking (October 2015) -- and look at some of the recent posts that attracted discussion from the Stockfish community *and* that provide some insight into the group's working methods. First up are a couple of forum posts about selecting a version for the TCEC final match, which is currently underway and nearing its conclusion.

  • 2018-10-20: Executable for TCEC superfinal? • 'Apparently we haven't submitted an update to play in the TCEC Superfinal yet. I suspect it is too late now, but perhaps the senior devs could prepare one or comment on whether the e.g. abrok one should be used and contact TCEC?'

  • 2018-10-21: Request for binary for TCEC superfinal • Marco Costalba: 'We have few hours left before TCEC deadline is met to provide a binary for the superfinal. We would need a fast binary of the current master within this evening (European time) and I will send to TCEC people. If not, we will fallback on current TCEC binary. If someone has a better idea, please share it.'

The open source nature of Stockfish development presents many challenges.

  • 2018-10-17: Discussion for optimizing methodology • 'SF has been able to progress with the same 7(?) year old recipe and without doubt still is, but in my opinion with some adjustments in accordance to its current situation and needs it can be helped tremendously'

A technical issue related to evaluation has been under investigation for months.

  • 2018-10-03: First results of contempt tests • 'Actually on STC it seems that optimal contempt value is somewhere between 30 and 70. I will try to build the same diagram for LTC and if it shows the same stuff then we can probably conclude that for TCEC divP optimal value of contempt will be much higher than default 21 - closer to 50 because field there is much weaker than SF9, in fact, every single engine there is weaker than SF9.'

As those excerpts show, the forum members use so many acronyms and so much jargon that they might easily dissuade casual chess engine fans from trying to understand what they are discussing. Here are a few terms that are worth knowing. These first two acronyms I covered in a recent post titled Catching Up with Engine Competitions (October 2018):-

CCCC = Chess.com Computer Chess Championship
TCEC = Top [formerly 'Thoresen'] Chess Engines Competition

The next two acronyms are used throughout the forum. If you don't know what they mean, there will be many discussions that you won't understand. The terms are key elements of a testing strategy that starts with STC games and, if no problems are discovered, continues with LTC games.

STC = Short time control
LTC = Long time control

This next term is not specific to chess.

SPRT = Sequential Probability Ratio Test

It is a type of statistical analysis that is explained in Wikipedia's Sequential probability ratio test. The next acronym stems from a web domain that is used for administration, e.g. Stockfish Development Versions.

ABROK = abrok.eu

The word 'contempt' in the last thread listed above is a current topic of experimentation. In Contempt Factor (chessprogramming.org), the wiki defines it as:-

The Contempt Factor reflects the estimated superiority/inferiority of the program over its opponent. The Contempt factor is assigned as draw score to avoid (early) draws against apparently weaker opponents, or to prefer draws versus stronger opponents otherwise.

In other words, even if the evaluation shows a small advantage for the opponent, treat it as equality (a 0.00 evaluation). The typical debate is about what numerical value should be used. If it's too large, you risk losing drawn games; if too small, you draw games that you might win.

Another thread that has been running for nearly two years points to a useful tool for understanding Stockfish evaluations.

  • 2017-01-22: "Stockfish Evaluation Guide" tool • 'I developed tool where you can investigate each part of Stockfish static evaluation function. It is standalone single HTML page with javascript. Every evaluation term is rewritten in single small javascript function. You can setup any position with FEN or by moving pieces on chessboard and see how evaluation is computed and what is result and if possible to attach score to individal squares it is visualized on chessboard.'

The thread eventually points to Main evaluation (hxim.github.io). If I had the time, I would definitely take a closer look at it. I could say the same for many of the threads in the FishCooking forum.

04 November 2018

Elegant 19th Century Austrians

For this November edition of Top eBay Chess Items by Price (March 2010), my first choice would have been a painting titled 'Charles Schreiber, 1845-1903 French, "A Game of Chess" Oil'. Then I realized that it was exactly the same item that I featured three months ago in Chess with Two Cardinals (August 2018) -- same title, same description. This time the painting sold for US $1500 after one bid.

Instead of the cardinals for a second time, I picked the painting shown below. Titled 'A Game of Chess Antique Genre Oil Painting Josef Morgan (Austrian, 1839–1898)', subtitled 'Attributed to Morgan - Signed with Initials - Inscribed', it sold for GBP 620.00 ('approximately US $803.71', according to eBay) after receiving 37 bids from 14 bidders.

The description added little to the information given in the title:-

A very fine 19th century genre painting depicting an elegantly dressed couple playing chess in an interior which is presumably the work of Austrian artist Josef Morgan (Austrian, 1839–1898).

The frame (not shown here) carried a small plaque, 'J. Morgan 1894'. Another variant of the name is Jozef Belohlawek Morgan. Whatever the artist's real identity might be, we have another reminder that Chess Paintings Require Dogs (April 2011).

02 November 2018

GPU Configurations

In my previous post, GPU Specifications, I closed with:-

The emphasis in the video is on gaming, but the connection with AI is clear enough. I'll look at that in my next post.

This led me to a page, Deep Learning Server - 8x GPUs (lambdalabs.com), about a specific product, called the 'Lambda Blade'. The essence of the product is captured in the following image, the physical machine on the left, its component list on the right.

The essence of the 'What's Inside' column is repeated below. The small print in the image lists options for upgrading this basic configuration.

OS: Ubuntu 18.04 (Bionic)
GPUs: 8 x NVIDIA GeForce GTX 1080 Ti
CPUs: 2 x Intel Xeon E5-2650 (12 cores, 30M Cache, 2.20 GHz)
RAM: 128 GB DDR4 ECC
SSD: 2 TB (SATA)
HDD: 4 TB (SATA)

This basic introduction leads to a 'Customize' page, which I'll look at in my next post.