My father would sometimes play that exact opening to see if he could rattle me. Of course he’d win in 15 moves or so.
As a teaching exercise, he’d then have me play the same opening. He’d then beat me in 25 moves or so.
My father was much better at chess than I was.
Perhaps, but Chess is not a game of random moves looking for a pattern. The key to the question, for me, in the OP is “with perfect play from both sides”, which to me states that white is attempting to force a mate, and also that they won’t do anything crazy to throw away their advantage.
With the advantage of a queen, a player who knows what they are doing and is trying to win, can force black to respond in particular ways, at the risk of getting further and further behind in material, so the number of moves that black has for the rest of the game, rather than being limitless, is extremely limited and only leading in one direction. All of this would be very different if the gambling element of chess had never been removed.
Junior Spaceman You are still missing the point being made. You’re putting “with perfect play from both sides” as if I’ve missed that, or don’t understand that, when I included that phrase in my post and I play chess competitively.
Put it this way: let’s say a 60 year old guy decides one day he wants to break the 100-metre sprint world record. I would bet my life savings, and probably my life, that that ain’t gonna happen.
The idea of there being a 50-move combination is of that level of probability. It’s purely theoretical. If your objection were “Stop being so pedantic, of course it’s not gonna happen”; sure I could agree with that (all I’d suggest is we drop the language of “forced win” which is only ever used when we can see a precise combination that leads all the way to mate or a solved endgame). But your responses up to this point imply you haven’t appreciated that I was speaking purely hypothetically.
We cannot “know” unless and until someone runs a computer calculation that explores every possible avenue. So far, we have no computers capable of doing so.
But we can “know” in the sense that we can be quite certain (as in with extremely high confidence levels) that black cannot “force” a win or draw. To “force” such a result would mean that, even with white making optimal choices, black still has responses that either force a mate, or force enough advantageous material exchanges that, in the end, white will not have a mating advantage. The chances of that being possible are so vanishingly small that we can dismiss them as not existing.
In short, to speculate that black could possible force a draw or win here is to take the basic idea of “solving chess” (determining if white’s first move advantage guarantees a win with perfect play) and ask the same question after handicapping black a queen down! If we cannot “solve” the first, we certainly cannot prove the second. But we can make a really, really good argument that white has a winning position, when a queen up after move four of the game (especially as black is constrained at this point to make a less-than-optimal move to avoid Nxc7!, which loses the major exchange). In light of this fact, I will assert that Yes, with perfect play, White wins.
Stockfish at a depth of 19 moves can’t see a way to force mate, so no human is going to be able to answer this question!
I would guess that with perfect play by black, mate probably isn’t going to happen until 30+ moves into the game. That’s another question that’s somewhat interesting to me. With perfect play, how long can black postpone mate?
Well, I was using Stockfish on chess.com for the first analysis. Leaving it running on my computer, after 20 minutes its searching 32 moves deep and still no mate.
MCO (does it still exist anymore, in this Internet era?) used to refer to 1. e4 d5 as the Center Counter Defense. I used to play it occasionally as Black, but didn’t do very well with it.
I agree that it’s a notch below the other replies to 1. e4. The problem for Black is that it gets you through the opening just fine, but it leaves you a bit flat-footed as you move into the middle game, with no obvious direction or focus of attack. A first-move reply as cute and counterintuitive as this shouldn’t lead to such a blah middle game, but it does.
Thus confirming my supposition! 
Seriously, 20 minutes to get 32 moves deep? Holy shit. I had no idea these chess engines were that fast to get that deep these days.
I’m going to leave in running for awhile to see where it goes. After analyzing 36 moves deep (over 1 billion different lines) it scores the position as +12.94 (i.e. it says white is ahead by the material equivalent of a queen, a minor piece and almost a whole pawn).
So, it looks like white is doing pretty good in the long run. 
Assuming he’s not lost here.
Being up a queen, a piece, and a pawn? Yeah, I’m pretty comfortable in betting the farm he’s not lost here.
DSYoungEsq you’re repeating back to me what I am saying.
I don’t know how to say it any clearer than “I’d be willing to bet my life that white wins with correct play”.
But nevermind I’ll drop it, it’s getting annoying.
Update: after 42 moves deep the computer scores it as +13.39 (I suppose it is only natural that it is tending to rise with each depth as with a fairly substantial initial advantage by continuing to search white should find better ways to exploit it).
However it is worth noting as has been pointed out it is still possible that in theory the initial set-up is winning for black.
White’s not up a Queen; white’s up a queen-knight exchange. Still an easy win for white, especially since black seems prone to blunders.
Why would black suddenly play perfectly after blundering his queen? It seems like a tactics problem, from Chess Tactics website, say, where the computer plays a blunder on the first move (to set up a tactic which the player must figure out), then suddenly plays “perfect” chess. But your question seems to be one of common players, OTB, not that of a tactics trainer.
There’s been some good answers already and having a computer analyse ahead is better than any human can do.
Meanwhile based on my own experience (I’ve had a 2300+ Elo rating for decades), it would be trivial to win against Kasparov or Carlssen (or even a computer) with a queen up. E.g. simply exchanging pieces increases this advantage.
You may be interested in the odds match Kasparov - Chapman (I think Chapman was rated about 2250 at the time) where Kasparov gave odds of two pawns.
Kasparov won by two wins to one, with one draw over the four games. Given that Kasparov would expect to beat Chapman 100-0 (if they played that often), two pawns was a serious problem for Kasparov.
A queen is far more of a handicap!
Inn a game where a player is giving a 2 pawn advantage is there a custom to which 2 pawns? (Kasparov’s choice, Chapman’s choice or a fixed order of removal)
I’ve always wondered about pawn handicaps, too, because having those pawns out of the way radically changes offense as well as defense. If, for instance, you give up the end pawns, then it suddenly becomes a lot easier to develop the rooks. Knights’ or royals’ pawns, meanwhile, enable quicker development of bishops, and queen’s bishop’s pawn enables development of the queen.