Reflections about judging

I wrote these reflections as a planned introduction to my judgement of Probleemblad 2001 Fairies that will be ready soon. But then I found out it would be better to publish them first in online medium at my disposition - CCM - and perhaps to receive some comments from readers... I also included below a few comparison problems that I searched during judgement - they have their own value and the comments sometimes are and sometimes aren't connected to the judgement.

I am very grateful to the organizers of any composing tourney for inviting me to be their judge. It shows their confidence in me and I try my best to make awards as good as possible. The last year I even have to refuse a few invitations as I thought I will have enough work. And yes, I did quite a lot of judging recently (and the entries for FIDE Album 1998-2000 didn't arrive yet!). Of course, some my compositions were judged by others too, sometimes positively strangely (I mean they were placed higher than I expected!). That's why I would like to explain a few my ideas about judging and personal preferences in general. Hopefully it will be interesting at least to some readers,

Judges of tourneys usually write in the introductions of their awards about criteria they've used for judging as well as about the influence of their personal preferences to the award. Yes, it is undeniably true that personal preferences play an important role in ranking the problems. Even the best experts can value and rank a same set of compositions in a very different order. The regular and the most important examples are FIDE Albums (or WCCI). Sections are judged by really experienced persons, usually holding the official title of International FIDE Judge. But their outcomes are often very different. How is it possible that a composition can get 4 points from one judge (it means in his opinion it should surely be in the Album), while other judge gives it only 2 points (it means in his opinion it is good, but probably not enough for the Album)? It is possible, I wouldn't blame the judges, I see why - and I guess you know the answer now too. Despite all more-or-less objective criteria for judgement, personal preferences of the judge play if not the crucial role, at least very important.

Why then there is so much criticism about various judgements? Let aside old well known WCCT reasons (giving high points to own country - it was pronounced only very silently even by the most courageous people in the press, it was always louder in private... but at least PCCC did something with it). One reason can be the breach of abovementioned more-or-less objective criteria. Among them, the originality is the least objective. Awarded compositions should be original. It may happen that judge simply doesn't know the precedessor - like in the problems by Castellari and Grevlund - that's why willing experts in the respective fields are asked to search for anticipations before an award is published (to name the most known - Chris J. Feather, Udo Degener, Harold van den Heijden, Zivko Janevski, Rolf Wiehagen, but there are many others) and there is a claim period after publication of a preliminary award too. Even then some anticipations remain hidden...

Also the articles about judging criteria are far from uniform when it comes to originality issue. What is original and great idea to one judge (or any chess composition amateur), can be harmless and outworn to another one. It depends largely on the experience of the judge, knowledge of the field, etc. - well, you see, this criterion is partly objective, but partly subjective.

But (taking into consideration all discussion about various awards with other composers) I see the mismatch of personal preferences as the main reason for disagreements with judgements. An example: a twomover judge has to decide between Lačný cycle problem with almost new mechanism with pawn-capturing key and theme H problem with flight-giving key and two changed mates. An expert in the one of the given fields (Cyclone themes or line combinations) would probably prefer the problem from his field. He knows it is perfect, original and it is very convenient to his personal preferences. But - what about the quality of key? Some judges don't care too much, some can't award any twomover with capturing key, even the most thematical... And now - imagine the situation when you have to decide among 5-10 compositions with various qualities and defects - depending on your own definition of the quality or the defect. It would be a miracle if two judges ranked 10 good problems into the same order. The same situation is in every genre, twomovers were taken only as an example.

(If you would like to see other example, look at the award of Mat-64 2000 - s# . Four best problems - according to the judge - are of 4 different styles: strategical, logical, bohemian and "letter". I don't say the award is bad, no, I rather agree with judge! But he wrotes very frankly that he was "not brave enough" to place bohemian higher despite it is the closest to his own style and he would like to see it higher.)

This everything is in tourneys usually evidenced by the solver's comments - not rarely the opinions even of the most respectable solvers differ very much. After all, chess composition is more an art than a sport, it's impossible to measure precisely what is nice...

It happens to me very often that I send two or more compositions to a formal tourney, really different compositions - and in the award they are ranked not according to my preferences (compare e.g. 2nd Prize and 2nd HM from JT Bunka 60, or 2nd Prize and 3rd Prize from Wola Gulowska 2000).

Or I don't believe my eyes reading the judgement... the ranking of various compositions is against my own preferences. Well then... some years ago it made dissatisfied, sad or even angry... but what can I do about that? Almost nothing. I just know some persons whose preferences are not compatible with mine and if I want to win something in the tourney judged by them, I have to compose and send there something they like (and I sometimes don't, haha). But as I'm getting older, I don't care so much about awards, I have enough of them. Moreover, I'm now regularly invited to judge good and important tourneys and it means that my personal preferences are more-or-less accepted and that my awards are not so bad. If I use in the comments some adjectives like "perfect" or "trivial", "mechanical" or "nicely motivated", it should be always taken as my opinion. Anyone can have his own opinion - and the time is the best judge. Widely reproduced problem praised by various column editors is in some sense surely better than any other awarded with better distinction. That's why I'm always so happy when any of my compositions is reproduced and positively commented somewhere...

I have a good luck that I'm not too specialized yet as a composer. I work in the field of tanagras (mostly longer helpmates with fairy pieces) as well as heavy twomovers and almost anything between this. I admire composers that are able to work better in any of my beloved genres, but perhaps even more those working in the fields that I just watch, like fairy moremovers. I don't think I'm the expert in any of the fields - but I'm in the know of the most of them. That's why I hope there are not many disappointed people after going through my award. It is highly probable that anyone would order the compositions differently...

My personal criteria are originality in the respective field (as far as I can judge it - here besides my experience the database of WinChloe is now of immense help), difficulty of a theme shown, intensity of use of the given fairy devices, technical tricks and general impression on me.
Umberto Castellari
feenschach 1973

1.Sg6 f8S 2.Sxf8(Sg1) Sf3#
1.Be7 f8B 2.Bxf8(Bc1) Bxe3(pe7)#
1.Qf1 f8Q 2.Qxf8(Qd1) Qxd3(pd7)#
1.Ra8 f8R 2.Rxf8(Ra1) Ra4#

In the Probleemblad 2001 there was published an interesting Circe h#2 by Jevgenij Fomichev and an improvement by Unto Heinonen. I'm not yet decided whether it will be included in the award, but during the search for anticipations (as I suspected there could be some and one of solvers pointed it is "an old hat") I found this very unexpected case. This composition and ...

h#2 (2+14)

Per Grevlund
3rd Prize The Problemist 1981

1.Sg6 f8S 2.Sxf8(Sg1) Sf3#
1.Be7 f8B 2.Bxf8(Bc1) Bxe3(pe7)#
1.Qf1 f8Q 2.Qxf8(Qd1) Qxd3(pd7)#
1.Ra8 f8R 2.Rxf8(Ra1) Ra4#

... this one are almost the same. I'm almost sure that Per Grevlund didn't know 8 years old problem as the judge didn't know it as well... It is probably due to very limited dissemination of feenschach in these days, limited communication and general belief that one can compose in fairy chess without dange of anticipation. It was common saying even in the years I was starting to compose, in the 90's of 20th century. Well, it is definitely not true in general, you have to know your field... I found this very special example only thanks to thorough search in WinChloe database - there are surely many other examples. Other one from the present tourney follows...

h#2 (2+14)

Steen Vestergaard
F286 Probleemblad 2001

1.Ra6 2.Rb6 3.Ka6 4.Sxc6(w) 5.Bxc6(w) 6.Ka5 7.Rxc6(w) 8.Ka4 Ra6#

The idea is clear, Black has to capture a few times to allow own king pass to the mating square. Some solvers commented positively, some didn't, but in my opinion it is clearly anticipated by ser-h#7 by Alexandr Shvichenko - and I realized there is the anticipation immediately after seeing F286 in the magazine as I studied Feenschach-50 award and (what is the most important) then I also wrote the page about WCCI for CCM - it makes one remember.

ser-h#8 (2+5)
Andernach chess

Uri Avner
Prize Nunspeet 2001

1.Sa2? zz
1...Bxd7 2.Rd2#
1...a4 2.Rb3#

1.Ra2! zz
1...Bxd7 2.Sd3#
1...a4 2.Sb3#
1...Rf1 2.Sxe2#
1...Kb2 2.Sb3#

This is another example of the process that happen relatively often. I talk about development of the idea - something rather positive.

Here we see the position composed in the limited time at the meeting of the Dutch Chess Problems' Friends Society. The idea is Mars Circe typical - fight for flights a4 and d7 guarded by wQ from d1. In the initial position there is a half-battery Bb4-Sc1-Rb2-Ka3. The try is flight-taking, the key is give-and -take. Then white waits for a black blocking and harmlessly closes one of lines d1-a4, d1-d7 (theme B2). There are 2 changed mates.

#2 (6+8)
Mars Circe

Uri Avner
Michel Caillaud

1st Prize Cheltenham 2001

1...Ra4 2.Bc2#
1...Rxd8 2.Bd3#
1...Rxh3 2.Bxh3#

1.Qc3! zz
1...Ra4 2.Sb3#
1...Rxd8 2.Sd3#
1...Rxh3 2.Sxh3#
1...Kb2 2.Sb3#

At the meeting of BCPS, very soon after Nunspeet, again in limited time, the idea is slightly changed with the help of the new co-author. The position becomes mutate. In set play White mates by battery Se2-Bb1-Ka3 and one variation without theme B2 is added (capture on h3). The key gives flight b2, that's why the set mates don't work anymore, but it also half-opens prepared half-battery and 3 changed mates follow.

+++ Composition In the Spotlight (CIS) No. 14 +++

Spotlight comment by Juraj Lörinc:

This composition was shown in Marianka 2009 by Frantisek Sabol as an example of excellent Mars Circe problem in his lecture dedicated to this peculiar fairy condition. (see bulletin Marianka 2009 - external pdf link). Actually I am not sure whether Frantisek have known the following Probleemblad problem further developing the idea of the present Cheltenham prizewinner, perhaps yes as both are contained in WinChloe database. What do you think: which one is better, the one with spotlight on or the next one?

#2 (11+7)
Mars Circe

Uri Avner
1st HM Probleemblad 2001

a) 1.Be4! zz
1...Bxd7 2.Sd5#
1...a4 2.Sc2#
1...f4 2.Sf5#

b) 1.Re2! zz
1...Bxd7 2.Kd2#
1...a4,Kb2 2.Kc2#

c) 1.Sg4! zz
1...Bxd7 2.Bd3#
1...a4 2.Bc2#
1...e3 2.Be4#

d) 1.Kxe2! zz
1...Bxd7 2.Rd2#
1...a4 2.Rc2#

However then, at home, Uri Avner managed to drill maximum out of the given scheme. 4x2 Zagorujko appears as a result of play of the half-battery Ba1-Kc1-Rb2-Ka3 and the created battery S-Bb1-Ka3.

#2 (11+9)
Mars Circe
b) f5 -» e4
c) = b) + g4 -» f5
d) = c) + g5 -» e2

Václav Kotesovec
Pat a Mat 2001

1.Gh1 Ge3 2.Gb1 Gb6 3.Gb7 Gb8#

1.Ge3 Gb1 2.Gb6 Gb7 3.Gb8 Gh1#

Formal cycle of moves is of a very unusual kind... white moves become black and vice versa. Václav worked a bit with this idea - he published besides Probleemblad problem e.g. this position.

h#3 (7+7)
1+2 grasshoper

Franz Pachl
3rd Prize StrateGems 2001

1.Ed5? A th. 2.Be5# B
1...Ec4 a 2.Bg6# C
1...Sf6 b 2.Sd4# D
1...Qxf4 c 2.Sxg7# E

1.Be5! B th. 2.Bg6# C
1...Ec4 a 2.Sd4# D
1...Sf6 b 2.Sxg7# E
1...Qf4 c 2.Ed5# A
1...Qxh5,Qxg5 2.dxe4#

By chance I was a judged StrateGems 2001 and I wrote about this problem:
"Djurasevic cycle 5/1 (according to Peter Gvozdjak's terminology) is and for long time will be something extraordinary. The mechanism is great. In try, 1...Ec4 defends by providing hurdle for Eb3, but unguards g6, 1...Sf6 guards directly, but interferes with Bg7, 1...Qxf4 captures threat unit, but blocks f4. In solution, 1...Ec4 guards directly, but provides hurdle for Ea3, 1...Sf6 pins Be8, but interferes with Ec3, 1...Qf4 provides hurdle for Ee2, but for Eh3 too. The paradoxical elements (key, key-threat and threat) are attained in a standard way, but the additional motives needed for other changes are nicely fitted together, note e.g. that key 1.Be5 closes Bg7 to d4 and c3, but activates Ec3 to g7, thus activates and deactivates the same line in different directions."

Franz Pachl was able to do this with the only fairy element, equihopper... but even more, he did also with grasshoppers only, in Probleemblad 2001!

#2 (13+12)
6+3 equihopper

Hubert Gockel
diagrammes 2000

1.Kh6! zz
1...Bd1+,Bc2+ 2.Bxg8#
1...Bxa2 2.Rxa2#
1...Bf7 2.Be6#
1...Be6 2.Bd5#
1...Bd5 2.Bc4#
1...Bc4 2.Bb3#

Hubert Gockel works with magnet theme with various fairy elements. Besides his Probleemblad 2001 problem he did e.g. #2 in Isardam or this problem in Beamtenschach. Usually moves are motivated by various line closings.

#2 (6+9)

Unto Heinonen
3rd Prize J.-P. Boyer MT 1988

1.Ra3? th. 2.Be3# A
1...fxg6(Sb1)! a

1.c5? th. 2.Rd3# B
1...bxa6(Bf1)! b

1.Be1? th. 2.Qxf6(Rh8)# C
1...Bxb4(b2) 2.Bf2#
1...Se7! c

1.Re8! th. 2.Rd8#
1...fxg6(Sb1) a 2.Be3# A
1...bxa6(Bf1) b 2.Rd3# B
1...Se7 c 2.Qxf6(Rh8)# C
1...Bxb6(Sg2) 2.Se2#
1...Bxb4(b2) 2.Be3#

Dombrovskis and a half (3 threat paradoxes) in Circe is one of Unto Heinonen's themes he tries for a long time. Here it is motivated disunitedly (guards of c3, f6 - guard of d3 from different directions - Circe selfguarding of wQ, check on 8th row), in 1st Prize Olympic Tourney Thessaloniki 1984-1988 it is more united by Circe selfguarding of wQ...

#2 (11+10)

Mircea Manolescu
Mat 1986

1.Be3? th. 2.Se1# A
1...Be1! a

1.Sf2? th. 2.Qe4# B
1...Se4! b

1.Kg5! th. 2.Rxg3(g7)#
1...Be1 a 2.Sxe1(Bf8)# A
1...Se4+ b 2.Qxe4(Sg8)# B

... while the mechanism in Probleemblad 2001 is the most similar to this two-element example - white King avoids selfcheck in variations.

#2 (12+6)

Comments to Juraj Lörinc.
Back to main page of Chess Composition Microweb.