The theme of this post is thinking strategically about the larger picture. Even if you disagree with how I play, at least you will have to think about your own strategy.
1. While you are doing things, your enemy is also doing things. This obvious observation is nevertheless thought about by a minority of Dota players.
If you abandon the off-lane to jungle because you will get more farm that way, you aren’t only changing the situation for yourself. The enemy safe-lane is now free farming. The enemy supports are now free to chain-pull for their own development, or harass your mid-lane and safe-lane. Even if you are left alone in your jungle, your team may not have gained from your decision. It’s usually preferable for you to contest the off-lane and suffer for awhile.
Let’s say you are a support and decide to farm a midas, so that at 40 minutes into the game you will have items. Disregarding the resources you have drained from your team, has it occurred to you that all the while, your opponent is farming too? Maybe a better way to secure the late game is by focusing on securing the early game.
2. Farm distribution should be done strategically. During the laning phase there’s no issue; the three “cores” take the farm. But what about after the laning phase? What about when there’s a creep wave at a tower, who gets it? It surely should not be decided by whoever is greediest and teleports first.
The 1,2,3,4,5 numbers we use to denote roles originally referred to farm priority. 1 would be your late game hard carry, who got first dibs on the best farm. Sometimes your 1 would be your mid-laner. It’s not because a hero laned mid that he is the 2; it is his rank in the carry hierarchy.
My point is, farm distribution should be determined according to strategy. It should not be a free-for-all.
Sometimes a support will be close to an important blink dagger, and you’ll give him a lane to farm. It just means you temporarily prioritized his farm.
If you watch some high level games, sometimes you will see supports use spells to weaken a wave for the cores. If you watch some low level games, often you will see the opposite!
If you watch a pro-level tinker, notice how he will leave farm for his teammates, especially his carry. Lesser tinkers (even divine/immortal level) will instead farm everything available, with a preference for the safest farm!
3. Lane direction. Generally, if you want space on the map to do things, you need to first ‘fix’ your lanes so that they are pushing away from you. If you smoke when there are creep waves knocking at your base, you are fooling only the dimmest opponents. If you do successfully catch someone, you won’t have an objective to take anyway. Smokes should generally facilitate objective taking.
4. Economy of movement. This is something that separates different skill levels, but is rarely credited. If you watch a low level carry farm, there’s probably a lot of pausing between camps and waves. A high level player has his whole farm pattern mapped out smoothly. Smokes in pro games are typically prepared in advance, with a clear plan in mind. In RD2L, it might take 20 seconds to bring a smoke out and organize the team, and it’s a cross-your-fingers-and-hope-you-catch-someone-to-win-the-game smoke. When smoking without a real objective (such as a tower push), I think RD2L teams lose more in opportunity cost than they gain by smoking.