Tomas Bragesjös French Defence and Caro Kann Defence games

French Defence

I play both sides of the French Defence. I tried it with black when I got an idea of ruling out some d-pawn specials by playing 1 .. e6.

I started to play French with black in late 2016 and my results has been excellet this far, I have even got several draws vs a local FIDE Master in both blitz and in rapid games in 2017. My first none blitz nor rapid games with it was in two team matches in 2017. I find the French Defence an excellent practical opening that works at all time controls but I prefers Sicilian Defence in over the board games.

Caro Kann defence

I have played both sides of the Caro Kann defence. However these days I plays it mostly with white.

As black I did not lose many games, I even won over two local FIDE Masters who both played double edged variations in blitz games but is difficult to play for a win in many lines. When playing it with black vs lower rated I did not always play best opening moves like early Queen exchange in order to get more inbalances.

In my experience Caro Kann is bit inpractical in blitz as well, particullary in advance variation, but good in games where a draw is a good result.

French Defence in Correspondence Chess

French is my main defence in Correspondence Chess. I have played various system like both Winaver and McCutcheon in team matches for Team Sweden and in rated class tournaments. I have even played different subsystems within Winaver since I played both Poisen Pawn and Petrosians b6 system and met the Winaver exchange variation. I have also played both Winaver and McCutcheon in over the board games too as well as Steinitz variation.

In a Correspondence Chess game I invented a novelty as black vs Wing Gambit in the opening that lead to a more or less forced draw (else black wins!) and my opponnet had played Wing Gambit for over 40 years and was surprised it existed.

Winaver with 4 e5 b6

Winaver with 4 e5 b6 is an excellet interesting choice for creative black players in Correspondence Chess since computers often miss evaluates the arrising positions in the opening fase. Many mainlines variations are evaluated more than 1,5 pawns better for white. But after every concrete accurate theory move played the evaluation drops for every move entered and suddenly the position is called equal.

Many of these unexplored lines are very tricky both positionally with hidden tactics as well and different engines evaluates the position compelty different regardless of depth.

5 a3 Bf8

This line can lead to very double edge play with chanses for both sides. But it can also lead to calm positional play in other lines. I have met several tame variations from this position not covered by theory and also hyper aggressive lines not covered by theory.

5 Qg4 Bf8 6 Nf3

However there are one trap in these relativly unexplored lines and that is the variation 5 Qg4 Bf8 6 Nf3. Since the databases probebly show that I have won games in the other mainlines (like vs 6 Bg5) and after I lost one game in this sideline I keept meeting it lots of times. I later also lost an other game in this line after a blunder in an equal middlegame.

There are two different authors, Marin (opening database) and Lakdawala (book) who advocated the b6 system. At first time I played a game in this line Marin originally did not cover immediate 6 Nf3 at all only when other moves are inserted. Lakdawala covered it and he went for traditional the main line 6 .. Qd7 but after 7 Bb5+ c6 8 Be2 Ba6 9.O-O Ne7 10.Rd1! is very strong as I learned the hard way where I got much worse quickly and lost some 30 moves later since Lakdawalas c5,that I played in a game are better for white if whites plays accurate.

The line it self is evaluted harmless by computers at first sight and have been played at elite levels but none critical play some moves further in. It is not a forced lose but it is a minefield where white is much better.

I found no other improvment than that I think black after 5 Qg4 Bf8 6 Nf3 black must play 6 .. Ne7 instead when 7 Bg5 is harmless becouse of Qd7 when 8 Bxe7 is met by h5. After 6 .. Ne7 whites knight is a bit missplaced at f3 and black is equal, and this was confirmed in several of my later corr game where I met this line and played 6 .. Ne7 and I won one game vs the player I lost when I played 6 .. Qd7 to and also even ended up with the better minor piece in one other game that became a draw and complety equalised fast vs several heigher rated players in several other draws while in two other games I made some inaccurate moves in tricky positions in the middlegame and lost one of the game but I managed to hold ther draw in the other game since we reached a tablebase drawn endgame where I was 2 pawns down and would have lost without tabelbases since the only move that held the draw was difficult to find.

Marin later updated hes Winaver repertour base to fix this error and we both reached the same conclusion that black is fine after 6.. Ne7.

Other none covered lines

Both authors has been lazy in other lines too. After the natural moves 5 Qg4 Bf8 6 Bg5 Qd7 7 0-0-0 h6 8 Be3 Ne7 I met the novelty and probebly computer top choice 9 Kb1 in two different games. However I equalised in both games after Ba6 10 Nge2 Nc6 and here they played two different moves and I won one of the games and the other game became a draw.

I also met 5 Qg4 Bf8 6 Be3 once but it is harmless and I equalised fast by after h5 7 Dh3 c5 8 f4 g6 and I even ended up with a slight advantage but it ended up in a draw.

Why I stopped playing Winaver with 4 e5 b6 and possible refutation

After playing many French defence games my opponnets became more and more booked up becouse I have more games and more in the variation in the databases. Thus opponnets I never met before and those I met before did not enter the computer missevaluted lines and played other better lines.

The concrete reason why I dropped Winaver with b6 French was the sideline 5 h4. Lakdawala did not cover it at all and apparently Marin did not originally cover it either. The line is also rare i databases, including in the Correspondence Chess databases.

Marin has a very light coverrage after an update but I met an improvment after Ne7 6 h5 h6 7 Qg4 Kf8 9 8 Nf3 Ba6 9 Bxa6 Nxa6 my opponent played the novelty and new plan 10 O-O where white has a slight risk free plus in every line. I did not find the right plan so I lost pretty fast. Maybee I am a bit depressed after the lose but I can not find any clear improvment.


The b6 Winaver is a risky opening for both sides since several lines are relative unexplored in most variations and thus lots of holes in repertour books (for both sides). Its like an ideal opening for hard working creative players. Computers has often no idea on what is going on in the opening fase.

I dont think that the opening is clearly refuted. I advise players to explore the line themself and learn the opening and only then draw there own conclusions about this system, maybee the reader handles the opening better than me.

Winaver Poisen Pawn

Winaver Poisen Pawn is one of the most concrete tactical lines in the French defence. Most moves are played for concrete tactical reasons and not by general principles. Many lines leeds to forced tactical draws for both sides. I have not yet played enough games to draw any concrete conclusions about every sideline but I played it a few times in both over the board games and as well as in Correspondence Chess. I have tried it in a few tournamnets games and white can very easy force a draw in more than one line so this variation is not recommended to play it unless a draw is an acceptable result.

I will play in a Winaver Poisen Pawn thematical event in 2021 to learn more about the opening, it will also be the last thematical event I will play in ever and instead focus on international open events instead.

French Defence in practical play

Over the board

At both club level and in over the board tournaments and in over the board team matches the exchange variation is the most common line to face, I met it in more than 75% of my over the board games. I do not understand the logical since black equalices easy and I even have a slight plus score with black but it usually leads to a draw. I belive that the line will become even more common since WCH Magnus Carlsen plays this as white and two serious repertour books recommends it, both with some new ideas in mind. Advanced is second most common where I met the dubios Milner Barry Gambit more than I expected. After that both Nd2 and Nc3 are equally common and I have also met KIA and also some odd looking lines like Wing Gambit a few times.

Correspondence Chess

I have also reached French via 1 d4 e6 2 e4 more than I expected, including in Correspondence Chess. In my own Correspondence Chess games I met 3 Nc3 most of the times and after that 3 Nd2 and 3 exd5 only once and 2 Nf3 once and 3 Nf3 (blunder) once.

I can also add that in over the board Chess, many players who play exchange vs French often also plays 2 c3 vs Sicilian defence.

In thematical tournaments at ICCF

Start yearEventPointsResult
2017WSTT/2/17/3 - Caro Kann - Panow Variation, B137/12: +3 =8 -14th in section with same points as 3rd .
2019WSTT/2/19/Final - French, Teisha Gambit, C023,5/6: +2 =3 -13rd place .
2021WSTT/2/21 - French Winawer, Poisoned Pawn, C183/6: +0 =6 -0Started 2021-03-01 .
Chess viewer is from