You said your method of analysis was based upon the rank of a move, but such way is inaccurate. In some positions the second best move is quite far from the best one, but in other positions the 20th best move may not be much inferior to the best one.

The fact that even lower rated players make normal moves mostly is obvious. Otherwise they wouldn't be anything but total patzers.

Setting engine parameters in a desired way is not completely sufficient. The aim is to make a computer to play exactly in a way as famous players from the past did. What constitutes a style of play is, in my opinion, more than just the dichotomy of tactical - positional. Kramnik, Petrosian and Carlsen are all positional players, but in fact treir styles differ nevertheless.

Also, playing strength is an integral part of how one plays. Strength and style are not fully detachable notions. The stronger the play, the less variety in style. So, for example, I think it is impossible to imitate Philidor playing on 2900 level.

<The stronger the play, the less variety in style.> Yes and no. I'd meant to say that a lot of extreme stylistic variety is found among players rated 2400-2550 - 'weak' GMs and IMs. Neo-hypermoderns like Suttles, complexifiers such as Simon Williams, and many other eccentrics as well as mainstream styles.

But I also think there's an 'amateur plateau' a couple of stages below that, around 1800 ELO. Where players have absorbed precepts but aren't strong enough to deviate from them effectively, resulting in a predictable style.

This is a crazy conclusion, though I accept that the current crop of elite players is stronger in some senses than its predecessors. As chess knowledge tends to increase, this has been true of every generation in the past 150 years.

However, the extreme chessmetrical approach - putting Morphy on a par with Tal and Shirov - is even more ridiculous.

I have a third option. Rating inflation is real, and so is a genuine overall increase in technique. But players from the past were capable of incredible feats of raw calculation - in some ways, deeper than anything a human does today. Perhaps, as in the case of Adolf Anderssen's two famous combinations, the attacking player was less aware of defensive ideas. Pre-Steinitz, Pre-Nimzowitsch, strong players calculated very deeply over a smaller set of ideas: fewer positional considerations, much less opening baggage, etc. Perhaps it freed up their brains for calculation?

Leading, somewhat paradoxically, to occasional deeper combinations despite a lower overall playing standard.

But it at least had an eval function displayed in a numerical way and was, too, able to show multiple variations simultaneously?

<I'd meant to say that a lot of extreme stylistic variety is found among players rated 2400-2550 - 'weak' GMs and IMs.>

To me this seems too categorical. Where did you get this idea anyway?

The lower one's level of play is, the more room his play has for stylistic preferences. Strong GM-s cannot afford too much style, because it would start hurting their accuracy of play and performance.

All GM-s have relatively similar chessic upbringing; they went to chess schools, had trainers/teachers who generally have similar approaches and methodologies. Most of the books they have read are the same.

These two principles put together means that GM games have less stylistic colour than coffeehouse players do.

The thing is that bottom-rated players generally have quite haphazardly learned anything other than moves. Some have properly studied beginner books, others haven't. Some have inclination to play wildly, do chess-gambling over the board, others never have courage to sacrifice anything. Some like to study opening moves and prepare as deeply as possible, others just don't care about it at all and prefer tactical puzzles instead, etc.

You never know what they have in stock for you. The variety of play is immensely large.

<But I also think there's an 'amateur plateau' a couple of stages below that, around 1800 ELO. Where players have absorbed precepts but aren't strong enough to deviate from them effectively, resulting in a predictable style.>

Sounds too arbitrary. They haven't mastered anything and nothing suggests they have something uniform in their knowlegde and skill. In fact, the possibilites for deviation diminish as one steps up along the ladder of chess mastery towards GM-hood. The accuracy of play becomes much more crucial.

In fact it's very similar to the first video I made except the person knows better how to use the program- particularly the zoom ins and zoom outs-

And her transitional fades are seamless, no jarring-

Also the <Byrne> video, which is actually quite a lot more literal than you suggest- I just checked twice and counted the number of objects that came on screen the same time as the word for that object- is making excellent use of a left to right pan- which I believe you can also do with Windows Movie Maker.

However the Byrne video was obviously made by a real artist with a real graphic arts computer program.

It is a triumph of visual art and music video making.

I found the image choice for the amateur Joni video to be a little tame and predictable- though I'd say the same of some of my videos.

She has good proficiency with tempo and simply operating the Windows program properly though...

At any rate your selections have galvanized me to do some further experimenting with the program and with new ideas.

My web friend <sony magellan>, who is also expert with the Windows Movie Maker program, just sent me some more instructions on how to make the image transitions more techniallly smooth- it requires some patience and skill with the toggle- some "manual dexterity."

Interestingly, you may not be aware that one of the greatest Mexican graphic artists was of course <Manuel Dexterity>.

At any rate, acting on Sony Magellan's tips, I managed to make the image transitions in the <Tarantula> video with only a few shimmies-

Unlike the previous <Warrior> video which has so many shimmies in it I can't watch it anymore it's too maddening.

Ok thanks again for such material help <Dom>.

My goal is to make a video with the Windows Movie Maker that is as technically "smooth" as the Joni Mitchell one but considerably more interesting and less obvious.

Well to be fair, one that is also considerably more interesting and less obvious than the 13 videos I have made so far with the Windows Movie Maker.

Ok then.

BTW <Nimh> as a 1800-1900 range coffee house chess player, I agree TEN THOUSAND PERCENT with your points about chess style.

My point, essentially, is that where style is concerned there is no 'ladder of chess mastery towards GM-hood' -- there are plateaux, zones of stability. A different phenomenon to the plateau effect when an individual player stops improving.

<This is only anecdotal info, but it matches PRECISELY EVERY LETTER you just typed.>

Good god man I'm a sewing machine not an umbrella. You're talking <ginormous negative entropy> there ... it could tear the whole universe apart.

Exactly. The universe isn't expanding -- it's escaping. From us.

The universe can take care of itself, the gingery little beggar.

<acirce> is still hundreds of posts ahead, but the fact that I have over 10,000 posts consisting solely of the word "quack" will render my inevitable victory Pyrrhic.

Not that I care. In battle, you win or you lose, unless it's a draw.

As in the immortal, and superb, Peter Gabriel song about British gang wars, "The Battle of Epping Forest"-

Timely-

A review of <Chien Andalou> is always in order, and you've reminded me of the very fine Dylan allusion to this film in his <immortal, and superb>, "Visions of "Johannesburg"

Why not. If different styles of wrtiting can be described and analyzed, why shouldn't it be the case with chess too?

Oops. The mile-long escalators are found in *tube* stations. Metro stations have escalators 1.414 Km in length. Ca plane pour moi.

I'm glad you saw the inevitable reference to <A Dog and a Bog>.

My esprit-de-l'escalier moment came just now when I recalled that Jethro Tull, in Passion Play, had provided a universal remedy and catch-all response to people visiting one's forum to say "Happy [insert festival]!"

It's:

<And who comes here to wish me Well?
A sweetly-scented angel
Fell.>

An alternative is <I do not like thee, Doctor Fell> but I would never be so impolite or indeed caustic.

Speaking of <caustic>, does he know they're onto him?

<When all of your advisers...>

Pic: still from <Advise and Consent> showing Charles Laughton as a US Senator.

<... heave their plastic ...>

Pic: a still from <Animal House - the Regurgitated edition> in which Beavis O'Brian, world vinyl LP eating champion, demonstrates his secret method...

<... at your feet ...>

Pic: anything from any Pedro Almodovar film, with high heels.

<... to convince you of their pain ...>

Pic: scene from *The Madness of King George, the Third of his ilk* in which George writhes in agony from porpoise disease while a doctor looks into his bag and says "Hmmmm, your majesty. Laudunum? ... or Leeches?"

... taking us up to G-6, father of Her Present Majesty. My main complaint is that they *still* go on calling it the UK or United Kingdom, when it is clearly a Queendom.

And I remain convinced that the only thing keeping Scotland, Sark, and other peripheral bits in place is the awful prospect of living in <FUK> or Former United Kingdom.

By 2900 rating I meant 'the least required level of the accuracy of play sufficient to attain rating of 2900 in the end of 2000s'.

Techically more precise would have been to say 'average expected error of 0.050', but it would have been too confusing for the majority. I use ratings because they are easier to understand.

<if you take the absolute highest rating achieved, as in the CG statistics, then Carlsen at 2810 is already the 4th strongest player in history. And Aronian and Ivanchuk and Morozevich aren't far behind. All, I think, have bested Fischer's 2785, or got very close to it.>

Being rated as the fourth strongest in history doesn't indicate he is the fourts strongest player ever either in <absolute> or <relative> terms. As you yourself understand very well, the inflation has made such comparisons fruitless. Chessmetrics ratings are more suitable for comparing relative dominance. And no rating system can describe <absolute> strengths throughout ages. This is something only computer analysis is able to do.

<This is a crazy conclusion, though I accept that the current crop of elite players is stronger in some senses than its predecessors.>

If by 'predecessors' you meant players from 20-30 years ago, then your claim is absolutely correct. But if we conentrate on players from 50-100 years ago, things become quite otherwise. Modern players surpass them in every aspect of game that can be learned.

<However, the extreme chessmetrical approach - putting Morphy on a par with Tal and Shirov - is even more ridiculous.>

Nobody has deliberately put anyone anywhere, it's the logical result of the way chessmetrics calculates ratings. Morphy had the third all time best rating gap over #2, whereas Tal's biggest dominance ever was just 27 points. So, in my opinion, chessmetrics didn't give any false information.

The reality is that Morphy was the most dominant of them and Shirov has the highest level of play.

<I have a third option. Rating inflation is real, and so is a genuine overall increase in technique.>

Correct, and I agree with you, but instead of 'technique', one should read 'level of play' or 'accuracy of play'. Technique is just one aspect of play among many. In the middlegame and openings, technique is not important.

Rating inflation has nothing to do with increasing level of play. but you already knew it, I presume?

<But players from the past were capable of incredible feats of raw calculation - in some ways, deeper than anything a human does today. Perhaps it freed up their brains for calculation?>

Analysis with Rybka has shown to me that this is just a myth with no grounds. Modern players are far better tacticians and calculate deeper.

Surely, having less theory to upload into one's brain cells, permits more efforts to go into tactical skills. But all 19th century players were essentially amateurs from modern point of view, and they started serious study of the game much later than today. Receiving money doesn't make the professional, attitude does.

<Leading, somewhat paradoxically, to occasional deeper combinations despite a lower overall playing standard.>

Since time controls at that time were much less restricted, I cannot argue. Yes, true, occasionally.

<Whatever the case, the past and present seem to be incommensurable.>

Taking into account the eval differences between move choices #1 and #2 is quite trivial for me. The bigger the difference, the less accurate the play is.

What makes analyzing tactical players more difficult and its results underrated lies in the fact that such players know when to concede the accuracy of play in favour of creating the play of their opponetns more difficult.

Computers are incapable of seeing if an inferior move is justified or not.

<It's a legitimate quest, but quantifying the results is just as difficult.>

No wonder, everything one hasn't paid attention to is difficult to comprehend. It's true that trying to extract trustworthy conclusions from computer analyses is difficult, but it's certainlt worth trying. Like everything in this world.

<FWIW, I think Tal's enduring popularity is a result of his high % of noncomputer approved moves. Which just goes to show you something or another.>

Thanks! A pity there aren't many those who appreciate what I write. What's your style of play like?

<60% observation, 30% analysis, 10% fantasy. On another day I might change the percentages.>

what about today? :)

<My point, essentially, is that where style is concerned there is no 'ladder of chess mastery towards GM-hood' -- there are plateaux, zones of stability. A different phenomenon to the plateau effect when an individual player stops improving.>

Of course, attaining GM-strength isn't the same as style of play, but what I actually said was that style varieties diminish with each time one gets closer to GM-hood.

<It might be described informatically as the predictability of a player's play in general, not just individual moves. Chess entropy, in a sense.>

I did was talking about player's play in general. And what does one's play in general consist of, if not individual moves? How one chooses his moves has an effect to his style and level of play.

Your computer method and system is vastly more complex than anything I attempted. I no longer have a copy of the software engine I used: but it did not provide evals for subsidiary lines. It could display its 'current best line' along with a (nonstandard) eval (and no info on ply depth etc). As I recall, forcing it to show 'next best move' was the only way to examine other lines. In fact, this was relatively sophisticated for 1986-88, when engines were sold as sparring partners for games, not analytic tools.

Around the same time, I was present at a tournament organized by Tim Harding (then a player and book author, later editor/publisher of Chess Mail) -- he rounded up as many different engines/programs as he could, and ran a 5-round Swiss between them and a number of human players. These were restricted to a maximum ELO of 1700 'to give the computers a chance'. One engine almost won the event, but lost to a clever swindle by a junior player in the last round. Harding wrote it up for Chess magazine (UK).

I estimated my Psion's rating at roughly 2000, but that was then -- I'd rate it as less now, and in any case I think ratings are increasingly unreliable, for the type of reasons you cite. I could beat it with 'normal' play - no anti-engine stuff - at tournament time rates, but only if I was careful, obviously. Like any engine, it quickly refuted tactical errors or speculative moves. I'd guess it had an effective ply depth of about 12-ply, reachable in about five minutes. (In blitz, it fell into cheap traps and was easily mated).

I promised a quibble. This may be simply semantics, but the word 'technique' is used in various ways. One (narrow) sense is endgame technique, ie knowledge of (or ability to calculate) the winning line in a known won position.

A slightly more vague sense talks of 'GM technique', where no single attribute is required. In this sense, the word 'technique' can be redundant - it just describes a GM being a GM.

I would still argue that there *is* a type of technique which applies to openings. It is not line memorization -- it tends to occur in closed and 'flank' games, where many transpositions are possible, yet one can find a novelty appearing before move 8. Some lines in the Reti, for example.

The player with a 'feel' for that opening -- a sense of its inherent strengths and weaknesses, knowing how best to develop pieces and apply pressure, is showing something that can be called 'opening technique'.

As, I think, are the IM/GM players who experiment with a new opening against weaker opponents, without preparing it. This has happened to me a few times, usually as black in semi-obscure lines of the French. The masters tended to play the opening slowly, working it out -- and invariably found lines that theory regards well. Of course one can hardly expect an IM or GM to blunder in the opening as white in a French -- but it was still impressive to see them avoid the danger of later problems and reach a good middlegame without prep. That's technique, isn't it?

A couple of years after Harding's event, engines were much stronger. Around 1989, I played in a normal open swiss where a computer (Mephisto) scored 5/5 and 5.5/6, nominally winning (it was not actually eligible for prizes). I remember watching the games with interest: a succession of good players, most over 2200, played suicidal wild tactics and lost. I don't know which surprised the onlookers more: the machine's apparent strength, or the weak play of highly-regarded humans.

Most players then had no experience of engines and took it as a challenge to out-calculate them. The widespread errors in classic games and the relative weakness of 'strong' players was only starting to be understood.

I cannot speak for its aesthetics or quality- but I can tell you that I know how to use the actual software program now.

I can't stand to look at any of my other videos except this one- i fact I feel like deleting them all.

Mind you I'm not saying this one is good- I'm just saying the image transitions aren't shimmying or jumping all over the screen anymore. Please be sure to watch this in "HQ" mode- you just click on the HQ button on the youtube screen there. If you don't, the photos look like crud.

Please let me know what you think of this one when/if you get a chance- and please don't hold any punches.

I need critique to improve.