What is it that makes a game good? I just finished reading a blog post on that issue by Tobold that talks about his views on the subject which essentially boil down to whether or not a game is entertaining. He goes on to talk about how if a game is entertaining it is going to have big sales numbers so from his perspective it is reasonable to say a game is good if people buy it a lot. Of course sometimes numbers get really messed up by marketing or other misconceptions so raw purchases are often going to be entirely misleading, but he does rely on subscription numbers to be useful. I figure that if you want to measure the popularity of a game you will get a fine metric by looking at how many people are willing to pay for it month after month.
Tobold, being a reasonable sort of person, suggests that his definition is by no means universal and that other people might use other definitions. This sort of behaviour is very much uncharacteristic of people making posts on the internet which is why I like his writing a lot. He has opinions but tries to couch things in terms of points of view instead of Good Vs. Evil, which means he ends up being right a lot but not nearly as compelling as some of the more strident agitators out there. I don't agree with his definition exactly because I put a lot of stock in analysis and replayability that is not necessarily well reflected in subscription numbers. If a game has a ton of people paying to play but isn't very interesting to think about I don't give it a lot of credit. It might be a bit of a stretch but slot machines have a ton of people who consistently go back to them to pull the handle and see what lights up but they are a good example of a wretched game. No decision making, no skill, virtually no interaction, but yet the shiny lights and lure of the big payout are enough to keep the customers coming back year after year. A little more on topic there are lots of war games out there that continue to sell copies and make money, particularly the miniature oriented ones, and yet usually it is completely obvious what the best units and strategies are and no real analysis needs to be done.
I feel like there is some kind of compromise necessary. There are games out there where the skill required is high, the depth of strategy and thought possible is tremendous and yet nobody actually wants to play them. Le Havre is a great example of this in the board game field; it is a great game and I could spend months trying to figure out the best possible choice in a given situation and mathing out the optimal route to play and yet the game takes 3-5 hours to complete so it is very difficult to actually get people to sit down and play. Titan is very similar in that it has incredible depth and requires a large chunk of a day to complete so only the most hardcore players actually get to see any of it. Heck, even Diplomacy fits in that category; so many gamers love the idea of it and never actually play it.
Of course by this compromise definition incorporating both depth of play and popularity the games I play are tops. Warcraft, Starcraft and CiV are all huge moneymakers, all are massively popular, and all have extreme depth in both strategy and in game decision making. I suspect that if you look for the game that is the pinnacle of monetary success in its field you will usually find one that has all of those things working for it; it is hard to be top dog without appealing to both casual and hardcore gamers.