THE LOGIC AND BEAUTY OF THREAT CORRECTION
by Valery Shanshin, in collaboration with David Shire and Marjan Kovačević
Preface
The main reason prompting us to write this article together is to try to increase the mutual understanding of our differing composing perspectives and to seek a collective common ground.
As an illustration of this process we have chosen a thematic field that was initiated long ago but is one that is still being interpreted ambiguously by problemists from different countries. A symbol of such theoretical differences is No 2 below, which has been perceived in the composition world, in the West and in the East, from opposing viewpoints.
While working on the article, we decided to divide it into three interrelated parts. Part 1 discusses primarily the historical and theoretical aspects of Threat Correction; Part 2 analyses the various opportunities for developing Threat Correction which have been adopted by composers over the decades; and Part 3 demonstrates recent successes in this area and outlines prospects for the future.
Through our discussions we discovered that we started with different approaches to understanding the logic of Threat Correction. All three of us have benefited from this interchange and we hope that a consequence might be an expansion of our creative horizons.
The authors would like to express their gratitude to Anatoly Vasilenko (Ukraine), Vyacheslav Pilchenko (Russia) and Wieland Bruch (Germany) for their active support and assistance in collecting materials and also to Andrey Frolkin (Ukraine) for his translation services. We are pleased that the editors of The Problemist and The Problemist of Ukraine have arranged for simultaneous publication, in English and Russian respectively.
Readers will have their own opinions about these articles. The best outcome will be achieved if constructive criticisms are sent to nihsnash@mail.ru, kovacevic.marjan@gmail.com and beatrice.shire@cantab.net. The most interesting observations can then be reflected in the later articles.
Part I
The Making of Threat Correction
Since the late 1920s, following the discovery of “White Combinations” (choice of first move/key) and “Mating Tries” (choice of mating move in a variation), the choice of move has become one of the key trends in the evolution of the #2. In the early 1930s, Soviet problemists working in the orthodox #2 genre focused on the development of the concepts of “Secondary Threat” and “Extended Defence”. Combined, these subsequently were labelled as Black Correction (BC). The “colour” of the correction generated a mechanism with at least two variations: the correction of a move by a black piece taking into account White’s response to its random move. Later years saw the appearance of #2s with an ordered choice of key spread across at least two phases, its characteristic element being a correction of the first move by a white piece taking into account Black’s refutation of its random move. This mechanism was named analogously as White Correction (WC).



In one of the earlier problems created along those lines – No 1, a random move of the white bishop contains a strengthening effect: additional guard of the e4square. This introduces threats of 2.Sb3# and 2.Sd3#; these are defeated by the pinning refutation, 1…Qс1! The wB seeks to achieve a better result by preclosing the c line with 1.Bc2!? However, in doing so it crosses the critical d3square, thus rendering 2.Sd3? ineffective. The remaining threat of 2.Sb3# is easily parried by 1…Qg1! It is necessary to incarcerate the queen in the corner; 1.Bb1! deprives Black of both effective defences. Again, the only threat is 2.Sb3#. Among the variations (1…Qd4 2.Sd7#, 1…Qс3/Qxb1 2.Qxс3#, 1…B~ 2.Qxа1#, 1…Sb3/Sс4 2.Rxе6#), one is of interest to us: the selfblocking 1…Sе4 is met by the transfer of the discarded threat; 2.Sd3#. This feature is both noteworthy and important.
The content was extraordinary for its time with tries 1.Qа4? (> 2.Sd3/2.Sd7) Qс3! and 1.Qb4? (> 2.Sd3/2.Sd7) Bс4! featuring White Combinations and double threats. This was supplemented with extended key correction (demonstrating a higher degree of correction), which is currently known as “Tertiary White Correction”!
Extended Threat Correction was probably first implemented in an accentuated manner (without a common sidethreat), with complete adherence to the order of strengthening and weakening effects of the first moves, in No 2.
In the diagram position of this problem, White can deliver mate neither by 1.Bc3 nor by 1.Rd3 because the bK will flee to e3 or c4. 1.S4~? opens the line of the wQ to e3 with the threat 2.Bс3#) but this is insufficient on account of 1…Rxс5! An additional attacking resource is brought to bear by 1.Sd2!? This guards c4 to set up a secondary threat of 2.Rd3# but eliminates the primary threat of 2.Bc3. However, 1...Bc5! subtly pins wRb3. The knight’s leap to d6 looks even less promising: the line of the wQ is opened to e3 and an extra guard over c4 is provided ... but the bK is provided with the c5flight thus rendering both primary and secondary threats ineffective. Nevertheless, 1.Sd6!! (>2.Sb5). The additional guards of e3 and c4 become relevant after Black’s selfblocking defences; 1...Bxc5 2.Bc3 (2.Rd3?) and 1...Rxc5 2.Rd3 (2.Bc3?). A pattern emerges: Try? (>2.A) 1...x! Try? (>2.B) 1...y! and Key! (>2.C) 1...x/y 2.B/A. Now we recognise this as the Hannelius theme. The problem’s accomplished logical form incorporates considerable tactical content: the opening and closing of many lines for white pieces, further emphasised by the byplay 1…Sе5 2.Sе6#.
In Britain, No 2 was initially viewed as an exercise in threat correction. Later, when reversal themes were thriving in the USSR, the same problem was regarded as the pioneer example of the Hannelius theme outlined above. Over the last decade this logical combination is referred to as Tertiary Threat Correction (TTC); more on this later...
The next problem was published in the same year as No 2, but in No 3 the same author was aiming for a record number of TC phases. 1.S5~? (> 2.Qа5#) 1…Bе1!; 1.Sс3? (>2.Sа42.Qа5?) 1…Sс7!; 1.Sс7? (> 2.Sxа62.Qа5?) 1…Sxс4!; 1.Sxd6? (2.Sb72.Qа5?) 1…d3! and 1.Sxd4! (> 2.Sxb32.Qа5?) 1…Bxd4 2.Qа5#. A point of great significance is the return of the primary threat as a variation after the modest key move. This transfer does not feature in the other phases.
The 50s and the 60s were a period when themes featuring multiphase change of play were developing rapidly in the #2 genre. Remarkably, the innovative ideas relating to choice of move did not go unnoticed in the West. One can see this from a number of eloquent examples presented hereafter.



In the diagram of No 4 there are set variations; 1…Sd3 2.Sg3, 1…Bb5 2.Qf3, 1…Sg~ 2.Qе3. A move by wSd4 will create the threat 2.Sd6. However, a random move by the knight would be erroneous – for example, 1.Sе6? (> 2.Sd6A), leads to new variations; 1…Bb5 2.Qxg4(В), 1…Bf3 2.Sg3, 1…Sе3 2.Sс5, but this is refuted by 1…b5! It is necessary to abandon the initial threat and play 1.Sf3! (2.Qxg4В and not 2.Sd6?A) with changed play 1…Sd3 2.Sd6A, 1…Sg~ 2.Sd2 and 1…Kxf5 2.Qh7. Again we emphasise the transfer of the primary threat, a case of the currently popular “change of move function”. There is a further eyecatching try of secondary logic; 1.Sb5!? (> 2.Sbd62.Sfd6?) with the changed mate 1…Sе3 2.Sс3. Unfortunately this is doubly defeated – 1…Kxf5! and 1…fxg5! – but this serves to prove that the opportunities offered by the mechanism have not been exhausted.
A later problem, No 5, demonstrates a threat correction made more paradoxical through the presence of interesting WC try play. 1.S~? (> 2.Qd5) 1…Bс6!; 1.Sе7!? (>2.Qd5) 1…Sxb6! (2.Bс5?) and 1.Sb4? (>2.Qd5) 1…Sс7! (2.Rb4?) Then comes the unexpected 1.Sxс3! (> 2.Sе22.Qd5?) 1…Bxс3/Saxc3/Sbxc3 2.Qd5/Bc5/Rb4 and 1…Kxс3 2.Qd3.
As with the problems mentioned earlier, No 6 was ahead of its time! This example provides a clear insight into the differences between TC and the threat avoidance idea substantiated by Y. Sushkov twenty (!) years later. 1.S~? (>2.Re5A/2.Be4B) is obviously refuted by 1...Sxe6! White can strengthen his hand by providing for 1...Se6 but by so doing a weakening is introduced through a competition between threatmates “according to Sushkov.” 1.Sg6!? (>2.Re5A/2.Be4B?) Sxe6 2.Se7 but 1...Sc6! is sufficient (1...Sf7 2.Qb7). 1.Sc6!? (>2.Be4B/2.Re5A?) Sxe6 2.Se7 but 1...g6! The white corrections and the black refutations occur on the same g6 and c6 squares. This idea was given the peculiar name of “Caprice” by the same Yuri Sushkov at the beginning of the 70s! The key phase presents the familiar threat correction; 1.Sxc4! (> 2.Sе32.Rе5A?/2. Bе4B?) with 1...Kxc4 2.Qb3. The primary threats return after selfblocking captures with mates separated by dual avoidance; 1…Rxс4 2.Rе5A (2.Bе4B?) and 1…Sxс4 2.Bе4В (2.Rе5A?) This same technique was demonstrated in No 2.



The concept of No 6 was used as the basis for a number of wonderful #2s, of which 7 is one. Set 1…dxc5 2.Rd1A and 1…Rс6 2.Bе6B. These mates become threats; 1.Sс~? (>2.Rd1A/2. Bе6B) but 1…Sс5! 1.Sе6!? (> 2.Rd1A/2.Bе6B?) 1…Sс5 2.Sс7# but 1…Bxе5(а)! 1.Sd3!? (> 2.Bе6В/2.Rd1A?) 1…Sс5 2.Sxb4# but 1…dxe5(b)! The key is flightgiving; 1.Sе4! (>2.Sxf62.Rd1?2. Bе6?), and now 1…dxе5(b) 2.Rd1A(2.Bе6B?), 1…Bxе5(а) 2.Bе6В(2.Rd1A?) and 1…Kxе5 2.S4c3. The Hannelius pattern again! The set variation, 1…dxe5 2.Rd3, is worthy of attention. Its significance in the interrelation of phases will be discussed in Part 2 of the article. Meanwhile we will note that, considering this strong defence that unblocks d6, White’s actions amount to secondary correction with changed play!
In the 70s, composers started studying the rich opportunities provided by the blending of various popular themes and these explorations continued successfully in the succeeding decades. Against that colourful background, those rare attempts to construct TC problems almost passed unnoticed. For example, in their day the next two problems were essentially considered to demonstrate the Caprice theme whereas they are also excellent examples of TC. In No 8, three white pieces enact analogous threat correction sequences. 1.Rb~? (>2.Sb6) 1...c5! 1.Rxc6!? (>2.Rc52.Sb6?) Sxc6 2.Sb6 but 1...Ra5! 1.Sd~?(>2.Rd7) 1...e3! 1.Sxe4!? (>2.Sf62.Rd7?) Rxe4 2.Rd7 but 1...Rf4! 1.Re~? (>2.Se7) 1...e5! 1.Rxe6! (>2.Re52.Se7?) Bxe6 2.Se7. In each sequence Black defends by unblocking a flight, the threat correction cedes the flight to Black immediately (the Caprice element!) but Black subsequently blocks the square again to admit the primary threat.
In No 9 a random move by wRe7 is represented by the Bristol move, 1.Ra7? (>2.Qb7) This is refuted by the familiar flightvacation idea, 1…е5! Consequently the threat is corrected by 1.Rd7!? (>2.Rxd62.Qa7?) and the refutation is unchanged, 1...e5! However, the motivation is now lineopening for bRh6. The key is a striking move that “combines” Black’s defensive resources. 1.Rxе6! (> 2.Sf42.Qа7?2. Rxd6?), leading to 1...Rxе6/Sg6/Kxe6 2.Qа7/Rxd6/Qf7. The Caprice element of No 8 is equally evident here.



The concept presented in No 2 is almost exactly repeated in No 10, but with the extra nuance of a choice between possible refutations to the tries. In the diagram position of No 10 Black has two strong defences; both 1...Rxb3 and 1...Rxe3 secure a flight for the bK. However, after the random first move 1.Sh3? White aims at guarding the c4square in the threat 2.Qf7A and so the only refutation is 1…Rxе3(a)! In the second try 1.Sf7!? White guards the d4 square in the corrected threat 2.Qxd6B (2.Qf7?A) With perfect analogy the sole refutation is 1…Rxb3(b)! Finally White outwits Black: 1.Sxе4! (2.Sf62.Qf7?2. Qxd6?) 1…Rxb3(b) 2.Qf7A (2.Qxd6?), 1…Rxе3(а) 2.Qxd6B (2.Qf7?) and 1…Kxе4 2.Sс3 with a surprising selfpin of the stubborn rook.
In order to implement the harmonious synthesis of TTC + Hannelius, the authors of No. 11 found a fresh mechanism for choice of first move with a wB as the thematic white piece: 1.B~? (> 2.Qс3) 1…Sd5!; 1.Bb4!? (> 2.Sе52.Qс3?) 1…Qd5!; 1.Bd4! (> 2.Sе32.Qс3?2.Sе5?) 1…Qd5 2.Qс3 (2.Sе5?) and 1…Sd5 2.Sе5 (2.Qс3?). The presence of the variation 1…b4 2.Sе5# in the first try suggests a direction for expanding the thematic complex.
Tactically interesting is the sequence of play in No 12. This is introduced by the set play 1…fxe5 2.Sf6. After 1.Sе~? (> 2.Sxf6) Black tries in vain to repulse this threat by unpinning bSe4 – 1…f3 2.Sе3. The refutation is 1…Bd8!, since the set response to any move by the black bishop – 2.Qxс6# – is no longer effective. By changing the direction of his attack, White himself unpins bSe4 with 1.Sf3!? (> 2.Rxd42.Sxf6?). The primary threat reappears when the released knight defends; 1…Sd2 2.Sxf6(A)#. However, 1…Bxс5! refutes because once again the wQ cannot mate on c6. White may also unpin bSd4 with 1.Sd3!? (>2.Sxf6/2.Sxf4) but 1...Sc2! secures a flight for the bK. The key is the familiar strike; 1.Sxс6! (> 2.Sе72.Sxf6?2. Rxd4?) when Black is helpless. 1…Bxс5 2.Sxf6 (2.Rxd4?), 1…Bd8 2.Rxd4 (2.Sxf6?), plus 1…Kxс6/Rxc5 2.Qb7/Sxb4#. Again TTC with Hannelius (as in No 2, No 10 and No 11), but with the transfer of the initial threat in the second phase.
During the recent decades of the modern history of chess composition interest in this thematic area, which is narrow but exceedingly interesting, has expanded noticeably in the West primarily due to works by British problemists. However, to most of their colleagues living in exUSSR countries this field still looks rather exotic. The time has come to review the logical structure of TC, to clarify the terms associated and to seek agreement on definitions.
In international terminology, seconddegree Threat Correction has long been referred to as Threat Correction (TC).
In The Twomove Chess Problem: Tradition and Development, a book published in Great Britain in 1966, the authors B. Barnеs, J. Rice and M. Lipton presented No 2 as an example of a twomover featuring TC. In particular, they emphasised the reappearance in postkey variations of the threatmates associated with the tries.
Nearly 40 years later, Chris Reeves published in the July 2004 issue of The Problemist an article entitled TTC: Ideal or Illusion? In it he presented a more detailed analysis of the logical structure of Threat Correction and proposed that its third degree be referred to as Tertiary Threat Correction (TTC). The same No 2 was quoted as the firstever implementation of TTC and again its content was reviewed, but this time for a different reason. Reeves specified the following: (1) The transfer of the primary and secondary threats to postkey variations should be regarded as a prerequisite for TTC. (2) The analogous transfer of the primary threat to a variation mate in the second phase of a TTC problem, while being logically desirable, should not be mandatory.
In connection with the latter specification, it is reasonable to refer to a rendition of a TTC #2 with the reappearance of the primary threat as a secondphase variation mate as the complete form of TTC.
In the relatively recently published Encyclopaedia of Chess Problems: Themes and Terms (2012), the authors M. Velimirović and K. Valtonen declared the complete form to be mandatory for TTC. However, considering the pioneering No 2 and the 2004 definitions, we suggest that the simple form should be seen as sufficient for TTC implementation.
Accordingly the following definitions are proposed:
TC is the transformation of the threat generated by the random move of a white unit to a different threat generated by its correction move. Consequently there is a choice of threats based on the avoidance of a dual threat. Phase 1 carries a primary threat A whereas in phase 2 there is a secondary threat B (not A?) – this is simple TC. Complete TC is to be found where A occurs as a variation mate in phase 2. Examples of this latter form include No 3, No 4, No 5, No 6, No 7 and No 8.
TTC is an extended transformation of the threat generated by the random move of a white unit to different threats generated by successive correction moves and spread across three phases. Thus, Try 1? (>2.A); Try 2? (>2.B2.A?) and Key! (2.C2.A?2.B?) 1...y/z 2.A/B. Examples of this simple form of TTC are No 2, No 9, No 10 and No 11. In Boyer’s problem we discover the pattern, Try 1? (>2.A); Try 2? (>2.BA?) 1...x 2.A and Key! (>2.C2.A?2.B?) 1...y/z 2.A/B. Thus No 12 demonstrates complete TTC. It should be realised that the strict sequence of threats and avoided threats across the phases is a prerequisite for implementing the patterns outlined above. Undoubtedly composers will seek further developments of this theme but for the sake of clarity any clouding of this logical framework is best avoided.
In conclusion, we would like to suggest that the logic of TC acquires great thematic significance when it embraces more than two phases and that this requires carefully considered expression in order to be successful. This issue will be addressed in the second of our articles.
Valery Shanshin (Russia), Marjan Kovačević (Serbia), David Shire (Great Britain)
November 10, 2014