Some info and a conversation with Valentin Albillo

Main (Dutch) page | Back to computer page | Correspondence Chess page

These pages will be divided in two big sections: one, that looks back to Valentin Albillo's work, and the other that looks to more positions which cause problems for computer programs to solve. For the moment only the first part (to continue where mr. Albillo stopped) is in the start-up phase.

In doing tests myself, I have wondered "what is the correct solving time of a given position?" For study-like problems, you can take the first time the program finds the move (and sticks to it), or you can take the moment where it gets a fail high (in other words: the moment it sees a win), or you can let it calculate for days, and give the last displayed ply (of which the evaluation will be probably the highest of all previously displayed solutions). If you think about it, you could say: the "fail high time" is the correct solution time.
Probably right, but the other two times are pretty important too. The time where the correct move is displayed is interesting, because that would mean that the program would play this move when given a short time (e.g. like in an actual game, where you get an average of 3 minutes per move). Even when not seeing the solution, after the opponent has played, it may actually play the whole combination, as it discoveres the correct continuation gradually, as with every full move played, it becomes more "clear" what the way to the win is.
And the "last" displayed time can be interesting to see what the limits of the program / hardware are. Especially from position 71 in this suite onwards, the "last" time becomes interesting.
On top of this, there are also mate problems, where the shortest mate needs to be found. Here the correct solution is easy: the shortest time in which the correct mate is found. And if the program does not find the correct mate, but another, longer one, then the "first" time of the shortest mate announcement counts.
Well, you will see what I have used, as you look at the pages...
What do I expect from people who want to contribute: every kind of honest info that fits in the tables you will find on the pages
If possible: give the first time / evaluation / ply depth / nodes / move sequence the program finds the correct move and sticks to it and for positions that are not mate problems: please let the program run for some more minutes/hours, to see how the evaluation gets better. It would be nice to solve some of the not yet solved problems.
Among my favourite programs I do not have (but have had the experience to experiment with), are MChess5 and Rebel6. Results of those programs are especially welcomed. Also I'm not a wizard in configuring freeware in user-friendly GUIs. In order not to obtain weak test results due to bad configuration, I give you, the experienced freeware user, the opportunity to contribute and defend the cause of freeware programs. With the arrival of Arena, even for dummies like me, testing freeware engines has become possible without too much configurationing, so you will see the names of Ruffian, SOS and others peep in here and there.

After a while, I realized something is missing. Nowadays, testing becomes even more difficult, as programs have developed a "memory", which enables them to store positions in which a drastic change of evaluation occured. This means positions which were considered good, suddenly do not seem to offer any salvation, and positions which seem lost, but suddenly a "great" move turns the tables around. This makes it very hard to verify or reduplicate results of one program on one computer. I experienced such learning proces first hand, when I changed my AMD K6-2 for a K6-3, and the program (Hiarcs) solved one position way faster than with the old processor. Taking all these "obstructions" into account, I will try to be as objective as possible. If the reader has corrections to offer, please do not refrain, they're welcome.

Declaration of integrity (message to Valentin Albillo):

Dear Sir,

Some five years ago, I sent you some test results from other programs (I recall Virtual Chess and Fritz were among them). I understand that due to limited time, you could not publish them. That is why I thought of putting up a sequel site to your original site. It has all the test positions, but will contain only solutions from other, newer programs on new hardware - for evaluation sake. Of course I will acknowledge you as the originator of the test suite, and if I make citations from your site, I state that it is an excerpt from your site. The whole project (I can now see how much work you put in this) is just in the startup phase, but I would like your opinion on it; if it is OK for you that I make such a site. It is not my intention to tag my name on your original suite, nor abuse it, I will present it as your idea and - as I have some interesting positions I want to show besides your suite - I will keep it separate completely from my test suite.

Your site has kept my interest for computer chess alive ever since I first visited it - and it even helped me in my earlly chess development. The purpose of this site is to keep the interest in your test suite alive and up to date.

I hope you approve of this project.

Best wishes,

Yves Surmont

I had asked on the newsgroups the email address of Valentin Albillo - and some time after that I got an email of Mr. Albillo himself:

Hi Yves,

Best regards from Valentin Albillo:

First of all, thank you very much for your interest in my humble computer chess web page, it's much appreciated that after all these years without being able to update and modernize it as I would have liked, there are still such enthusiastic and worthwhile persons like yourself that do appreciate my efforts.

I just happened to find your post to a newsgroup asking if anyone knew my e-mail address (since the one at the former Mygale is no longer active), and then searched the web for references about you. I promptly found your web pages at: and was very moved to see your "sequel" to my site and reading your words. I will now comment on your "Declaration of integrity (message to Valentin Albillo)", point by point:

"Dear Sir, some five years ago, I sent you some test results from other programs (I recall Virtual Chess and Fritz were among them). I understand that due to limited time, you could not publish them."

Yes, regrettably I found myself back then in a series of circumstances of all sorts that prevented me from maintaining the page, as I fully intended to do. It was unexpected, and I couldn't do anything about it, which was particularly sad to me, because a number of gentle contributors like yourself had sent me many very interesting results, for faster and better programs, and even some managed to solve a number of the most difficult positions.

So, I had a vast amount of material, plus my own results, book reviews, articles, and analysis, but I found it impossible to get enough free time to update the page. However, I did my best to not let it die, and every few months made a trivial change here and there, so that Mygale-Multimania-Lycos would detect activity and didn't delete my site from their server. I succeeded, and to this date, both the original multimania site at: and the one at Geocities are both still alive, if pretty static.

"That is why I thought of putting up a sequel site to your original site. It has all the test positions, but will contain only solutions from other, newer programs on new hardware - for evaluation sake."

It seems a very good idea to me, as simply mirroring my site would be pointless, as all results are for rather obsolete software (though most comments still apply even to modern programs). Your brand-new results are certainly very interesting, and the first thing I did after finding your "sequel" site was to print them to study them at leisure while at home.

"Of course I will acknowledge you as the originator of the test suite, and if I make citations from your site, I state that it is an excerpt from your site."

It's certainly extremely kind of you to do so, and I thank you for it. You certainly can quote as much or as little as you want from my site, you've got my full, explicit permission.

"The whole project (I can now see how much work you put in this) is just in the startup phase, but I would like your opinion on it; if it is OK for you that I make such a site."

I see that you've noticed the extreme amount of work (and thus, free time) necessary to create my site on the first place, just to write the HTML code, not to mention getting the results. If you did print my whole site on paper, you'll have seen that it is a very thick stack of pages, nearly almost book-size. I hope that you will be able to find the free time I lack and will update your site with ever interesting contents.

As for my opinion, yes, it's absolutely OK for me the idea behind your site, the fact that it took mine as an starting point, and your proposal of developing it even further from that basis.

"It is not my intention to tag my name on your original suite, nor abuse it, I will present it as your idea and - as I have some interesting positions I want to show besides your suite - I will keep it separate completely from my test suite."

Again, thank you very much for being so honest and kind, it's certainly rare in this modern world around us to find persons with such an integrity as yourself. Thanks for it.

"Your site has kept my interest for computer chess alive ever since I first visited it - and it even helped me in my earlly chess development. The purpose of this site is to keep the interest in your test suite alive and up to date. I hope you approve of this project."

I certainly do, I'm really indebted to you for your words, for your work, and for your utmost integrity, and I'm extremely glad that you've found my site useful and likeable. As you'll see with your own site, there's nothing more rewarding for all the hard work invested than to discover someone actually does likes it and finds it useful, so that all the effort wasn't in vain.

Finding your site and reading your words has done wonders for me, I feel most honoured to have inspired such an effort on your part, and I certainly wish you all the best with your site, with your chess club, and expect to hear from you and visit your site frequently in the future.

Best regards from Valentin Albillo.

P.D.: I would be very obligued if you would keep my e-mail address private, without either publishing it and/or telling to any other persons. Thanks in advance.

Of course, I will not give Mr. Albillo's email address to anyone. My reply to his mail was this:
I guess the real start of the computer chess page (with lots of results on the chess positions) will be around may-june [ ... ] I plan to publish your comments on my "declaration" online (small separate page), so it is clear for everybody that I have your permission to do this. Of course, I will not publish your email address, nor send it to anyone else.

Hi again, Mr. Surmont:

Thank you very much. I look forward to visit your chess site very frequently in the future, and meanwhile, please receive my best wishes for your site's success, I'm really very obligued to you for your worthwhile and kind efforts.

Best regards from Valentin Albillo.