H.E.Atkins

[Geoff Chandler]

Hi John,

I've copied that from a site somewhere 10 years ago.

At one time the game was known as Atkins - Gibson 1924 so I've got it from an uncorrected site. (somewhere.)

See page 84 'Great Short Games of the Chess Masters' by Reinfeld. (published 1961)

Same game given as Atkins - Gibson, Southport 1924. 'The Soul of the Game'

If you do not have the book - surely you do it's a classic! here is a link.

https://books.google.co.uk/books?id=Otn ... 24&f=false


Edit: there is also a vid of it with the same players.

https://www.youtube.com/watch?v=vhNOB8T9rfg

So it's out there 'somewhere'. (that vid was posted in 2019 - maybe they got it from here!)

[Geoff Chandler]

Hi John

(see above) think I have traced where I got it from

At one time Chessgames.com must have had the game listed as Atkins and I took it from there.

This is the George Thomas game: https://www.chessgames.com/perl/chessgame?gid=1242920.

But down the bottom of that page some lad has used that game for his section on Henry ernest Atkins (1972-1955) (sic)

when you follow that link it takes you to here

https://www.chessgames.com/perl/chessco ... id=1007202

And you can see the lad has that game titled:
Henry ernest Atkins (1972-1955) (he means 1872)

but the link goes to the Sir George Thomas game and he has not corrected the title of the game.

Also see:

https://www.chessgames.com/perl/kibitzi ... 3&reply=10

One lad refers to Atkin - Gibson 1924, click on the link and it takes you to the Thomas-Gibson game.
So they must have had at one time as Atkins - Gibson.

They might have got it from Reinfeld who possibly got it from Chernev.

Chernev's 1000 short games (pub 1957) has it as Atkins - Gibson (game number 466)

Chernev got it from...

[Keith Arkell]

In terms of natural ability and training a sub 4 mile must to this day be in the same ballpark of difficulty as becoming a GM.

Maybe a bit higher than GM even, if we benchmark the best middle distance runner in 1954 to the best chess player, Smyslov.

I did not remember this thread and am surprised by Keith's assessment. I am not sure how much better than a modern 2300 Atkins was, but I would say a lot. So we have a problem if he was also much weaker than 2500!

I don't really want to get drawn into another debate about continually rising and evolving levels of chess playing down the years. The problem is that the observation can be taken the wrong way - as some kind of attack on the amazing chess geniuses of the past, from whom we all learnt so much.

In 'John Nunn's Puzzle Book' the author went to some lengths, back in 1999, to compare the playing level of top flight chess players across a large time-span. Eight years later, IM John Watson quoted the essence of this work. I guess I'm allowed to copy and paste this as it is freely available on the internet, at the foot of Mark Crowther TWIC) https://theweekinchess.com/john-watson- ... tallment-3

Here , then, is what John Nunn wrote:

'The method I chose to examine the games was a two-step process. I reasoned that a good way to eliminate differences resulting from 80 years' advance in chess theory was only to look for really serious errors - if you blunder a piece, it doesn't matter whether you understand Nimzowitsch's pawn-chain theories or not."

[Watson: Notice this important step. I'm always hearing (and reading) that "If the players of yesteryear could only catch up with opening theory, they'd be as good or better than today's players". The funny thing is that the many years (usually decades) of study that modern players put into opening theory should not only count towards their strength, but that study and practice contributes vastly to their understanding of the middlegame and even some endgames. The silly idea that you can just 'catch up' in opening theory ignores the vast undertaking that this would involve, especially to absorb the vast number of openings and opening variations necessary to a complete chess education. Nunn removes this factor from the equation, to the enormous detriment of the modern masters' strength assessment! Surely this will roughly equalise things? Let him continue: ]

"To analyse almost 800 games from scratch by hand would take years, so first I used the automatic analysis feature of Fritz 5 to look at the games without human intervention. It was set in 'blundercheck' mode, which fitted in with my objective of looking for serious errors. Then I examined 'by hand' all the points raised by Fritz to decide whether they were genuine blunders or products of Fritz's imagination.

I had no particular preconceptions about what the results of this search would be. Like most contemporary grandmasters, I was familiar with all the standard textbook examples from the early part of the century, but I had never before undertaken a systematic examination of a large number of old games. I was quite surprised by the results. To summarize, the old players were much worse than I expected. The blunders thrown up by Fritz were so awful that I looked at a considerable number of complete games 'by hand', wondering if the Fritz results really reflected the general standard of play. They did. By comparison, the Fritz search on the 1993 Biel Interzonal revealed relatively little; many of the points raised had already been examined in the players' own notes in Informator and elsewhere. I had originally intended to have the Karlsbad and Biel positions side-by-side in this chapter, but the results were so lopsided that I decided to concentrate on Karlsbad here. Some of the more interesting Biel positions may be found scattered throughout the rest of the book.

In order to be more specific about Karlsbad, take one player: Hugo Süchting (1874-1916). At Karlsbad he scored 11.5/13.5 or 'minus 2', as they say these days - a perfectly respectable score. Having played over all his games at Karlsbad I think that I can confidently state that his playing strength was not greater than Elo 2100 (BCF 187) - and that was on a good day and with a following wind. Here are a couple of examples of his play:"...

[Watson: You have to get the book to see these examples of Süchting's horrendous mistakes and misunderstandings. Nunn also has talks about more positions, and then includes a section of 30 Karlsbad "puzzles", representing all of the players. The positional mistakes by the top players are particularly telling.]

"How, then, did Süchting manage to score 11.5 points in such company? Well, he did have a couple of slices of luck - Duz- Khotimirsky overstepped the time limit while two pawns up in a completely winning rook ending and Alapin agreed a draw in a position where he could win a piece straight away. However, there were some games where Süchting might have hoped for more; he certainly had Levenfish on the ropes (see puzzle 184), and he agreed a draw in the following position against E.Cohn:" [Diagram follows] "It is hard to understand this decision, as with a clear extra pawn Black certainly has very good winning chances and could proceed without the slightest element of risk."...

"Returning then to the question as to how Süchting scored 11.5 points, the answer is simply that the other players were not much better. If we assume Süchting as 2100, then his score implies an average rating for the tournament of 2129 - it would not even be assigned a category today. Based on the above, readers will not be surprised when I say that my general impression of the play at Karlsbad was quite poor, but the main flaws did not show cup in the areas I expected..."

[Watson: Here Nunn shows that openings weren't a problem in this tournament for the older players (who specialised in a few systems). Then he points out the generous time-limits in Karlsbad. Having eliminated those factors, he gives three reasons for his the weak play of the Karlsbad group:]

"The first was a tendency to make serious oversights. It is quite clear that the Karlsbad players were far more prone to severe errors than contemporary players. Even the leading players made fairly frequent blunders. Rubinstein, for example, who was then at virtually the peak of his career (1912 was his best year) failed to win with a clear extra rook against Tartakower ... He also allowed a knight fork of king and rook in an ending against Kostic..."

"The second problem area was an inclination to adopt totally the wrong plan...[examples follow]..."

"The third main problem area was that of endgame play...[horrendous examples of elementary blown endgames follow]..."

[Watson: In the course of research for a book, I made a lengthy look at endgames from a comparable period and found similar butchery, including some terrible blunders by top players such Lasker. The endgame skills of the great masters - excepting Rubinstein - are much exaggerated in books, for the reasons that Nunn gives, i.e, the understandable selection of a very small set of games for reasons of instruction and beauty.]

Nunn goes on to discuss the treatment of games in older tournament books, comparing them with today's analytical approach. He addresses an obvious objection:

"Doubtless, some will respond by searching through contemporary tournaments and finding errors just as serious as those presented here. However, a couple of words of caution. Remember that all the examples given here were played in one tournament. Of course, it is easy to present a player as an idiot by listing the very worst blunders from his (or her) entire career, but that is hardly the point - it is the frequency of errors which is important. The second cautionary word concerns the method of measuring the frequency of errors. You cannot just take a tournament book and count the number of question marks; modern players are far more critical and objective than their predecessors. Although there are exceptions, tournament books from the early part of the century seem to be strong on flowery rhetoric but weak on pointing out mistakes. You actually have to analyse the games to obtain a realistic assessment of the standard of play; one day, perhaps, you will be able to feed a selection of games to Fritz and it will come back with the players' Elo ratings, but that day has not yet arrived.

'Nunn's argument makes sense to me, and I can subscribe to its conclusions. Of course, it would be interesting to hear from someone with a contrary point of view.' (John Watson).

[Paul McKeown]

I find it really strange how people are referring to all the athletes from the present day who have run sub-4 miles, perceptibly diminishing Roger Bannister’s feat in their eyes. I strongly doubt that most of those athletes would sub-4, if they had to do it on the crappy leg-deadening tracks that Bannister had to do it on, using Bannister’s lumping great clogs instead of today’s highly engineered running shoes. The mechanical saving of the two factors is more than significant. Sub-4 miling is, in my view, a higher standard in track than GM is in chess.

[Paul McKeown]

The sanest comment on this thread remains Andrew Martin’s right at the start, ten years ago:

You'll just have to give him his due; surely he played some decent games?

[Paul Cooksey]

I don't really want to get drawn into another debate about continually rising and evolving levels of chess playing down the years.

Keith has blundered by saying interesting things about the subject he does not want to get drawn into!

This is an quote that interests me

You actually have to analyse the games to obtain a realistic assessment of the standard of play; one day, perhaps, you will be able to feed a selection of games to Fritz and it will come back with the players' Elo ratings, but that day has not yet arrived.

Twenty years later AlphaZero probably has the raw power should this be a task that interested DeepMind. But then you get in to the difficulty of what question we are asking, familiar to Douglas Adams' fans.

I don't think many people have Karpov as the greatest player of all time. Neilsen and Gustafsson had him at 6 in the chess24 series. But who beats him consistently on his best day? No-one in my opinion. Beating him once in a high pressure game is probably a key factor in being ranked number 1. Had Karpov played just fractionally better on 18 December 1987, probably we would see legacies differently differently. I'm not sure how AlphaZero assess that situation objectively. Worse, I don't think our possible score against best play is a definitive measure either. It undervalues those whose approach is to take their opponent into a deep dark forest where 2+2=5, and the path leading out is only wide enough for one.

So, I don't think even AlphaZero could really know. I know I can't. It always amuses me when people argue that Fischer is better than Kasparov or vice versa. You could show me a game by Leonid Stein, tell me it was by either of them and I would probably believe you. I don't have the competence to judge.

But, many years ago when we discussing the history of British number one players I though it would be interesting to work out when the British number one was better than me, higher than 2300. I thought it would be interesting to play through the games of some players I did not know, which it was, and that I would be able to tell, which I think I could. Not more talented of course, clearly they all were, but better even if I turn up standing on the shoulders of giants and having used chessbase to prepare.

I think Blackburne was probably better than me. But I am not absolutely sure. His best games are fantastic, his worst games poor. I suspect I would get good positions against him and then get blown away tactically.

Atkins is the first player I am sure was better than me, based on maybe 10 hours playing through his games with the computer running and thinking "could I have found that?". Although for a given value of sure. Nunn's point that to know how well someone plays you need to know how strong their opponents were means I am not absolutely sure. I would need help from AlphaZero.

(PS when I started I thought I might be better than Alexander. I'd like to apologise to his ghost for my hubris...)

[JustinHorton]

one day, perhaps, you will be able to feed a selection of games to Fritz and it will come back with the players' Elo ratings

How far from that point are we? I don't mean an exact rating of course, but say a fifty-point band. Is this a feasible proposition? (I ask while being aware that, as per Paul above, we need to know what question we are asking, e.g. are we talking about FIDE ratings, are we talking about normal time limits etc. But one of my reasons for asking is that I am interested in establishing how different chess strength may be in different disciplines, e.g. between rapid and normal limits, or between normal and correspondence.)

[Roger de Coverly]

How far from that point are we? I don't mean an exact rating of course, but say a fifty-point band. Is this a feasible proposition?

Those who attempt to detect the presence of external assistance from the relationship between quality of moves and current rating are claiming to be able to do this.

[Nick Burrows]

I think Blackburne was probably better than me. But I am not absolutely sure. His best games are fantastic, his worst poor. I suspect I would get good positions against him and then get blown away tactically.

Atkins is the first player I am sure was better than me, based on maybe 10 hours playing through his games with the computer running and thinking "could I have found that?". Although for a given value of sure. Nunn's point that to know how well someone plays you need to know how strong their opponents were means I am not absolutely sure. I would need help from AlphaZero.

(PS when I started I thought I might be better than Alexander. I'd like to apologise to his ghost for my hubris...)

Very interesting post Paul, thanks.

Similarly, in the intro to Quality Chess' excellent version of 'My System', Aagaard says that Nimzowitsch was definitely better than him.

[John McKenna]

"... Aagaard says that Nimzowitsch was definitely better than him."

FWIW -

Jacob Aagaard, now about 48 years of age and currently FIDE rated 2477, reached his highest rating of 2542, May 2010.

Aaron Nimzovitch (Nimzowitsch) who died aged 48 years (07 Nov 1886 - 16 Mar 1935) is given a "best 5-yr. average" rating of 2615 (peaking at about 2650 around 1927) by Prof. A. Elo in his seminal book The Rating of Chessplayers - Past & Present, believe it or not.

[Richard James]

If I might venture a wild guess regarding the average strength of say the top five or top ten players throughout the [19th] century I would say it gradually went from about 2000 around the [1830s] to 2400 near the end of the century.

[John McKenna]

If I might venture a wild guess regarding the average strength of say the top five or top ten players throughout the [19th] century I would say it gradually went from about 2000 around the [1830s] to 2400 near the end of the century.

"2400 near the end of the century" seems a bit of an underestimate according to Prof. Elo as a graph of his shows the following approx. historical ratings in 1900 -

2700 Lasker
2600 Tarrasch
2590 Maroczy
2575 Janowski
2550 Chigorin
2500 Marshall
2475 Blackburne
2450 Mieses
2400 Gunsberg
2370 Mason

Giving an average rating of 2521 for 10 "lifetime ratings of selected chessmasters" (and does not include Pillsbury and others who had shorter, meteoric chess careers) then.

NB - The 11th, and final, player on the graph ar the time is Schallopp - rated about 2320. Including him in the calculation brings the average down to 2503 (rounded up).

[JustinHorton]

How far from that point are we? I don't mean an exact rating of course, but say a fifty-point band. Is this a feasible proposition?

Those who attempt to detect the presence of external assistance from the relationship between quality of moves and current rating are claiming to be able to do this.

They're not really, and I am specifically not interested here in questinos of cheating. I am asking whether there is sufficient reliable data, reliably interpreted, that would allow us to take a sufficiently large selection of an individual's games and say from the moves, rather than the results, that their average level of performance falls within such-and-such a band.

[John McKenna]

Got to say about what is immediately posted above - Prof. Elo started that ball rolling, and I agree with RdC that Prof. Regan's s/w has some relation to it, also.

However, perhaps this is better -

http://www.edochess.ca/

Don't know if anything will pass the Horton reliability test, though.

[NickFaulks]

http://www.edochess.ca/

Unfortunately, Edwards' work is founded on the now discredited theory that being good at chess is about beating the guy on the other side of the board. Modern pundits understand that it is really about matching computer moves.