CHESS#1274

GM Mathew Sadler recently pointed out with faux nostalgia, that circa 1990, it was entirely possible to analyse a position for weeks and still miss a huge tactical shot. On one famous occasion, Michael Adams set up an opening position wrong while preparing for a Candidates match against Viswanathan Anand!

Using engines and databases has not only eliminated such accidents, it has also led to improvements in human calculating ability. Humans have learnt how to do this better and there's an important change in attitude. Players expect to find hidden tactical resources in all situations. They are more concrete in approach, not relying on general principles. The average tournament player of 2017 is, therefore, a far more dangerous opponent than the 1997 version since he knows more tactical motifs and looks harder. 

Something similar may happen in Go. Go is way more complex than chess. The first four moves in chess can result in some three billion positions — in Go, that's over 16 billion. Last year, the AI, AlphaGo, beat two world champions, for the first time. One of them later studied the program's style and this led to a sharp improvement in his play. As off-the-shelf Go programs become available, it's a certainty that average playing standards will improve. 

In chess, the difference really shows up in defence. “Half-correct" romantic sacrifices are more likely to be refuted. Somebody sacrifices material for a dangerous attack. The defender has one narrow path to safety and possible victory and many losing options with limited time to calculate.

This was how everybody played back in the early 19th century before the first world champion, Wilhelm Steinitz, derived his principles. Even in the 1950s, Mikhail Tal, and later others like Boris Spassky and even Garry Kasparov back in the 80s, regularly won at the highest level with such "romantic gambles". The 21st-century defender will find the best defences more regularly. 

The teenaged Chinese genius, Wei Yi, has a penchant for this sort of romantic play. But he was out-calculated in a recent game by an elder statesman from the pre-computer age. 

Hoogeveen, Netherlands, is hosting two six-game matches. One features Vassily Ivanchuk versus Wei Yi (Ivanchuk leads 3:2) ; the second is Adhiban Baskaran versus Jorden van Foreest (Tied 2.5:2.5).  

In the diagram, BLACK TO PLAY, (White: Ivanchuk Vs Black: Wei,Yi , Hoogeveen Matches 2017, Game 3)  Black can try 17...Rxe3! with the idea 18.Kxe3 d4+ 19.Kxd4 ? Rc4+ and Qe5# mate — white will struggle to hold with the forced 18.Kxe3 d4+ 19.Kf2 dxc3. 

Wei tried to improve with 17...g5? 18.Bg3 Rxe3? Now the Bg3 controls e5. After 19.Kxe3, white won with 19...d4+ 20.Kxd4 Re5 21.Bd3 Be6 22.Qxb7 Kg7 23.Ne4 Nd5 24.Bxe5+ Nxe5 25.Rc5 (1-0). 
 

Devangshu Datta is an internationally rated chess and correspondence chess player