The chance tesuji is a form of play which transforms a game of strategy and tactics into a game of hazard. Principally possible in many games it is conspicuous especially with some Go players, possible through the high complexity of positions on a Go board.
Said players make a group of violent moves, each of them not trivially refutable, having no sense in itself, belonging to no strategy apart from the chance tesuji, but creating together a complicated enough position which is not evaluable for the opponent. Often only one move is necessary to reach such a position. The aim is to win by chance.
In the extreme this way of play results in a winning chance of 50:50, regardless of strength. Luckily it is necessary for successfully playing chance tesuji to at least learn the basic tactics. And the victim can reduce the impact by bettering his own tactical ability. Nonetheless chance tesuji is destroying the chance of playing true Go.
One form of this play is the speculative invasion. There is some overlap with trick play. Overly aggressive play sometimes results in similar positions. The distinction to an experiment is the aim to learn and therefor more coherence in play and situation. The same is true for simple errors.
The chance tesuji is one of the unrespectful bad habits used by people to much fixed on winning instead of playing Go. It sometimes is rated as appropriate for a tournament, but the distinction between tournament Go and true Go should then be clear.
togo: I invented this term after being deeply annoyed by way to much of chance tesuji, boredom tesuji and trick play happening to me online. The problem there is, that stopping playing or giving up disturbs the rating system. And "way to much" means that substantially more than half of my time was wasted. I tried to get something from these games, too, but they just stayed annoying and boring.
Gresil: I contest the summary labeling of the described technique as a bad habit. For one thing, I am reminded of what Calvin writes at Zone Press Park, specifically the bit about turning "finite" positions into "infinite" positions. The player who is behind in a game must seek complications. Who is to say that that is not a part of "true go"?
Eluusive?: This description is nonsense. As if any random style could result in a win 50% of the time.
Herman: For those who were as confused as I was by this position, here's the last few moves in the SGF file: WL[X.XXX] means White has X.XXX seconds left, BL means the same for Black. OW[X] means White has X overtime (byoyomi) stones left in this period (Canadian).
;B[hk]BL[77.01]OB ;W[gl]WL[1.858]OW ;B[hj]BLOB ;W[cc]WL[1.412]OW ;B[bj]BL[173.25]OB ;W[bi]WL[1.018]OW ;B[gi]BL[165.391]OB ;W[aj]WL[0.758]OW ;B[fg]BL[159.558]OB ;W[gh]WL[0.312]OWC[ Pyoveli [2d\]: so, it's fifty fifty Pyoveli [2d\]: G13 or H12 ] ;B[gg]BL[101.517]OB B+Time
Before the last black move, white has 0.312 seconds left for 1 stone. The position is such that there is a large black dead group whose death is absolutely vital to White's position. There is a group of three white stones that are part of the surrounding group keeping it dead, which have two liberties. (G13 and H12). Black can play one of those, and white will have to play the other. But White does not have time to wait and see which move Black will play, White must gamble one of those points and hope he guesses right (in the game, he guessed wrong).
MrTenuki: Actually, I would disagree with the assertion that the chances of winning are exactly 50/50. White could always tenuki, to be safe-- after all, losing 60+ points due to letting that Black group live is still better than losing on time immediately. Think about it this way: by picking between G13 and H12, White has a 50% chance of losing immediately and a 50% chance of generating an almost sure win. By playing elsewhere, White has a lower chance of losing immediately, but will also end up with a much worse board position. Whether this is better than the original scenario really depends on one's risk averseness and playing style...