It is arguable that if you are in a must win situation as black, then you should avoid 1...e5 against 1.e4 altogether in favour of a more imbalancing choice, such as 1...c5. I was just curious as to what you guys think is the best fighting choice against the Ruy Lopez from black's point of view. I don't necessarily plan on taking up this line, but since I am learning the opening for both sides (especially white) I think it would make for a good discussion topic. For all you die-hard 1...e5 players, what do you guys turn to when you need to create positional imbalances necessary to outplay a weaker opponent, without making serious concessions.

I am basically taking a theoretical standpoint here, meaning that with best play what defences do you think leave black with enough scope to play for a win? Assuming both players know the theory quite well and play at a reasonably high level. Personally, I think the Berlin is quite a good choice as it leads to an imbalanced endgame, and white can't just simplify to a completely lifeless position if he wishes. I know Aronian has scored quite a few wins at the top level with this defence.

It seems like the sharp attacking lines, such as the schlieman or Marshall, leave black defending a worse endgame with best play, often a pawn down in exchange for two bishops, and aren't ideal for playing for a win. In many of the closed lines it seems like it is tough for black to do anything if white doesn't overpress, and white often gets a small but persistent initiative on both sides of the board, while black just defends and trys to neutralize the pressure. So what do you guys view as black's best choices in a must win situation?

sometimes, those "weaker opponents" tend to surprise me with their attacking lines, so I stick to the Petroff. 

Back in the days when I used to defend the Spanish Game, I chose the Moller Variation of the Ruy Lopez or the Breyer Variation. But now, I use the Petroff defense most of the time because its full of tactics

If you're set on playing the Ruy as black in such a situation, you might want to consider the Arkhangelsk variation. It's tactically very sharp, yet doesn't leave black for wanting in an endgame like the Marshall does. Perfect for playing those lower rateds you mentioned.

How about the Breyer? Featuring Nb8, which relocates to d7 to protect e5, followed by Bb7? Bf8, and Rf8?

Both breyer and archangel are quite good. I think white can get a slight edge against both variations, but the play is with chances for both side. The breyer leads to a stragetic fight while the archangel leads to a more tactical fight.

Actually one of the main lines of the arkhangelsk variation does lead to an endgame you simply have to draw.
An example of this variation from the Kasparov-Kramnik 2000 world championship match:

Though to be honest, I think you can even use the berlin wall to generate winning chances. It's simply about outplaying your opponent from an equal or slightly worse or slightly better position.

I follow the spirited fighter Korchnoi in favoring the Open Defense of the Ruy Lopez.

---

My database begs to differ with your database. For third moves with 10 or more games played, my DB agrees that 3. ... f5 has the highest winning percentage for black (33.2% wins for black, of 4339 games, with only 39.2% wins for white, and 27.6% drawn).

Just looking at black's winning percentage, my DB gives 3. ... Bc5 as second best, at 30% winning for black, though white's winning percentage is a very high 45%! 3. ... g6 is third with 29.9% wins for black, but only a 34% win rate for white, so it does appear to be the 2nd best choice overall.

But hold on a second! There are enough sub-choices after 3. ... a6 that they're probably worth considering separately. If white does go for the exchange with 4. Bxc6, then black has a respectable 28.2% win ratio, so it might be worth chancing, given that white only plays this roughly 20% of the time. After 4. Ba4, black's best winning chance seems, surprisingly, to be 4. ... f5 - the delayed Schliemann, with 37.8% wins by black in 201 games.

Of course, you can't make your decisions based solely on statistics. The statistics say that white's best opening is the Blackmar-Diemer Gambit. Playing style of both players, knowledge of the opening, new theoretical novelties, etc all have to be taken into account.

What about the Marshall, I thought that is good for black ?

In my experience it has not worked out very well.  It is a very difficult opening to understand; it does not compare to other gambits.  I think that I must get a great deal stronger before I will attempt to learn that again.  You have to be very careful about keeping counterplay otherwise white will slowly grind you away.

Personally, I like the Chigorian Variation.  Black gets strong queenside pressure and can easily repel white's kingside attack.  I really think that white struggles to get any tangible advantage.

Thanks for the replies so far everyone. I plan on playing the closed lines as black, probably with a focus on the Zaitsez and Chigorin. If I take up a secondary "fighting" choice as black, it will probably be the Moller variation, which Shirov has used to good effect over the years if my memory serves me right. Here are the reasons I don't like the third move alternatives for black:

3...Nf6. This probably is the most principled move for black, developing a piece. However, the resulting endgame that arises in the main line is quite strategically complex and is probably best reserved for players over 2200 strength, who have a real good understanding of strategy. I will play into this as white, but before I play it as black I would like to be much stronger.

3...f5. This line may have good statistical results for black, but when I study the theory from the white side I can't help but feel that black is just fighting for a draw in the critical lines. For example 1.e4 e5 2.Nf3 Nc6 3.Bb5 f5. 4.Nc3 fxe4 5.Nxe4 and now 5...Nf6 6.Nxf6 Qxf6 7.Qe2 leaves black a pawn down for the two bishops as compensation after 7...Be7 8.Bxc6, and it isn't enough compensation in my opinion. Can black really expect to play for a win here against accurate play from white? It doesn't seem like it to me. I don't know what the current status is on this line, or on the 5...d5 6.Nxe5 line, but from what I have seen I would much rather be white in both cases. I am surprised Radjabov gets away with playing this defence at the top level to be honest.

3...g6 just looks wrong to me, as does 3...Nge7 followed by ...g6. White plays d4 quickly to open the position for his better developed pieces, and black's set-up just seems too slow. I know concrete analysis is required, but I am just going by principle that black should develope quickly in an open position. In general it seems wrong to me to play ...g6 early in any line of the Ruy.

3...Bc5 seems alright but I would much rather be white as he seems to get his way in the center.

As for more mainline stuff... The marshall isn't practical because if white really knows his stuff black can't hope for much more than a draw, and there is just so much theory to know. Worst of all, white will often avoid it by playing an anti-marshall set-up so you may rarely get to use your preparation. I know when black trys the Marshall againt me I always avoid it with 8.h3.

As for the Open, it seems good in principle, but doesn't black have serious problems in many of the lines? Not that it would matter much at my level anyway, so it is definitely a candidate. Likewise, the Archangelsk seems to me similar to the Moller in some ways, but also seems theoretically weaker. I remember reading somewhere that white can often get an improved "anti-marshall" set-up against it.

That was a bit long-winded... Let me know what you think.

Your post indicates you're not interested in something that is sharp against weaker players, but rather you want something that somehow gives black a better game than white no matter what playing strength! Let me tell you now, it doesn't exist. You even said it youself while investigating many of these lines, that white gets a slight advantage with best play. Yeah that's called CHESS.