Questing about armageddon playoff

If the time control was going to be 5 minutes for white and 4 minutes for black and that black has draw odds (a draw = a win) which color would you chose?

I've heard it's preferable to chose black.  I'm wondering how effective psychologically it would be to play white, blitz out the hippo defense, and so threaten to win purely on time.  As for general game play, white will play for a draw on the board.  Black's main idea in such a game would seem to be never to overreach and while he shouldn't play passively black will generally avoid any complications.

So black's ideal reaction would be to calmly develop his forces, and calmly prepare a break somewhere.  However if it's too calm, white can continue playing for a draw on the board (and a win on time).  Black seems to have to walk a fine line, but in practice I wonder how effective this would be and if black would still be preferable.

White does not play for a draw, black plays for a draw.  White MUST win.

Well I can see my strategy would completely catch you off guard, and so I'd guess be pretty effective 

But against GMs they wouldn't overreact like that (omg white MUST play for the win I'm confused) so I'm wondering how effective it would be on an experienced player

To clear up the confusion, a win on time is a win for white.  The premise is white plays for a win on time as leverage vs black's draw odds.

Fine, just let me know where the cut off is for "strong player"

http://www.chessgames.com/perl/chessgame?gid=1106734

http://www.chessgames.com/perl/chessgame?gid=1102287

Also one of Gonnosuke's who advocates it here http://www.chess.com/forum/view/chess-openings/hippopotamus-defence?page=1

It's obviously not the best opening, but this is blitz we're talking about.

Obviously you don't stay passive the whole game, just the first 10-12 moves.  Especially if black opens anything up you kick into the standard, make it as sharp as possible.  Because you have more time, if it ever gets very sharp you have the advantage.

If it was this time control, as black I would play as in a KID for white, with ...c5, ...d5, ...e5 etc, because this was breakthroughs in the center are more straightforward to achieve. Also quite possible would be an austrian attack type setup: ...d5, ...e5, and ...f5. It's not necessarily better than other setups, but these ones wouldn't be too tough to play: just develop your pieces, castle, and if the opponent doesn't strike in the center then play ...d4 or something, and if the center is closed  black has a textbook queenside attack (with pieces like the b2 bishop for white to need repositioning), while if it opens it should favor black's better pieces. If the game just happens to draw, then great, but that's not what I'd be aiming for. If you aim for a win and still miss, it'll probably at least leave them begging for a draw, right?

Indeed, this is probably how I would play a regular blitz game against a hippo too.

I would say black still has better chances, if only because I think these setups are effective against a passive hippo, and that draw odds is just a nice bonus.

Hmm, on 2nd thought this might be too one dimensional.  If you're going to play for a win on time it would be better to do it more covertly, go for a sharp opening, but then never quite commit to an attack, set it up but just leave it lingering.  Of course against an experienced player as black they wouldn't hesitate to seize an advantage and play for a win.

It just seems like a 1 min advantage in blitz between equally matched players is a lot and yet black is generally favored.  I guess it's obvious to say that in general white will keep the tempo high and black will play defensively with both players being willing to switch depending on the position.  The guys I usually play with hate any kind of variant even if it's just a time variant, but I'll see if I can't get them to play a few with me so I can get a feel for it.

Except you can't draw, it will be impossible to draw.  Because white is going all out the only time a draw gives black the win is when there is a perpetual or stalemate.  If the pawns become locked white can just shuffle back and forth and win on time.  Yeah it's stupid way to win, but it's an even more worthless way to decide a champion.

I like the draw odds though. If white picks the hippo he's pinning solely on the fact that black overpresses rather than builds slowly or screws up due to lack of time, but otherwise he's under pressure to win in that type of position. I've never played this time control though, so it's just my guess. I think a strategy like that would be more effective in like a 2 min to 1 min control.

Wait, is this with no increment too? If so, then it makes draw odds a lot less useful! I would probably want to be white then, as he has more time, the first move, and can avoid even a clear draw for many moves until stalemate or 50 move rule happens.

I guess it all depends on the skill level of the opponents involved. Anand-Topalov? Anand has a clear advantage playing blitz. The white player has the advantage of prep on infrequent lines, but the black player knows he only has to draw. Not too sure on the stats for this type of tie-break game...but my money is on black...especially at the GM level.

That's right Elubas, I'm thinking no increment (increment severely undermines white's side of the bargain imo, although I know it can be played this way).

Anand-Topalov I'm not sure Anand has the advantage in blitz.  Strong players are monsters at blitz, and Anand is a bit older now, and not as fast as he used to be.

At the GM level, black is supposed to be favorable with this setup.  Because the players would be evenly matched, and the drawish nature at that level, it doesn't surprise me.  I guess that's why they sometimes use 4 for white and 3 for black to make the time advantage even greater.

About how much better is Fritz 12 than Fritz 5.32? I need to know because I'm going to do an experiment.

I guess it still won't be very helpful, because I don't think that the engine knows I'm giving it draw odds.

Fritz 5.32 isn't on the current computer rating lists anymore, but about 8-9 years ago it had a computer chess rating of 2500, fritz 12 is at least 3000 if not 3100 even.

So a 500-600 point difference gives a 96%-98% expected result (respectively).  If your experiment got it to score as low as 92% for example (400 points) it would show a large improvement so maybe a bit hard with such a large difference in ability unless you're using 1000 games.

Hah! Thanks for bringing Gonnosuke's game up again. You got to appreciate the pawn-structure after move 12. Ujtelky even managed a straight line. Great stuff.

I think I'd rather be white. I almost never draw in blitz so I figure that the draw advantage for black is minimal. I'll take the time.

i agree with thesexyknight about the draw thing, i also have never won at blitz by playing passively.

Prepare to be amazed:

Fritz 12, playing white every time, scored 11 wins

Fritz 5.32, playing black, scored 9 wins/draws

That doesn't seem... possible :)

Then I probably screwed something up. I do that a lot. Maybe somebody with less screw-up tendencies can perform a similar experiment?

Or play some unrated live games and watch for when black's clock gets down to one minute.