Early Nh4 in the Exchange Lion?!

Engine analyses don't help me understand the plans in these positions. Anyone have a solid grasp of the lines here, especially those without short castling for Black?

I think 6.dxe5?! is premature, but 6...dxe5 7.Nh4! looks like a good followup to me.

Compare: 1.e4 d6 2.d4 Nf6 3.Nc3 e5 4.Nf3 Nbd7 5.Bc4 Be7 6.O-O O-O is the typical position of the Improved Hanham variation. These days white usually plays 7.Re1 or 7.a4. An older idea for white is 7.Qe2 c6 8.Rd1 Qc7 9.a4 h6 10.dxe5 dxe5 11.Nh4! with advantage to white. This has been known since the 1970s, but fortunately for black either 7...exd4 or 8...exd4 or even 9...exd4 gives adequate counterplay, which is why masters switched to 7.Re1. The point of all that is, if white can arrange it then d4xe5 ...d6xe5 Nf3-h4! is exactly what white wants.

I consider the plan with ...h7-h6 and ...g7-g5 to be unsound. Nevertheless it's interesting. I looked in Van Rekom/Jansen (2008) The Black Lion. With the 5...h6 move order they analyze only 7.Bxf7+, on pages 165-172. It seems they mostly prefer the 5...Be7 move order. I also looked in the ChessBase database and found two relevant games. These are not perfectly played, but you can check them with an engine just as easily as I can.

(1) Van Rekom/Jansen, pages 112-115 : 1.e4 d6 2.d4 Nf6 3.Nc3 e5 4.Nf3 Nbd7 5.Bc4 Be7 6.O-O h6 7.a4 c6 8.dxe5 dxe5 9.Be3 (a) Qc7?! (b) 10.Nh4! g6 (c) 11.f4! exf4 12.Bxf4 Ne5?! (d) 13.Nf3?? Qb6+ 14.Qd4?? (0:1) R.Morlock - M.Bergmann, Steinhaldenfield 2001

(2) database.chessbase.com : 1.e4 e5 2.Nf3 d6 3.d4 Nf6 4.Nc3 Nbd7 5.Bc4 h6 6.dxe5 dxe5 7.Be3 c6 8.a4 Qc7 9.O-O Be7 10.Nh4 (e) Nc5 (f) 11.Nf5 Bxf5 12.exf5 O-O 13.Qf3 Rad8 14.Qh3 Kh7 15.g4 Nfe4 (1/2, 54) P.Zagrapan (2194) - T.Liscak (2028), SVK-chT2 2003-04

(3) database.chessbase.com : 1.d4 d6 2.e4 Nf6 3.Nc3 e5 4.Nf3 Nbd7 5.Bc4 h6 6.dxe5 dxe5 7.Nh4 c6 (g) 8.a4 a5 (h) 9.O-O Qc7 10.Kh1 Nc5 11.f4 Be6 12.Qe2 Bxc4 13.Qxc4 Ne6 14.f5 Nc5 15.Be3 b6 16.h3 Be7 17.Nf3 Rc8 18.g4 Ncd7 19.Rad1 Bc5 20.Bc1 Rb8 (1:0, 38) Z.Runic (2427) - A.Jurkovic (2330), Zagreb op-A 2010

1. It seems white could play 9.Nh4! right away (not mentioned in the book), before black can play their recommended improvement 9...g5. After 9.Nh4! Nc5 (idea ...Bc8xNf5) 10.Qe2 is better than 10.Be3.
2. They analyze 9...g5!, which also prevents 10.Nh4.
3. About 10...g6 they say "The usual reply, which keeps the white knight from f5". This sort of note is why you shouldn't trust books by non-masters. Then they analyze 10...Nb6! 11.Ba2 (11.Bb3, not mentioned, is a tiny improvement) Ng4 12.a5 Na4 leading to equality.
4. They actually spend two pages analyzing the complications after 12...Qb6+. I won't get into it, because 10...g6 is just a bad move, but I will point out 12...Qb6+ 13.Kh1 g5 14.e5! "White obtains a crushing attack ..."
5. After 10.Nh4 it's the same position as the Morlock - Bergmann game, where 10...g6 was played and 10...Nb6 was analyzed.
6. 10...Nc5 is logical, covering f5.
7. Certainly not 7...Nxe4? 8.Bxf7+. The only try to "punish" 7.Nh4 is 7...g6 8.Qe2 Be7 9.a4 Nxe4? 10.Nxg6 Nxc3 (a typical desperado which is also quite common in the Sicilian) 11.Nxh8! Nxe2 12.Bxf7+ Kf8 13.Bxh6#. Of course black doesn't have to get mated, but 9...Nxe4? just loses.
8. Again not 8...Nxe4?! 9.Bxf7+ Kxf7 10.Qh5+ Kg8 11.Nxe4. But there are a bunch of moves which are close to equal, possibly a bit better for white: 8...Bb4, 8...Bc5, 8...Qa5, 8...Qc7. e.g. 8...Qc7 9.Qe2 Nc5 10.O-O Ne6 11.Qd3 Nc5 12.Qf3 Bg4 13.Qg3 Be6 14.Bxe6 Nxe6 15.Be3 Bb4=.

First of all, thank you for your very detailed reply! For some reason, I haven't bothered checking the thread in a while, but now I have to try and not TL;DR this, since you were only so kind to provide so much info. 😅

I also used much of the Van Rekom/Jansen book in my own explorations, and I'm trying to find out what ultimately makes the plan with h6 and g5 work in these positions. I'm not a tournament player, and I play mostly only rapid time controls, so I find it to be a very practical weapon.

Having said that, I agree that there seem to be many lines which aren't covered at all, much less properly! For example, if the point of the line with 5... h6 is to delay the placement of the Bishop, then they sure didn't go into any detail about it whatsoever, even on the very next move. If White just castles after 5... h6, they even give 6... c6 in the sample game, but they don't cover the delayed sac on f7 at all: 7. Bxf7+ Kxf7 8. Nxe5+ Kg8 9. Ng6 and now Black can't get away with playing 9... Rh7, because 10. e5! Qe8 isn't a pin anymore. Instead, Black has to play 9... Qe8, sacking the exchange, and after 10. Nxh8 Kxh8 White has a Rook and two pawns for two pieces. It's still at least slightly better for Black, as my engine shows me (and it's really one of the rare instances in which I used it, because I was so puzzled); nevertheless, why such an idea was left without any coverage in the book is beyond me.

I will now briefly go over the eight details you mentioned. I just wanna preface this by saying that I don't want to consult an engine on this one, as I'm just analysing and trying to evaluate on general positional grounds and also with the most usual sacrifices from the book in mind. I want to figure things out on my own, to the best of my purely human understanding.

Now, on to the details:

a. The problem here seems to me to be the fact that Black played ...c6 AFTER White had played a4 (which usually serves to deter Black's expansion with ...b5 after ...c6), so Black effectively lost a tempo on playing a move for which the most principled follow-up had already been prevented. If after 7... g5 8. dxe5 dxe5 White plays 9. Nd5 (which could be seen as a natural drawback of delaying ...c6), then just 9... c6! anyway (Black can't keep his pawn advantage after 9... Nxe4 Re1 or similar), and White is either losing time on retreating or relieving the position with a piece trade. Black just needs to avoid being greedy about the e4-pawn and focus on achieving the usual setup first.

b. This is my preferred move order, maybe not through the given move order (because of what I just covered in (a)) but certainly via transposition.

c, d, e. I wish you explained why ...g6 was a bad move, but I have to say that I agree, even on my gut feeling only! While it does serve to prevent knife-f5, it weakens Black's kingside too much in return, especially without the fianchetto. I remember trying to make some delayed Pirc lines work from the Lion setup with ...h6 and ...c6 but before ...Be7: I'd play ...Qe7, ...g6, ...Bg7 and ...0-0, but due to ...h6 already having been played, Nh4xg6 would be possible due to the pin along the a2-g8 diagonal, and White would get way too much initiative even otherwise.

f. 10... Nc5 doesn't seem like a bad move at all, but:

1) is 9... Be7 really the only move in the position that's any reasonable? I have some other ideas here that might be just a tad too crazy (so I've decided to withhold them for now xD), but if Black could get away with bolstering either g5 or f7 while leaving his Bishop on f8, then he could take the edge off of the early Nf3-h4-f5 manoeuvre;

2) is there any concrete reason why Black shouldn't play 8... Be7 instead in this variation? If 9. 0-0, then 9... g5! transposes to (b); but if 9. Nh4, then 9... Nb6, and as soon as the Queens get traded and the Bishop retreats from c4, I don't see why ...Ng4, either winning one of White's Bishops or doubling his central pawns, would be anything less than slightly better for Black.

g, h. I agree that 7... c6 should be played here any way you slice it, but now I think 8... Qc7 must be a better move for Black than 8... a5, and that also gives me the main answer that I'd been searching for when I posted this topic. Again, the Bishop still being at f8 only helps Black against the early Knight jumps to f5. I see two main variations here:

1) 9. Nf5 Nb6 and now f7 is protected by the Queen, so the Bishop must retreat to d3, else ...Nxe4 would at least require a desperado check from White in return: 10. Bb3 Nxe4 11. Nxg7+! Bxg7 12. Nxe4, but if that's too dangerous for Black, also available is just 10... Bxf5 11. exf5 Rd8, or just trying to transpose into the next variation with 10... g5. After 10. Bd3 Bxf5 11. exf5 0-0-0, it seems to me that Black has at least a slight initiative;

2) 9. 0-0 g5!? and now White has to play 10. Nf5, because Nf3 would transpose back into an even better version of (b). Following that, 10... Nb6, and it ends up as almost the same position as the previous one. It should be said, though, that after, say, 11. Bb3 instead of 11. Bd3, 11... Nxe4 seems to me like it works, but White also seem to have some additional tricky moves at his disposal, even without the desperado on g7: 12. Nxe4 Bxf5 13. Nf6+ is the difference, but after the bongcloudy 13... Ke7 14. Nh5 Bg6, White seems to me to be just a clean pawn down.

I hope my own TL;DR is up to your standards as well!

I only posted that long stuff because I play the Philidor with both colors, and I too wanted to know what to do about an early Nh4.

Every opening has problems. The more you analyze at home the bigger your practical advantage when something similar crops up in your game. It's okay that you didn't use an engine. But I really recommend you take the next step -- put all your variations in a pgn file, write up your conclusions, and then go over _everything_ with an engine. I always learn something that way, and sometimes what I learn is big enough to overturn my conclusions.

Some general non-engine points.

1. ...g7-g6 just wastes time and creates additional kingside weaknesses. The only possible point is to fix the white knight on h4 until black can make a good discovery on it. Unfortunately, black can never make a good discovery.
2. ...c6-c6 is a good positional move. If white hasn't played a2-a4 yet, then it has an _incidental_ tactical threat of ...b7-b5. But if white has played a2-a4, black still wants to play ...c6-c6, there just won't be a second idea behind it.
3. ...Bf8-e7 is about the worst possible reaction _after_ Nf3-h4. Discoveries will never work! In the meantime, one idea for white is to play Nh4-f5. By playing ...Bf8-e7, black unguards the g7-pawn and puts the e7-bishop in range of the future f5-knight. Makes no sense. Since white made the early d4xe5 exchange, black should look for ...Bf8-c5 or ...Bf8-b4, otherwise leave the bishop on f8.

About the gaps in the Lion book, this is usually intentional. These guys play the opening regularly, if they covered it in a systematic and complete way then their opponents would know exactly where the problems are. So they give a ton of "interesting" analysis, while in the critical line(s) it is just crickets. Similarly they don't have a well-defined repertoire move order in the book, solely to confuse future opponents. I have seen the same shoddy treatment by some IMs and even GMs going way back. But these days it's harder to get away with, when even the lowliest amateur has Stockfish and ChessBase for checking everything, and the internet for broadcasting any criticism.

Thank you for the PGN recommendation! Kinda weird about the intentional gaps, though, because it certainly doesn't help the players themselves, either.

Im no expert in this line but I get the feeling 7...Nc5 equalizes immediately. The ending seems fine for black and if 8. Qe2 Bg4 9, Qe3 Be6 or 9, f3 Be6

 @jamesstack - 8.Bxf7+

Looks like I need to get back to studying some tactics. Dont know how I missed something simple like that.

The other thing I was wondering is why is black playing h6 so early? if black isnt really intending g5 then wouldnt some other move be more useful like c6 or Be7? Someone mentioned that they didnt think the g5 castle long ideas werent rreally all that sound.

Well, there's your answer right there in the question: I ALWAYS intend to play ...g5 eventually. 😃

Yes, it's true that in some lines it might be more technically sound for Black to just go back into the good Hanham lines with ...0-0 and the slow queenside expansion (i. e. ...c6, ...b6, ...a6, ...Rb8 and ...b5, or just ...c6 and ...b5 if White omits a4), in which case ...h6 probably is an objective loss of a tempo; however, if I played 5... Be7 instead of 5... h6, even if I was more comfortable than I currently am with the most dangerous sac on f7 (6. dxe5 dxe5 7. Ng5 0-0 8. Bxf7+ Rxf7 9. Ne6 Qe8 10. Nxc7 Qd8 11. Nxa8), I'd have to contend with the early Nh4 lines all the more, seeing as the ideas of knife-f5 and/or f2-f4 are no less dangerous when Black's already castled short (plus the fact that Nf5 potentially actively harasses Be7 as well). I'm honestly not ready to play such positions, so I'd rather take my chances with a weakening my own kingside pawn structure for the promise of an attack.

Also, the tactics against f7 come much more naturally once you've played the Lion enough to be on the receiving end of at least a few of those. 😅 It's a sort of a venom that just comes inherently with the opening: Black spends several moves worrying about Bxf7+, then a few more worrying about Bxg5 hxg5 Nxg5 and Bxf7+ again, and then perhaps a few more worrying about Nxe5 and Bd4 if his Rook is still on h8 in the positions with ...g5. The price of Black's slow-burning attack is his equally slow-burning tactical vulnerability beforehand.

I'm going to cover all three of your comments on this topic and provide some answers to them:

1. Wait, you literally mean 1. e4 d6 2. d4 e5?!, right? While it might be true that the position would be virtually equal after 3. dxe5 dxe5 4. Qxd8+ Kxd8, White's best third move is Nf3!, after which 3... Nc6?! 4. Bb5! tricks Black into the pure Steinitz variation of the Ruy Lopez (after which it's too late to have played ...a6 first, which would've been a clearly better option for Black in comparison), and 3... Nd7?!, the old Hanham move order, runs into 4. Bc4! (after which White can forcibly win a Bishop for a Knight and get a position in which only he can play for a win, effectively). Because 4... exd4 would then land Black in a clearly inferior Exchange Philidor (not Lion!), the best option is to play 3... exd4 instead, after which White can choose to recapture with either the Knight or the Queen. The true Exchange Philidor isn't unplayable for Black, but my personal experience tells me that it's much more enjoyable from the White side if White doesn't get too impatient with his pressure. Also, after 2... e5 3. Nf3, 3... f5? (the Philidor Countergambit) has effectively been refuted by modern analysis. The English Wikipedia page on the Philidor Defence has most of what I've written so far concisely yet effectively explained, honestly.

2. "Passive, non-developing, weakening and running contrary to what Black should be trying to do"? That's highly debatable. Passive and non-developing? Yeah, kinda, but only on its own, as it's angled towards a structure with g5, which is both active and with its own development scheme in mind. Weakening? Only if you either really need to play ...g6 (which is either bad, as yetanotheraoc already wrote, or doesn't happen at all in the Lion structures) or if you castle short too quickly (i. e. before initiating at least some sort of a Queenless middlegame) and thus invite sacrifices on h6... otherwise, like in castleless Lion, it's not an automatic fatal flaw by any means. Running contrary to what Black should be trying to do? Bold of you to assume that transposing back into Hanham with ...Be7 and ...0-0 is the only viable option!

3. Looks can be deceiving... and as for what the Lion is trying to achieve, well - haven't you read it anywhere above here in this topic at all? A solid yet flexible defence (especially with 5... h6), and an attack on White's castled King, with the standard Philidor queenside expansion plan remaining an option in almost all scenarios.

Well if you always intend g5 then i guess it doesnt matter too much.

For some reason, I didn't even register that you wrote "g5 castle long ideas", so I'll have to amend my words after all: I want to play g5 as long as White is more likely to castle short than long. If White wanted to castle long, then he would delay castling (in which case I'd delay playing Be7), he would play something like Be3 and Qd2 (which pretty much telegraphs it), and he wouldn't play a4 in response to ...c6. Then I'd play b5, and my attack would be ultimately be quicker than his, in all likelihood. The point is that trying to castle long doesn't go all that well with the idea of Nf3-h4-f5, as Black will have more than enough time to stay flexible and consolidate his entire position.

Even though it seems as though ...h6 isn't useful in those positions, my experience so far tells me that it's an excellent waiting move in the event of White wanting to castle long. I've played some very exciting games in which White even forewent Nf3 and played f3 instead (though this isn't the variation we're discussing in this topic), and at some point White threatened to play Bg5 and Bxf6, either damaging my pawn structure or winning my e5-pawn in turn. Taking said option away from White allows Black to transfer some of his pieces to the queenside quicker, in which case his chances against White's long-castled King are even better.

When I wrote "g5 castle long ideas" I meant black castling long.....reasoning that black probably shouldnt leave his king in the center or castle kngside if he wants to play g5. I didnt even consider the possibility of white castling queenside..

I would like to suggest something. After 5...Be7 instead of 5....h6 if white plays 6. Nh4 is there any reason why black cant play 6....Nc5 with the ideas I mentioned earlier now that Bxf7 wont work snce the queen is protected by the bishop.?

Well, first of all, if you play 6... Nc5, White can play 7. dxc5. xD Nh4 never really works out for White if he doesn't make sure Black can't respond with ...exd4 first, especially if Black can then also safely place one of his Knights on e5... but that's okay, though, I know you both actually meant to say 7... Nc5 instead, after 5... Be7 6. dxe5 dxe5 7. Nh4??.

..."Wait", you might say, "why did you mark it as a blunder?" Well, while 7... Nc5! (threatening the e4-pawn and therefore more active than 7... Nf8) might not be bad per se, why not simply 7... Nb6!! instead? I've prepared two main variations to illustrate the power of this move:

a) 8. Qxd8+ (White might as well keep his castling rights, right?) Bxd8 9. Bb3 (or Bd3) Nxe4 10. Nxe4 Bxh4 leaves Black a clean pawn up in a soon-to-be endgame;

b) 8. Qe2 seems to be trickier because 8... Nxe4?! 9. Qxe4! Bxh4 10. Qxe5+ doesn't really win any pawns, and also because of 8... Be6? 9. Qb5+! c6 10. Qxb7; however, after 8... Nxc4 9. Qxc4 Qd4!!, White has to either exchange or retreat his Queen, after which his e-pawn falls as per the established schedule. A sample line: 10. Qb5+ (10. Qxc7?? Bd8!!) c6 11. Qd3 Qxd3 12. cxd3 Nxe4. In any case, Black ends up a pawn up with the Bishop pair advantage to boot, so he should win.

Having written all this, I think we can conclude that both 6. Nh4 and 7. Nh4 are basically bad after 5... Be7. So, if that's the case, why wouldn't I simply play it instead od 5... h6? The answer lies in the following variations:

That's basically 3-5 variations to learn! In contrast to that, after 5... h6, I only have to deal with 6. dxe5 dxe5 7. Bxf7+ Kxf7 8. Nxe5+ Kg8 9. Ng6, as well as basically the same thing on the very next move but only if I play 6... c6 (instead of 6... Be7) after 6. 0-0. Dunno about you, but I know what I find easier to learn and understand thoroughly.

That's not an unfair point, which is why I mentioned the 6... Nxe5 line in the first place! If you wanna play the Lion in serious tournaments with longer time controls, perhaps a variation like that one can serve you as a whole complex of airtight positions, kinda akin to the Berlin Defence of the Ruy Lopez; however, if you're like me and you happen to feel like you would regularly lose said positions from either side simply because they're symmetrical to the point of boredom, then I would recommend picking something else.

The Lion is a personality test of sorts, indeed. After 1. e4 d6 2. d4 Nf6 3. Nc3, do you prefer 3... Nbd7 or 3... e5? The former allows the aggressive 4. f4!?, while the latter allows 4. dxe5 dxe5 5. Qxd8+ Kxd8, with 6. Bg5 and 6. Bc4 as the main continuations (which I find tricky in general, even if only slightly so; still, I think I at least slightly prefer even such positions to those after the 6... Nxe5 from your question).

1.e4 d6 2.d4 Nf6 3.Nc3 Nbd7 4.Nf3 e5 5.Bc4 Be7 6.dxe5
6...Nxe5  {is certainly playable, although Harding called it dubious.
Here is what Larsen (1971) Why Not the Philidor? said about it: }
7.Be2 {White may have an edge but it is not fully clear}
7...Nxf3+ {most natural}
(7...Ng6!? {the Dutch master Barendregt has achieved good results})
8.Bxf3 O-O 9.O-O Re8
(9...c6! {better})
10.Re1 h6 11.b3 Bf8 12.Bb2 g6 13.Nb5 +/- {Pachman - Dunkelblum}
https://www.chessgames.com/perl/chessgame?gid=1104862

Bauer just stops after 7.Be2. Pickett analyzes 9...c6, but as usual Pickett's analysis is strange. The best source is Kosten (1992) Winning with the Philidor. Kosten doesn't like 9...c6, instead he analyzes 11...Nh7! as a strong improvement over 11...Bf8. Van Rekom/Jansen don't give anything new compared to Pickett and Kosten.

The biggest objection to 6...Nxe5 is that black can very easily get the same type of += position via 1.e4 e5 2.Nf3 d6 3.d4 exd4, without the extra options white gets after 1.e4 d6 2.d4 Nf6 3.Nc3 Nbd7, in particular 4.f4 and especially 4.g4!. Also white certainly does not need to play 6.dxe5.

I dont know...seems like white is attacking with just the knight. I would let white sac and do my best to fight through the complications to win the game. I dont think this is about trying to figure out all the lines and memorize them. its more about a willingness to enter an unclear position and trusting your calculation ability.