A tribute to Touw Hian Bwee (1943-2025)

The great problemist Touw Hian Bwee of Indonesia has passed away on 12 December 2025, at the age of 82. He became a Grandmaster for chess composition in 1984 when he was 40, the only person from Asia ever to have acquired the prestigious title. Touw (his surname) primarily composed orthodox two-movers, though his output also includes some top-notch longer problems. His composing style can be described as classical, with a predilection for strategic ideas like batteries and pins that make his sophisticated works relatively accessible.

In 2008, A Collection of Chess Problems by Touw Hian Bwee, edited by Milan Velimirović, was published – a wonderful anthology downloadable from the British Chess Problem Society site. The book is divided into five chapters: (1) traditional (single-phase and block problems), (2) thematic tries without changes, (3) thematic tries with changes, (4) multi-variations changed play, and (5) three- and more-movers. The first four parts reflect the development of two-move problems from traditional to modern (generally increasing in complexity), and showcase Touw’s mastery of the form across this broad spectrum. The three selections below illustrate some of these different categories.

Traditional two-movers include single-phase problems that focus on the post-key play, and block problems with interesting set play that gets changed by the key. The latter type is represented by the first diagram, a complete block where all legal black moves are already prepared with mating responses. Thus 1…Se2 2.Bc2, 1…Ke2 2.d3, 1…Ke4 2.Bc2, and 1…f2 2.Bc4. White can attempt to preserve the zugzwang, but 1.Be5? Ke2!, 1.Bf2? Ke4!, or 1.Ba4? f2! falls short, while creating threats with 1.Re1? (2.Re3/Bc4) Se2! or 1.Qxh2? (2.Bc2) f2! also fails. The delightful key 1.Rda1! (waiting) abandons two of the set mates, but compensates by clearing the 1st rank for the queen and guarding d1. Two changed mates result: 1…Se2 2.Qb1 and 1…Ke2 2.Bc4. The remaining play is as set, 1…Ke4 2.Bc2 and 1…f2 2.Bc4. A fine mutate solved by a Bristol clearance key that increases the queen’s utility.

Modern two-movers usually contain thematic tries that complement the key phase. Try-moves naturally occur in many problems, but to be thematic, they should share certain motifs with one another or with the key. In this example, seven tries and the key are all executed by the e4-knight, to produce a knight-tour. All moves by that piece threaten a diagonal battery mate, 2.Rf-any, but with one exception each involves a subtle weakness that Black can exploit. 1.Sf6? Sf8!, 1.Sg5? Rh2!, 1.Sf2? Rg1!, 1.Sd2? Rxe2!, 1.Sc3? b2!, 1.Sc5? Kb1!, and 1.Sxg3? Rxg3!, These tries show not only unity but diversity, as their various errors provoke seven different refutations by Black. The key 1.Sd6! succeeds by avoiding interferences with the rook: 1…Sf8 2.Rxf8, 1…Rh2 2.Rh5, 1…Rg1 2.Rf1, 1…Rxe2 2.Rf2, 1…b2 2.Rf3, and 1…Kb1 2.Ra5. Three extra variations are 1…Sc5 2.Rxc5, 1…Se5 2.Rxe5, and 1…Sf6 2.Sxf6. A terrific knight-tour problem that also incorporates nine distinct mating moves by a rook.

Try-play problems, following the modern trend towards greater complexity, may involve one or more tries that introduce their own variations, different from those occurring after the key. Changed mates are a standard way to achieve such a contrast between the virtual and actual phases. Here shifting the d3-knight will threaten 2.Rd3, and the piece has two candidate moves that form different batteries with the bishop and the queen. The first option 1.Sxc5? is the thematic try; note how the move guards e4 and releases the f6-knight. Five variations follow: 1…Rbd8 2.Sxb3, 1…Sc1 2.Sb7, 1…Rgd8 2.Sxg4, 1…Sf2 2.Qxf4, and 1…Sd4 2.Sd5. (It’s a pity that 1…Sxc5 2.Bxc5 is not a real variation because the self-pinning black knight doesn’t stop the threat.) The try is defeated by 1…Sd2!, which gains a flight on e2.

The second option 1.Sxf4! is the key; now the knight controls e2 and frees the b2-rook, besides aiming at d5, a potential mating square. The same five defences seen after the try are playable, but their mating replies are all strikingly changed: 1…Rbd8 2.Rxb3, 1…Sc1 2.Bxc5, 1…Rgd8 2.Sxh3, 1…Sf2 2.Sg6, and 1…Sd4 2.S4d5. Unlike 1…Sxc5 in the try phase, the by-play 1…Sxf4+ 2.Qxf4 is real because of the black check, and this shows a mate transference relative to 1.Sxc5? Sf2 2.Qxf4. (The post-key 1…Sc1 2.Bxc5 would be a mate transferred from 1.Sxc5? Sxc5 2.Bxc5 if the latter had been forced.) Two attractive pairs of battery openings are found among the eleven variations rendered in total.