London System & the Greek Gift: Bxh7+

London System & the Greek Gift: Bxh7+

Sep 12, 2018, 1:01 AM |

This is my 15th(!), and probably last (for now) post on the London System and related openings. It's been very instructive for me to write them and I hope you've enjoyed them (find the others here).

Today we'll look at another thematic idea...the Greek Gift sacrifice, which is one of the most recognizable combinations in chess: 


There's a ton of material available on the Greek Gift on For example, here are two articles by GM Serper (1 and 2), here's here's one by GM Naroditsky, and here's one by IM Silman. Furthermore, my fellow bloggers UAArtur and Simaginfan have written several very good posts on this sacrifice (see herehere, and here).

For book-length material, nothing can beat Jon Edwards's magisterial "Sacking the Citadel":


Edwards's book is especially fascinating since it looks at how the theory surrounding it has involved through Edwin Voellmy's original article from 1911, Znosko-Borovsky's The Art of the Chess Combination in 1936, and Vukovic's Art of Attack in 1959, before giving his own analysis. In what follows, I'll summarize a few critical points, but I strongly encourage you to check out the book for yourself.

* Greek Gift Theory

Let's start with this didactic French Defense position from Vukovic:

We've seen that white needs at the minimum a bishop attacking h7, a knight ready to come to g5 (as we will see, not necessarily before the queen), and a queen ready to come to the h-file, most typically h5. The three main lines are Kg8, Kg6, and Kh6 (the other two lines, Kh8 and a non-losing case of Qxg5, are much rarer, but have to be kept in mind). White needs to calculate all before going for it.

Now, Vukovic had postulated that for the sac to be correct or sound, White needs at least 2 extra assets out of the following:

  1. A secure pawn on e5
  2. A dark squared bishop (1+2 were present in our first example)
  3. h4 with the rook on the h-file
  4. A second knight
  5. A rook on an open file ready to lift to the 3rd rank

Edwards thinks that the 2-asset rule is a good rule of thumb for assessing a potential sac. However, he points out that there are situations where fewer assets are needed:

  • the White queen can come to h5 before the knight (This will be important in what's to follow...)
  • there's a rook on the 3rd rank
  • Black Rf8 is blocked in
  • Black Qe7 blocks the king's escape route
  • White has a knight already on g3 or f4.

Similarly, there are situations where more are needed:

  • B(lack) has a pawn on f6
  • B bishop or queen can reach the b1-h7 diagonal
  • B rook has already left f8 making room for the king
  • B has protection of f7
  • B knights can force one back to f6
  • B has a rook on the open f-file, especially f6
  • B has a counterattack on the e5-pawn
  • In the Kh6 line where White doesn't have a DSB, but has played h4
  • In the Kg8 line where W needs more than one move to play Qh5
  • White has a significant material deficit (thus allowing B to non-losingly play Qxg5)

Hopefully this gives you some food for thought about the potential complexity in deciding whether to go for the sac and can help you better analyze the examples in the above articles. However, we will next look at a case of an unsound, but incredibly powerful sac to see that it might be worth playing even in a speculative manner.

* Colle's "Unsound", but Powerful Bxh7+

The Greek Gift is most frequently associated with the French Defence where, with the pawn on e5, the enabling conditions are most frequently met. For example, in Edwards's book the French accounts for 98 out of the 309 games (that's 32%). However, Edgard Colle's (1897-1932) the Colle System is also partially built around angling for the sac.


I want to show here a classic game which shows clearly that sometimes, even when the enabling conditions are not met, the sacrifice can still be incredibly powerful.  If you haven't seen it you're in for a treat:

Whoa! Edwards's final assessment of the Colle game is instructive. He says that even though the sac was unsound in the sense that Black has a path to at least a draw, it would've been incredibly hard if not impossible for his opponent to find, as is evidenced by the fact that most commentators couldn't find it for more than 50 years after the game!

For those who prefer Mato (but he doesn't give the best line in Kg8):

Thus, sometimes, even if the sacrifice is not sound you should still consider going for it because it sets your opponents overwhelming problems.

* London and Bxh7+

The London System frequently features the Greek Gift, especially the enhanced version with h4 and the rook on the h-file. In fact, some of the most beautiful recent Greek Gifts have come in the London. We will look at three variations. However, before doing so, let me point out that the London Gifts are in some ways more special and easier to play than in other cases. The main point is that White almost always has an open diagonal or rank for the queen and thus he gives the second check directly with the Queen on h5. This means that the Black king will inevitably have to return to g8 (instead of the trickier g6), which considerably eases the attack.

1. The Most Basic Case: Black plays QGD style with Be7 and captures Nxe5

The first line where the sac occurs is when Black plays with Be7 and captures Nxe5, after which you capture with dxe5. This is an enhanced version with h4 and the rook on the h-file:

Here's Mato with two games in such lines:

A Variation: Black plays Nbd7 & Be7

There is a variation on the same theme where Black plays Nbd7 instead. The following game is a another London classic by Kovacevic:

Here's one of my own Daily Games which followed this game up until the sac, after which Black decided to give back the material in hopes of surviving. I did miss a brilliant win, but Black resigned nevertheless:

2. The Old Main Line: Black plays Bd6 and Qe7

The second line where the sac occurs is the old main line with Bd6 and Qe7. The story goes that this was discovered by someone on the ChessPub forum and then first used by Eric Prie in the following game from 2009 which unfortunately ended in a draw:

Most people who've read on the London have seen this sac in the following modern classic by Gata Kamsky against the reigning US Champion Sam Shankland (also discussed in the above Naroditsky article and by Sam Copeland here):

The best thing here is that even if Black reacts correctly by 10...Qxd7, White can simply take the draw as in the following game. Here the lore goes that Shaw mentioned that this line had been discussed by Prie in ChessPub to Brechin's teammate and then forgot about it, only to be promptly reminded of it in the game:

3. The Main Line: Black plays Bd6 and b6 and White pushes e4

The third line where the sac occurs is the main line with Bd6 and 8...b6, after White's 9. e4. Black has to screw up here a bit to make it possible, but it's very easy for him to do since the ensuing position doesn't look typical for Bxh7 at all: