enter image description here

Is not bishop to e2 better? Why does the kingside has to be weakened like this?

Thanks everyone for your answers. You all argue that the knight must remain on f4 so h4 prevents kicking it out. However, I just found another game in which h4 is played and this time there is no keeping the knight on f4. Could someone tell me what else happens here?

enter image description here

  • 1
    Can I vote +3 to this question?
    – thb
    Commented Jun 22, 2018 at 13:35
  • @thb Of course you can
    – Maths64
    Commented Jun 22, 2018 at 17:44
  • @Timetraveler please, give a link to the game, otherwise it's very time consuming to find the game for analysis.
    – lenik
    Commented Jun 23, 2018 at 15:23
  • Sorry for that. The link of the last game is: lichess.org/caIfDMgT/white#10
    – Maths64
    Commented Jun 23, 2018 at 15:37
  • 3
    Please don't add on to the question with a totally new question. The second diagram is a totally different position.
    – Ywapom
    Commented Jun 25, 2018 at 16:47

5 Answers 5


You're trying to analyze the very equal position with no significant tactical potential. Depending on minor tweaks of your engine settings, you may get a few "good" moves, including Be2, Ba3, c5 -- all approximately equal.

Computers at lichess have limited resources, hence they spot only obvious stuff, and have a limited precision when evaluating moves.

If you move h4 -- you're down about 0.1 pawn, I would not lose a sleep about that, Be2 is good enough and you should not worry what stupid engine says =)

Disclosure: I'm actually running one of the lichess analysis boxes, so I took a deep dive into the details. Here's what I found:

  1. From your screen shot, you're running analysis on your local iPad, which is slow and may take some time to arrive to the good result. It would be much preferable to run the server side analysis, much faster and more precise.

  2. I have found the game in question among your games and analysed it using various engines available. Most of the engines recommend c5 as the best continuation for White. However, for the lichess analysis we use a very conservative settings of Stockfish ("Analysis Contempt" is turned Off), this changes the focus from the active attacking moves to the "safe" defensive ones. And h4 is one of those nice safe moves.

As I have said earlier, it all depends on the minor tweaks of the engine parameters, the score difference is negligible and h4 is no better and no worse than a few other moves -- play what you like and don't worry about what computer says.

Answering the second part of your question, here's my analysis:

enter image description here

Bd3 seems to be "the best", with h4 and Be3 sharing the second place. Actually, while analysis was running, 4 or 5 top pretenders constantly swapped their places, sometimes h4 was on the top. Probably, if I run the analysis deeper, there will be a few moments when any of the top 5 moves will be declared "the best".

All in all, the analysis is not exhaustive, and it does not guarantee to give you the "absolute best move". The main purpose of the analysis was to spot your mistakes (8. h3? was clearly a mistake, losing 0.5 pawn), and propose something better, not necessarily the best.

Here's the analysis @ depth 24, h4 is way below.

enter image description here

  • 1
    Engines also assume opponents make the GTO play (at least, as best as they can calculate) - but it doesn't take long to play with an engine to realize the difference between a human move and an engine move. You see it a lot when playing long clock games and people have lots of time to go coddle their engine.
    – corsiKa
    Commented Jun 22, 2018 at 17:28
  • 4
    @corsiKa what's GTO ?
    – lenik
    Commented Jun 23, 2018 at 6:15
  • Was the engine intentionally tuned to be defensive or why was this configuration made? I would have expected the engine to be tuned for maximum strength.
    – usr
    Commented Jun 23, 2018 at 10:59
  • @usr the engine tuned to the "maximum strength" is a very broad definition. There are ways to play against ELO1500 players, which differ from 2000+, which, in turn, differ from the settings used for computer-to-computer matches. The current setting is a bit conservative, but this might prevent an unusually high scores for very active, tactical positions where a bit of risk usually pays off. IANAL (I am not a lawyer), it's my personal opinion, which might differ from the "official" point of view =)
    – lenik
    Commented Jun 23, 2018 at 15:20
  • @lenik. Thanks for answering. But I still cannot see why h4 is any better than h3. With h3, I wanted to prevent the knight from going to g4
    – Maths64
    Commented Jun 23, 2018 at 16:21

As mentioned, it's all about keeping the Knight on f4. That outpost needs to be maintained and ...g5 is stopped by h4! Yes, g4 is weakened but to quote Bobby: "You have to give up squares to get squares".

Also note that if Black plays ...h6 hoping to play ...g5 you can play h5 and that knight is not going anywhere.

Piece activity can be more important that weak squares or even weakened structures. You can go over the Kasparov - Short world championship match an see how Short took on doubled pawns, etc. but was compensated by good pieces.

  • 2
    But why is that important to maintain the Knight on f4? Let's say 10. Be2 g5, 11.Nd5 it looks perfectly fine for White.
    – Akavall
    Commented Jun 23, 2018 at 2:30
  • @Ywapom. But still, with h4 you are not ensuring that the knight will remain on f4. I mean later on, one knight can move to e6 or g6 to exchange your knight on f4.
    – Maths64
    Commented Jun 23, 2018 at 12:50
  • @Akavall Whites Nf4 is his best active piece, he doesn't want it made to leave so easily. Black will have to make a lot of maneuvers to try and get rid of it, which allows White time to improve his position.
    – Ywapom
    Commented Jun 25, 2018 at 16:52

With h4 you prevent g5, which would kick your knight.

  • 3
    h4 seriously weakens g4 though.
    – ggorlen
    Commented Jun 22, 2018 at 17:58

As a novice, it's hard for me to see why 1. Be2 g5 2. Nd5 is putting White at a major disadvantage, but I do feel it puts White on the defensive. Black has gained space and potentially a kingside attack, but may be overextending. White's knight is centralized, and black may trade it away leading to a centralized pawn structure for White with a potential target pawn aiding the Black's light bishop and weakening White's. White needs to protect their center while anticipating heavy kingside pressure in terms of pawn pushing backed by rooks.

Compare that to 1. g5, where White seems to be restricting Black's expansion. This may allow White to develop into a stronger attack without worrying about the kingside pressure.

I'm unsure about responding to g5 by moving the knight to the h-file. I don't see any benefits to White's position there.

  • One thing you're not mentioning is that white currently has space and so should try avoiding pieces. After Be2 g5, Nd5 Be6, white is forced to trade the central knight. Commented Jun 22, 2018 at 19:34

I am writing this answer in order to explain my thoughts about the second position I provided. I have come to the conclusion after playing similar positions that h4 is played in order to weaken my rival’s kingside. That is why they have to play right away h5. Otherwise, my h pawn will have to be exchanged by the g pawn, opening a file for my rook. Although admittedly this does weaken my own kingside, there seems to be no danger and probably, I will eventually castle queenside. The sinequan reason for playing h4 though is that my rival has played g6 in order to fianchetto his bishop and so a weakening can be induced in my opponent pawn structure. In case someone has concluded something different feel free to correct me.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.