By popular request:

• Buy as a team, always coordinate your buy.
  
• Aim. One accurate shot is better than many sprayed. Tap shoot for the head.
  
• IMO the two above points are the most important, but there are many small but important things in this guide, which can't really be compressed into this section. You may know many of them already, but in my expereince every game I play a fair few of these tips seem to be unknown to several of the players.

• First round buy armour, don't keep your 800 in the bank. EDIT: This one is getting a lot of comments contradicting it, so see the bottom of the guide for an expanding discussion.
  
• Never buy without checking your team mate's cash situation first, if you can't all buy (or drop those who can't) its usually best to save. Communicate and buy as a team. If half the team saves round 3 and half buys you'll be playing a 3v5 for the next few rounds while the guys who did the right thing and saved can't help their team mates. Sure buying alone might get you one or two kills and put you top of your team score board, but in reality you are making it hard for your team to win... you are the problem.
  
• Don't all buy AWPs. Saving for the first three rounds, then buying two or three AWPs is going to lose you the game because you'll probably lose the round and hand the enemy free AWPs in the process. Have only one support weapon on the team, or occasionally (very occasionally) have two. Discuss this with your team and agree before buying.
  
• When saving there is a minimum amount of cash you'll need to have so that next round you'll have enough for a M4 or AK + armour. This varies as the round reward depends on how many you've lost in a row, but buying a deagle or something on a save round is disastrous if it leaves you with less than 3700 you need for an AK and helmet next round, now you are saving twice!
  
• If you are CT and you know the terrorists are using AKs and AWPs then you don't need the helmet. It will do nothing but waste 350.
  
• If you win pistol round then use an SMG but not the P90. SMGs give 600 kill reward and will kill with a head shot on round two because your opponent doesn't have a helmet round two. AKs aren't needed and don't give as much kill reward. P90s only give half kill reward.
  
• Don't buy pistols like Deagle or Duelies. Just don't. The Five-Seven and Tec-9 are great pistols to grab if you are AWPing or sometimes with armour as a semi-save.
  
• Don't buy a gun and no armour. If you can't afford armour you should be saving this round. Without armour you get aim-punched if hit (your aim jumps off position) and you'll lose fights. The only way I can see this being ok is for someone using an AWP at range, or maybe a silenced M4 where you are trying to be stealthy and avoid being shot. But really, you need armour.
  
• Don't grab cheap weapons like the shotguns or the Bizon unless you have a specific reason to want to use them. If you can't afford an M4 or AK then SAVE.
  
• If one of your team is on 14,000 cash then even if someone else has 9,000 or something the richer player should drop for them. You can't earn more than 16,000 so that player on 9,000 will still earn money while the one on 14,000 is not going to unless they spend some. If you lose a bunch or rounds and someone has to eco you'll regret sitting on 16,000 for several rounds and not earning money that could now be used to drop for that player.
  
• On an eco round don't send the players with pistols in first. Send in a player with a rifle, if they die a player following can pick it up.
  
• Don't rebuy armour as soon as you lose a few percent of it. If it drops to 50 then rebuy, but it its at 70 or 80 then control your OCD and don't waste your cash.
  
• Don't buy grenades for no reason. Rightly or wrongly I get the impression a lot of players buy nades because its what pros do, it makes them feel more competant, and seem more competant. Often they get thrown randomly to little effect. Focus on one or two grenades. Early on I decided to buy a smoke and a HE and learn to make use of them. Its better to learn than to grab flashes decoys and molotovs only to throw them one by one in any direction after making a bomb plant. Ideally watch some online guides for smokes and coordinate with another team member on site takes.
  
• Controversial warning for this one! Do you really need a defuse kit or not? Kits aren't cheap, pros need them because the bomb timer is 35, but in competative its 45. Is it worth $400 every round to extend that timer by 5s? You can lose a round and lose a match if it comes down to having that kit or not... but if your team already had two or three kits bought is it good use of your money to buy one too, or better to just grab one thats dropped if it looks like you might need it. Arguments either way, but don't just buy it because "thats what everyone does".

• Only run with your knife when you know you are safe, e.g. round start, or you have information on enemy positions from your team. Otherwise keep your pistol out.
  
• Running makes noise. Walk to stay silent.
  
• Do not walk for a bit, then run a step or two, then go back to walking. You might as well just run.
  
• Do not crouch walk. Ever.
  
• Jumping up is silent, jumping down makes a noise.
  
• If your team is all walking don't run around next to them.
  
• Be careful how you use your nades in a 1v1 or clutch situation. A good nade will give you an advantage to take the other guy down, a bad nade will alert him to your position when he wasn't sure where you were. Sneaky may be a better choice if you don't know precisely where he is yourself.

• Watch corners. Don't hide in an obvious place. The enemy can pop around the corner and spray you before you can respond.
  
• Do not stand close to allies, it just makes it easy for an enemy to kill you both. Worse, do not stand close to a silent ally and then run, jump or shoot. You'll draw the enemy to their position they worked hard to get to without being seen.
  
• Don't stand in front of a sniper unless you want to be team killed.
  
• Don't stand close to an ally at a corner. They may wish to run back and if you block them you'll get them killed. Always give your team room to move.
  
• If you are running through a tight space like a corridor or doorway do not stop because it blocks your team. If you are flashed either move to the side or keep going forward, don't stand still because everyone will bunch up behind you and the enemy will have several easy kills.
  
• Try to stand back from a corner you are watching. If you peak a corner from close to it you become visible to someone around the corner before you can see them. Stand back and make them come close to the corner.
  
• Don't look the same direction your team mates are looking. You need to cover them, watch their backs.
  
• If you are defending a position engage, then whether you get a kill or not fall back to a new position. Once engaged the enemy will know where you are so change position and kill them when they look to where you used to be. Especially useful for using an AWP.
  
• If you kill an enemy from one position they will expect you to be there next round. Change position. A good idea is to move closer to them next time. They may rush to where you were last time and be caught out when you kill them before they expect it.
  
• Be careful your gun doesn't stick out around a corner and give away your position.

• Lower your sensitivity, and keep lowering it until you really can't play any lower. Accuracy accuracy accuracy.
  
• Don't ever remove the silencer from the USP-S or the M4A1-S. Removing it makes the gun worse in every way.
  
• Always aim head height. When you are moving do so with your cross-hair at head height.
  
• Shoot for the head. Unless you use an AWP, then shoot for the body as it doesn't matter.
  
• Low rank players often crouch as soon as they start shooting. Aim for the head, but be aware you may need to adjust down!
  
• Most weapons are not accurate when running or jumping. SMGs like the P90 can be used like this but do not fire an AK or M4 while moving. Stand still, aim for the head and fire. If you shoot while moving, particularly with the AK, you'll miss every shot by meters.
  
• You can fire the scout while jumping. Don't jump and shoot with other weapons though.
  
• One accurate head shot will kill an enemy. Fifteen rounds sprayed around them does nothing. Aim.
  
• At RANGE tap fire. Single shots only! Leave a small pause between each shot. Shoot a wall to learn the timing.
  
• When confident and at MEDIUM distance you can fire a burst, 3-4 shots. Learn recoil control (pull down a little) to do this accurately.
  
• Only ever spray at point blank range. And learn the recoil.
  
• The silenced M4A1-S and USP-S are quiet, but also they don't show up on the map when you shoot (see expanded points for more), and they don't have tracers so the shots are invisible. Use this advantage. Be stealthy.
  
• Pay attention to your ammo and plan reloads. Probably the #1 reason I die in a game is when I'm shot while reloading. If you use a gun like the M4A1-S then you need to be aware of when you have ~6 or less bullets and choose a safe place to reload. At the same time do not reload after every kill. Your instinct will make you want to, but if you have 16 rounds left in the gun you don't need to reload immediately, if you do you are vunerable to the guy who might be just around the corner from the one you just killed and more than able to hear your reload start.

• When you die call out information.
  
• Information is useful in the following order: enemy position, enemy weapon, enemy action (looking X direction, running towards Y location), damage given to enemy.
  
• Don't yell into your mic when you die. Don't talk constantly in the mic. Doing this prevents your team hearing crucial information and will get them killed and lose you the round and maybe the game.
  
• If you are watching a location your team will assume they are safe from that direction. If you can't stop an enemy getting past you at least call out to your team. They won't be happy if they get shot in the back because you didn't tell them something.
  
• Make sure you call out the number of enemies you see and if they do or do not have the bomb. This is crucial information your team needs to decide if to rotate and which route to take as if lets them know where enemies are likely to be on the map.
  
• Enabling music lets you know when the bomb timer reaches 10 seconds as the pace of the music changes.
  
• When you die you can see for a few seconds what happens. So if you kill a player and immediately run to hide, they will be able to inform their team where you went to. Throw them off by running one direction, then turning around a couple of seconds later.

This section is added on request and is more general advice which you have to adapt for yourself, rather than simple dos and don'ts.

• You can't control your team, but you can encourage them. If a team member makes a mistake its hard not to yell/insult/groan... but this can make them feel frustrated and they'll stop concentrating on the game. They'll probably feel under pressure in a 1vX and they might even try to TK you.
  
• On my friends list there are guys who "know" they are superior to you and will let you know it every time they don't like how you play. Then there are others who will say things like "nice try man, bad luck" even if you made a real stupid mistake. The second kind of person makes me more relaxed, more inclined to listen to their advice and confident that I'm going to get it right next time.
  
• Try saying "oh really bad luck there man, next time you could do X" and see how people respond.
  
• In short always be encouraging, you can't correct all the many hundreds of mistakes your team will make (and you'll be making mistakes yourself), but a player just needs encouraging even if they are doing something wrong. Don't call them on it, let them ask you for help, or offer it very sparingly if you want them to listen.
  
• CS:GO lets you choose the map you play, so make use of this! Pick one or two maps and play them repeatedly, rather than playing a different map each time. Pick maps like Dust 2, Cache or Mirage if you are solo queuing, rather than Nuke or Inferno because Nuke and Inferno typically require a good coordination on the T side to get any rounds. Solo queues are often less coordinated and often faster to become frustrated and give up just because they lost 5 rounds as T on a CT sided map. Your mileage may vary with this one as different maps appeal to different players.

In this section I just add some more discussion over some of the points people have disagreed on. I'm not saying my answers here are the only right answers, but I hope to explain why I think you should do X and not Y. Remember that this advice is all geared towards players who are new or who want to improve out of Silver/Gold Nova, and therefore need to focus on mastering basics. A high ranked player would probably do something different, harder to pull off... more nuanced. But don't try that until you have mastered the basics and you are also a high ranked player.

• First round buy armour
  Why? First round you can buy armour, a pistol, nades, or save. Which to do? Buying nades requires a coordinated team strategy to be most effective. If your team can do that then go for it, usually this isn't the case in silver/GN solo queue though. A better pistol then? You can do more damage this way, but you leave yourself vulnerable to aim punch if the enemy gets a shot on target. For lower ranks I recommend getting armour to stop this and focusing on improving your pistol aim. A head shot will kill an opponent first round because they have no helmet, so you don't need a better pistol if you can aim. Get armour to help with that aim and practice. Finally, why not save? IMHO the first round is the most important round of the match. Its the only round that everything is equal. If you lose this round you often lose the next one or two as well, you really want to win the first round. Therefore why wouldn't you throw everything into winning this? Having a helmet second round doesn't help if it risks you losing first round, if you won first round you'd be better off.
  
• Silenced weapons and the radar
  I had watched a YouTube tutorial that claimed the silenced weapons mean you don't show up on the radar when shooting, and I never questioned it but it now seems it may be wrong. Searching around most sources say that you only show up on radar if you are seen by an enemy (some saying within a 90° FOV. However, some sources also say that non-visible enemies don't show up unless they make sound, which then seems to contradict the previous statement. Its unclear if sound means shooting or steps or both. Therefore I've removed the statement about the silenced weapons. Of course they are still less likely to be detected because enemies will be less drawn to their positions. I could jump in game and check out precisely what sets the radar off once and for all but I don't have the time right now.

"I have now 9 competitive wins in row and I didn't rank up (silver 2 rank with 160 competitive wins). Any specific reason i don't rank up?"

This is a pretty common complaint. The short answer is: 'we don't know' because in any specific case we cannot see the calculations that were used to determine rank. Now for the longer answer:

Rank is determined by a mathematical system usually referred to as "the ELO system". A very brief description of this would be:

• Rank of players before the game is used to calculate the "expected outcome"
  
• Actual outcome is compared to expected outcome
  
• Depending on how close they match ranks are adjusted
  
• For CS:GO its unclear if a "game" in this context is a "round" or a "match". I can see some problems with using rounds, specifically that 16-0 games would then require many more 14-16 games to balance out. Also rounds don't start equally as previous rounds impact the buy a team can do.

This system is designed for a two player game, specifically chess. Applying it to a multi-player game results in complications, since now we don't know how much of the outcome was the result of each players actions. You can read about how this is applied to gaming in this Microsoft Research paper on their "Trueskill" system: http://research.microsoft.com/en-us/projects/trueskill/details.aspx

So what can we learn from this to understand why we do or don't change rank?

• Simply knowing the results from a small number of games doesn't help much. If you were expected to win those games and you did, then your rank won't change by much. A game you are expected to lose but you win might have much more impact on your rank.
  
• That said, a moving average of your game results is what determines your rank, we just don't know how significant each result is compared to the other results.
  
• Given that there are 10 players in each game, you are contributing only 10% to the result. Therefore it is expected that changes over time will be slow, which is probably why there are so many ranks compared to say Starcraft 2, because small changes are actually quite significant in CS:GO.

I personally started tracking my game results in a spreadsheet similar to one posted on reddit, and I just passed the 100th game in that sheet (I have over 500 wins though, so its only a fraction of my results). I've changed rank three times in that period, and each time it was within a game or two of my moving average Win/Loss going positive, actually in all three cases I had just hit a win rate of 67% over the previous 10 matches. I arbitrarily picked a 10 and a 20 game moving average to plot. I can post the sheet somewhere if people want to see the data.

EDIT:
Link to my spreadsheet. You can probably make a better one but this is what I've been using.
https://docs.google.com/spreadsheets/d/1sD_ev1bT8demVT4ZwWOHm8uFmYPVPJehL64GcJaTvzs/edit?usp=sharing

I don't believe looking at your wins is very important. Any method that tries to introduce an systematic way to improve at a game pretty much ignores match outcome and instead focuses on improving a limited part of your game. You just can't do everything at once. My personal benchmark right now is my position on the scoreboard. This is also an imperfect assessment, but since it factors in several things in an attempt to show "player value" to the team it is quite useful to look at it over the course of many games (that last part is critical). If you are constantly near the top of the scoreboard then you can expect your 10% contribution to match result to be positive, and therefore that you will win a bit more than usual and therefore rank up.

Currently I'm (based on 28 games since last promotion):
32% rank #1
29% rank #2
25% rank #3
14% rank #4
0% rank #5
So I'm hoping for a rank up soon. If your rank was accurate, then over a sufficiently large number of games you'd expect to see these all be 20%. A skew up or down should indicate a promotion or demotion if you can maintain that skew.

Of these last 28 games I've won 46%, lost 39%, and drawn 14%. As previously noted we don't know which of these games I was expected to win, to lose, or were equal chance of outcome... so the Win/Loss itself doesn't help much. However we can note that despite being significantly skewed to the upper end of the score board that hasn't (yet) translated into a large increase in wins for me. Due to the 10% impact on game result effect random chance dominates and it takes a lot of games to see an impact. And of course any time there is a smurf, or any player with an inaccurate rank, the game will incorrectly calculate the expected outcome and therefore incorrectly adjust ranks at the end of the match. Smurfing is always bad folks.

Also as a player improves and wins games they get progressively harder games, even before they are promoted. The result is that again Win/Loss isn't the full picture because a set of losses or draws might be the result of moving up and stabilising at a new skill level. In fact my most recent promotion came after a draw, my moving average had met the level for a new skill group.

I have thought about, but not yet tried, doing some maths on how many games it would take to see a significant improvement in your win/loss given a 25% (for example) improvement in skill, taking into account your 10% contribution to the match, and factoring in random variation in skill for all players (day to day variation and random luck). I might have a go at doing that soon.

Finally, don't focus too much on rank. I know its hard, but I suggest its better to focus on playing your best, as you'll enjoy it more if you play better... and as a side effect you'll probably rank up.

My Clan (we are recruiting):
http://www.vortex-esports.co.uk/forums/
UK Based but members from all over EU and NA, Teamspeak, Forum, In-house torunaments and more.
If you are interested in joining a clan check us out, put my name as referer (Spirit).

I stream occasionally in 1440p or 1080p.
Hitbox: http://www.hitbox.tv/VRTXSpirit
Twitch: http://www.twitch.tv/vrtx_spirit