does the Scandinavian actually work??

Of course it's not 'great'. Just playable. The same though can be said of all sorts of openings/defenses. Few are truly 'great'. I guess the Berlin can guarantee Black a draw if White heads into it and Black is 'booked-to-the-T'. That could be considered a 'great' defense for Black I guess if you only need a draw.

low, its a dubious opening choice for black and is only played for tricks

just look at my games. i play it every chance I have and win with it more than I loose.

Depends on...about the same things as any other choice you might make:

1. How well you know it
2. Which line you play it
3. How good a player you are
4. How good your opponent is
5. Time control
I check some chess.com blitz stats - the first variation I come upon, I encounter a lot and it says 45% win rate over 335 games. Now, keep in mind my rating range for games is set for 'no less' than 50 points below me all the way up to infinity above me. Not bad.
My OTB, the last time I checked my games stored in Chessbase is about a 85-90% win rate....but then I'm usually in the top couple of percent of players in these weekend swiss tourneys and you simply have to play for a win in those...which makes it a good choice. I think though I might do as well with various other defenses...I've certainly play a lot of others over the years.

I took a look at 365chess database of all ratings: After 1 e4 d5 black scores 45.4% (27.4% draws, 31.7% wins). In the masters and above only database section black scores 41.95% (43.7% draws, 20.1% wins).

By comparison, in the all ratings 365chess database the Caro Kann scores 46.3%. In other words, it scores about a point more every 100 games. That's to be expected, it is the better defense generally speaking, but the Scandinavian is close to it in scoring. At the master+ level the Caro Kann scores 45%. That's 3% more than the Scandinavian in the upper level, which is also not surprising, very strong players are in general better able to convert advantages.

At below master level the Scandinavian appears to be a very practical choice.

Breaking down the Scandinavian into a bit more detail, in the all ratings 365chess database

3...Qa5 scores 44%

3...Qd6 scores 45.5%

3...Qd8 scores 41%

2...Nf6 scores 45.6%

@UnsidesteppableChess. That really seems to be about right inline with my chess.com blitz ironically. 3. Qa5 mainline says 45%. I rarely play the Qd6 or Qd8 lines but have played 2...Nf6 a lot. My most recent game on chess.com is a 3...Qa5 vs someone about the same blitz rating.

I find I often get 'Coach' saying what he said about that one...something like 'Great game! Your opponent was a bit better out of the opening but you were 'on another level' in the middle game. You saw your chances and took them'.

Basically...most openings are playable...if you know what you are doing.

You pray your opponent does not play the one defense you do not want to see and probably have not seriously worked on because everyone tells you is so bad: 1...d5! 

Just flip what was mentioned a few posts ago by someone...ie, if it's 45% Black, then it's 55% White (draws not counting or factored in...not sure). But it's basic math.

Cornfed,

Speaking of percentages, you might already know but I was wondering if you are aware that in a main line after 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qa5 4 d4 Nf6 5 Nf3 Bf5 6 Bc4 that 6...e6 is the modern way of playing instead of the 6...c6 that was very popular for a long time. 6...e6 saves the c6 square for a possible ...Nc6 and prepares to answer 7 Bd2 with 7...Bb4. 6...c6 scores 43% for black, 6...e6 scores 47%.

UnsidesteppableChess,

Absolutely! It's referred to as the Laznicka variation. I found it in a...think it was a Chess World course...or maybe it was Robert Ris's Killer Scandinavian. I've been playing it a lot...still need to work on it. Perhaps not the best....but certainly the quickest yet against a pretty decent player is this miniature: