How These Pages Work
Othello Pages
2005 US Open

Home Page
Othello Blog
Othello Gateway Site Map
Othello Links
Othello RSS
nbrowser
Ben's Blog Posts
Eddie's Blog Posts
Ed's Blog Posts
Randy's Blog Posts

The F.A.T. opening has become one of the most studied and played openings in the last few years. Fellow OthelloGateway bloggers Ben Seeley and Edmund Yiu played major parts in the study and advancement of this opening, and are known for using it in important matches. This, of course, is nothing less than a ringing endorsement of its potential and playability. F.A.T. is dynamite, and white in particular has to be very careful with it. This article is somewhat of an extension of my last one, "Toeing the Line”, as diagonal control plays a big part in the nature of the opening's volatility.

On the http://www.reversi.se/forum/forums.php message board, Thai player Thammatorn Kraikokit posted some examples of some unusual and effective x-squares he played, much like the "seeley-esque” example I described in "Toeing”. Thammatorn clearly has a solid grasp of the winning elements of the F.A.T. opening, as he displayed in the game he posted (applet above), against a fellow Thai player he named as "Taey”

ey makes an error with f3 instead of e2. E2 threatens access to c5 while denying access to h7. Now, Zebra lists this as going from a draw to -4.24. After 27.h7, let's break away from the actual game to look at what is supposedly best, 28.h3 29.h2 30.e2 31.c7, shown below.

32.d1 is listed as best, but still leads to a pretty solid black win. A very likely human move here is 32.c5, which is listed as -6.96. 33.c2 34.b3 35.b6 36.b7 is now a 26-disc loss (36.e1 is best at -22)! It's somewhat amusing to analyze from here, as 37.g3 is at first shown as being only +2.xx, and while it upgrades to +4.xx it shows "+26 at 99%”. This is a sign of how deceptively strong this position is, as the short-term analysis shows a fairly close game, while the final solve shows the opposite. To drive this point home further, let's look at another human-style variation on this. Play out 34.d1 instead of b3 in the above line, followed by 35.b6 36.b4 37.g3 38.b3, shown below.

Here Zebra looks at 39.b2 with an initial evaluation of +6.69. Upon final analysis, it is all the way up to +46! To show just how powerful black's c7-h2 diagonal control is, 39.b2 40.b7 41.a8 is a wipeout (and the only win after 40.b7), as white has no way of getting access to b8, and by extension no way of obtaining any stable discs.

Back to the original game, after 27.h7 Taey tries to break the e3-d4 mini-diagonal with 28.e2, but it is too late. Thammatorn replies with 29.e1!, gaining more central control and keeping white off of c5. After 30.c2 31.c7 32.f1?? (these question marks are actually somewhat unfair, as the recommended 32.f2 33.c5 34.b7 is still -24!). After 32.f1 black can easily take 33.g1, as there is no danger of 34.d1 35.c1. 34.c1 clears the 3 row, and 35.g3 is safe as it controls its diagonal. Now white plays 36.c5, which he has been striving for, reaching the position below.

Thammatorn wrote in his post that he did not play the natural-looking 37.f2 here because he was afraid of a stoner trap. This surprised me, as I saw 38.g2 39.b3! as unstoppable. What I think he was really looking at was after 37.f2 38.g2 he wouldn't be able to play the move he really wanted to – b7! So he played 37.b7 38.a8 39.f2, and white has no access to b8. Taey is now losing by 14, and 40.g2 takes it down to -32, and Thammatorn plays the rest perfectly while Taey makes two more minor errors. In fact, 37.b7, while being very clever, is the only mistake Thammatorn made in this game, going from +30 to +14. To break away from this fascinating game one more time, lets take a look at what black actually was afraid of, 37.f2 38.g2., pictured below.

My instinct of 39.b3 is indeed quite strong, but it is not best. What is? 39.h3!! Once more this shows the power of the c7-g3 diagonal, as if white plays 40.h1, 41.d1 leaves white with only one move, 42.g7. If white responds to 39.h3 with 40.g7, 41.d1!! 42.h1 43.h8! still controls the diagonal, as would 40.g7 41.d1 42.b7 43.h8 44.h1 45.b8. All of these lines insure a mass of black discs in the southeast and big wins at the end.

In addition to Seeley, Yiu and Kraikokit, Canadian Master Jacky Fu is also a well-known F.A.T. advocate. In 2002 he won the Canadian Championship against Tim Krzywonos using it, which was actually my introduction to it as being a strong opening. It came as little surprise that he played it again in the first game of the 2004 Canadian Championship finals against Krzywonos, transcripted above. While Krzywonos gives black the uneven edge on the south as Taey did in the previous game, Tim plays 20.e3 instead of Taey's 20.e8, and both are good.

players are playing strongly. Here Fu deviates with 29.c5, and Krzywonos plays the correct 30.h3 31.h2. I'll break away from the actual game here also, as this is a very close and interesting position. Best here is 32.g3, and both sides are controlling important diagonals, as shown below.

Black owns the c3-f6 diagonal, while white controls d3-g6. Black wants to gain access to h7, while white wants to beat black to that punch with g7. This is tough for black to accomplish, as 33.f1 34.b3 regains white's diagonal while crossing black's, and 33.e1 34.a4 does the same. It seems there's no way for black to win this particular battle (although both 33.f1 and 33.e1 are very close), so maybe some other trickery is in order. 33.b6 34.c7 and my first instinct is 35.b8 which regains the c3-f6 diagonal but fails after 36.a4, as black has no effective way of both preventing white's g7 and gaining access to h7. So, again black cannot fix his uneven edge, so 35.f1 makes at least makes the threat, and 36.g7 attacks the east edge. 37.e1 38.a5 39.f2 40.a3 41.b8 creates the position below.

The potential swindle here is 42.b5 43.a4!! and white's threat to the east edge is doomed. If black plays h8, white has no wedge at h7, and if white plays b7 black gets a quiet h7 with h8 to follow. The only way for white to win here is 42.b3 43.a6 44.b5 45.a4 46.b2! While black eventually gets the free h7 swindle, white can control the c6-f3 diagonal and the southwest is less playable for black, which adds up to a narrow white win.

Since white may see the potential swindle with that 44.b5 45.a4, and may not see the winning ways of the following 46.b2, he may be tempted to play 44.g2 instead (above), which leads to another example of effective diagonal control. 44.g2 45.b5 46.a4 47.b7! 48.b2 49.c2 50.a8 51.h8 52.h7 53.a7! 54.h1 55.a1 and white has no immediate access to g1. Most of these lines are beyond human vision, but they show the flipside of the Kraikokit-Taey game; while only a couple of reasonable errors sent Taey on a brutal downward spiral, Fu's 29.c5 deviation leads to an extremely close and complicated game. The element of diagonal control dooms white upon his mistakes, and may still save black when he strays from optimal play.

The secondary point of this tangential analysis is that 32.g3 leads to a close position, while Krzywonos played 32.f2? (above). The problem with this move is twofold; white's c6 makes 33.g3 very playable, as flipping d3 does not provide any good c2 move (as it would if c6 were black), and that g3 move also keeps control of g3 and g4 for black, threatening a 6-4 formation on the west as soon as h7 is played. A third problem can be seen a little further down the line, as white has no effective way of gaining access to g7 while preventing black's h7. For example, 34.b5 can be met with 35.a5! or 35.b3 (which flips f7 and makes white's g7 useless), and 34.c2 may gain that g7 access but gives black deadly free moves to f1 and c7. Again, black's deviation leads to a close position, while one real error sends white into a tailspin. Fu played the correct 33.g3 and Krzywonos responded with the best 34.b3. After 35.h7 36.b6 37.a3 38.c2 (38.g2 is -16 at this point) 39.c7 40.a5, the position below is reached.

While 41.a6 is best, Jacky played the striking 41.b7!, which is similar to Thammatorn's move except white has absolutely no access to a8. Also while Thammatorn's 37.b7 was a drop from +30 to +14 (after which 40.g7 was the -14 for white), Krzywonos' 42.a4 is a drop from -14 (also with g7!) to -30. These numbers may not be coincidences worthy of Ripley's Believe it or Not, as the positions stem from the same opening, but I still find them entertaining. The position after 42.a4 is seen below.

The problem with 42.a4 is now Jacky plays 43.b2!, controlling both main diagonals, and Krzywonos has no way of accessing either in a profitable manner. From here Fu only makes one error in the late endgame, going from a possible 55-9 win to 52-12. After Krzywonos won the second game, he prevented Jacky from playing F.A.T again by playing No-kung instead, which worked out much better.

Much of Othello skill is recognizing patterns, and these two games show a distinct pattern with this particular F.A.T. extension; black's diagonal control, strong uneven edges and potential for unusual and strikingly effective x-squares make this explosive opening particularly dangerous for white.