Most important trait? Have the last name Fulop.
In all honesty, there are alot of things that go into being a great player. I break it down into key broad categories: In game playing, deck building, and metagaming. All are equally important at the end of the day. In game playing simply means how well you handle your in game moves. Don't make mistakes, know all of your matchups. KNOW PROBABILITY. WELL. The # of times per game you need to try and get a card, or figure out odds your opponent has a card, or if they can draw into a few cards to "go off" is incredible. Multiple times almost every game you need to make choices that probability influences. The decisions may not be terribly relevant every time, but are often enough that this is critical to learn to do. It gives you a useful edge. Most of the times I end up taking a long turn, almost the only times, is when I am working out odds. Many times when luck is involved, unless you have a read on someone, or think you do, probability is the best judge of what the right moves are. Many decisions take into place alot of different factors that need addressed. To play well and do all of these calculations quickly is difficult, but gets easier the more you play. ALSO pointing to the previous point, KNOW YOUR DECK LISTS. And try and figure out your opponents. All of this is needed towards your calculations. Mikey is 100% accurate with that. I can USUALLY come EXTREMELY close to knowing an opponents exact list by the end of a match with them. You can usually even make blind assumptions based on a format: Assume a stage 2 deck has 4 RC, 4 Bebe, 4 Roseanne, 2-2 Claydol, at least one Uxie, etc.
On the topic of in game decisions: Ross nailed it. Alot of players can play a game turn by turn. Yet one of the things I feel I excel at is playing a matchup. Know the "whole picture". Know the general way matchups flow. Know what type of board positioning favors your deck,a nd what favors theres. Know key turning points in the game, and what, theoretically each deck wants from that point. Than, work your entire game plan around making sure those vital points favor you. THIS is a huge difference between a solid, and a good player.
Next point: deck building. Some people are better at it than others. Someone said some players are good players, but bad deck builders, or vice versa. I disagree. If your a great deck builder, you are at the very least going to be a good player. Not great, but at least good. There is not a player in this game who can claim to be a great deck builder without understanding the very specific fine tuned points of competitive playing. Unless they just blindly get lucky on every build. It doesn't make sense at all for them to be. A good player can be BETTER at making a deck than at playing, but there is rarely a great difference between the two.
A number of things go into building a deck. You need to focus on consistancy, yet also making sure to have options. Make sure your deck does what it wants to do as quickly and often as possible, while making sure what that thing is actual wins games. You need a balance between consistency and speed. Have a goal for your decks win condition. Some decks benefit off of being fast, and disruptive. Others aim towards an inevitable set up that overpowers most other decks. Some decks want to have an answer to all of the other popular decks. Focus your deck. Pick either a strong Pokemon and focus around it ( Kingdra for example ) or pick a couple that have inherent synergy that combo together ( Emp w Bronzong, for example ) and make sure the other pokemon support it. An example of a BAD combination of Pokemon? Machamp Lucario from a few years ago. Machamp was a good pokemon. As was Lucario. Yet what synergy did they have? They used the same energy types? They both served the same purpose: attack. They didn't cover any of their weaknesses. They were utterly redundant. If anything, you'd want to run a different type Pokemon to cover any of your weaknesses. "Doubling Up" on two attackers of the same type is rarely a good decision.
Now onto the third and final point: Metagaming. You can play near flawlessly. You can make a great list. Maybe even the best list possible for a given deck. Yet you can walk into an event and get steamrolled. You need to know how well a deck fits in with the other viable and popular decks. There is a difference between making a smart choice, and making the BEST choice. It isn't that difficult to make a smart deck choice. That should be your goal. You can, and should, expect that of yourself. Making the BEST choice is extremely difficult. After an event, you can always look back and think to yourself what deck would have given you the best opportunity to take first. Worlds 2005 for example? Muk EX would have been the correct play. Worlds 2007? Metanite with Sceptile d EX. You have to predict what decks will be played, and how they interact, and play a deck that will beat the other decks that will do well.
A few examples: In Worlds 2005, I made a terrible deck choice. I went with Rock Lock. Here was my thinking: The BDIF was Medicham. Medicham was pretty hard for Rock Lock to beat. I knew Rock Lock pretty well, and I wasn't that comfortable with Medicham. I was debating what decks to use, and had issues finding a strong counter vs Medicham that would beat the rest of the format, or at least hold up. I expected many of the players to either be afraid to use Medicham due to mirror match, or some archtype phobia ( I LOVE when players refuse to use good decks: please, keep it up ) or that they would be running decks aimed to BEAT it. Therefore, I went with a gamble: Rock Lock lost to Medicham, but I had a number of reasons to suspect it would be a fine choice none the less: I expected Medicham to be hated against, and therefore not do as well as it did at nationals. I expected people to play less of it out of fear. I could afford as many as 2 losses to it. So ASSUMING I didn't get paired against 3 of them, and not be able to win at least one, it was a loss I'd be willing to take. I also knew that against every deck except Medicham, that Rock Lock was by far the strongest choice: I gave up one matchup for the sake of being extremely strong everywhere else. This also meant that any deck that was good against Medicham would likely get torn apart by Rock Lock. I made my gamble. What happened? Medicham got played, and it did well. Nidoqueen, the most successful of the counters to Medicham, convienently did very well against Rock Lock as well. So Rock Lock turned out to be a poor choice. Yet, I feel I was justified in my theory behind choosing it. It didn't pay off, but even looking back, I had no regret about my choice. Muk EX would have beaten Rock Lock, Nidoqueen, and Medicham, the three biggest threats. It also would have done well against ZRE, which Go had used. I feel that of the decks present, he should have done the best, if not won. He made a great choice.
Great players can make bad metagaming choices. Worlds 2007, Jason K used Banette. He is arguably the best,a nd unarguably the most successful player the game has ever seen. It is one of the hardest parts of the game to master. You need to be able to know the format as a whole, not merely your deck. You need to guess what other people will play as well. The more information you have, the better you will be. Metagaming also involves knowing how to tweak your deck accordingly. A great example of this would be my efforts with LBS in 2006. I ran a number of tech cards, and I would switch between them depending on what I felt would be popular at any given event. LBS was extremely hated against, so I had to remain one step ahead of everyone else with techs to keep it viable. This is the case with many decks.
Sometimes, regardless of tech options, and regardless of how comfortable you feel with a deck, and regardless of how powerful it is, a format turns against a deck, and you need to play something else. There are points where your fighting a losing battle to keep your deck viable, when it would be better to switch to a new deck entirely.
Onto the lovely debate of rogue vs archetype. Archetypes are archetypes for a reason: They win. Consistantly. 99% of rogue decks are bad. Unless you are 100% confident that your deck is extremely good, and you have tested all of the key matchups against GOOD players, stick with something proven. It is easy to be misguided about the results from rogue testing: Many players are bad ( quoted Matt Alvis! ). One of the defining differences between an average or bad player and a good one is the ability to adapt. You notice this at Nats, and Worlds often: A player does well against matchups they have tested, but against new decks, or new techs, they play very poorly. The good players will do fine against them. This happen alot with Delta ( Raichu Eggs ) at Nats 06. This is the main reason behind the "Secret Deck" phenomena. Secret decks are rarely the best choices. Yet they overperform because players are bad. An example would be with the deck Jimmy Ballard unveiled at Nats 2007. Jimmy is arguably the best rogue player/deckbuilder in the game. The Empoleon deck he brought in force to Nats 2007, despite not winning, performed rather well. Well enough to test it rather considerably after the event: It did very poorly for my testing team. It turned out to be roughly 50-50 vs Infernape, borderline unable to beat metanite, and terrible against Speed Spread. Yet it tore up most of those matchups throughout the day by surprise, and player inadequacy. This is not a slam against Jimmy at all: Hes produced a number of great decks, and even decks that may not be ideal have been at the very least solid. Sometimes a format has an opening where a rogue deck can slide in and be one of the best decks, but a number of times, much more often than not, this isn't the case. Play a rogue deck when a rogue deck is better than the BDIFs. If it isn't, don't play it to prove something. All it will prove is you made a bad deck choice. It doesn't make you a better player if you succeed with a bad deck.