OpenTal was a rebirth…
A rebirth of that cool, cigarette smoking 21-year old, with dark glaring eyes, supreme confidence and ultra deep concentration – the guy who left veteran chess grandmasters red-faced and shaken by the relentless fury of his play – reborn today as a chess computer program.
OpenTal is a chess engine designed to emulate the dazzling playing style of the Latvian chess genius Mikhail Tal as closely as possible.
This means daring attacks, terrifying piece sacrifices and whatever it takes to bring down the enemy king as quickly as possible.
So if you’re a fan of Tal, and who isn’t, OpenTal is your chance to challenge him, analyze your games with his help, and to truly relive the brilliance of the “Magician of Riga”
But how did it come about?
As you guys know, I’m a fan of the Rodent chess engine by programmer Pawel Koziol, and recently I’ve been working a lot with this engine creating around 8 personalities for training games, analysis and so on, to be released with the next version.
During an exchange of emails with Pawel, he mentioned that some of the settings for a “Tal” personality I’d been testing, played such beautiful chess, that it was worth making an entirely separate engine based on it.
I’d tuned it to play not only in Tal’s style, but to even choose the very same sacrifices that Tal made in his games, whether sound or not!
The result was an engine that was truly a reborn version of Tal – christened OpenTal (due to it’s also being open-source).
Want to see some examples? Keep reading.
Before seeing the following games, you might like to review this post on Mikhail Tal to get a refresher on how brilliant Tal was.
Then return and witness the uncanny similarity in how this crazy chess engine plays!
OpenTal makes 2500-2600 level engines (which would be 2650-2700 Elo) look as ridiculous as the real Tal made his opponents look.
Won’t it be an invaluable tool for your own training games, or for getting killer ideas in your favorite openings?
Let’s see a few more brilliancies…
And tell me if you can count the sacrifices in the following game.
Poor Johnny is a strong engine, but if OpenTal gets “his” type of position, the tactics are going to flow.
Look at the crushing combination which decided the following game.
The following game, in which OpenTal puts his own slant on the famous “Greek gift” sacrifice is another jaw-dropper.
To see an engine such as Alfil humiliated in such a way is very rare indeed.
The last game I’ll show you is against Tao 5.6, an engine which has a reputation for being quite aggressive in it’s own right, and is much loved by computer chess enthusiasts.
In this game however, just like Alfil in the previous game, Tao is completely humiliated like a beginner…
OpenTal gambits several pawns in the opening and then, just as Tao is preparing to castle, he sacrifices a whole piece to keep the white king in the center – the followup is brutal as well as pretty.
Of course, if you place OpenTal against Stockfish, or any engine over perhaps 2800 on the CCRL rating scale, he will be made to look silly – just as any human would.
OpenTal is designed to play like Tal, which means he will have a lot of human prejudices and weaknesses as well, and will generally succumb to the cold, calculating precision of regular engines.
And this is what makes the games posted above even more impressive – OpenTal is not supposed to win such brilliant games against engines – but he does!
Here are a few quotes from people who have been testing OpenTal in the still early days since Pawel released it.
“This is the most aggressive , suicidal , sacrificial thing I have played” – FossilHunter
“OpenTal still plays at about 2500-2600 Elo, despite sacrificing pieces right and left.” – Pawel Koziol
“I just played it. What a dammned beast. I am beginning to hate Norman…
What did you do with Rodent?
This Tal thing has defeated me, a strong expert player after all, like a child.” – Fernando
Want to challenge Tal and see what the fuss is about?
Please log in again. The login page will open in a new window. After logging in you can close it and return to this page.