If you carefully observe your workers mining, you will notice that they sit around for half a second after they mine a mineral before realizing they need to return it to your CC. I'm not sure why the AI is slow on this, but the delay is actually not necessary. By removing this delay, mining speed is increased by up to 7%.
Select a worker that just finished mining some minerals. He should have the 'Return Cargo' ability, which has the hotkey C by default. Hold shift and right click back to his mineral patch. Then as quickly as possible press C then right click and repeat both of these until the worker reaches the CC. The last command issued should be right click. Release shift, and repeat this with another worker.
The reason this works is quite simple. You are just queueing up many copies of the commands "Gather Minerals, Return Cargo, Gather Minerals, Return Cargo ...". Then rather than waiting for the AI to tell the worker to return cargo, the worker follows your instructions and returns the minerals immediately. However, if you end with return cargo, your worker will just wait at the CC after he goes through the whole queue. Hence, you should either end with right click, or keep extending the queue.
Select a worker that is about to return minerals. Hold shift, and then hold C. Just spam right click on the appropriate mineral patch.
Holding C after shift effectively spams the Return Cargo command very quickly, as with making units. This method requires less effort and is limited only by your clicking speed. However, the extra 'C' presses will fill up your queue and your maximum amount of commands will be more limited. Also, often the last command will be Return Cargo, so make sure you add a right click at the end of the queue.
This is more of a specialized method in order to get the worker to return quickly on the first trip (you can't execute the above methods until after they return). After a worker goes to mine, just shift+right click his mineral patch. It will make the worker run immediately back to the CC once his mining is done and continue mining. However, without using one of the above methods, the later trips will not be boosted.
The main use for this method imo is on mules. On some close patches, the mules will return an extra 30 minerals, but only if you use this trick on the first return (and boost the rest of the mining as well).
I tested this trick by mining with two workers (so I could do the trick perfectly) and counting the number of minerals returned after 5 minutes. This was done for both close patches and far patches. This data is also used to determine how much time this trick saves from each mining trip.
Minerals after 5 Minutes
Close Patches with Normal Mining: 495
Close Patches with Boosted Mining: 530
Percent Increase: ~7%
Far Patches with Normal Mining: 440
Far Patches with Boosted Mining: 470
Percent Increase: ~6.5%
Round trip mining time (seconds)
Close Patches with Normal Mining: 6.74
Close Patches with Boosted Mining: 6.25
Time decreased: 0.49
Far Patches with Normal Mining: 7.69
Far Patches with Boosted Mining: 7.14
Time decreased: 0.55
Thus, this trip removes about 0.5 secs from each trip, and gives up to a 7% increase in minerals.
This trick is difficult to use optimally as I mentioned. After trying several methods, the best way I've found to use this trick is to start with a 3/3 split. One worker will arrive (and hence will return) first for each group. Select the first worker ahead of time and quickly spam. Then try to get the other two workers of that group of three. Repeat on the next return trip for the other group. Then keep extending these queues as much as possible, while boosting the newly arriving workers as well.
Once the workers start running around looking for patches, this trick doesn't really boost mining rate anymore since they will keep trying to return to the same busy patch. However, careful pairing up of workers can reduce this problem, and I don't think it actually hurts. Eventually there is no point to this trick however, because the actual mining time becomes the limiting factor.
I've found that by using this trick, mules mining from close patches will return 300 minerals instead of 270, which seems worthwhile for your first few mules. Some far patches normally only return 240, so this trick will give the full 270 in these cases.
Using this trick early game can save seconds of your build. For low level players, this may not be worth the hassle, but I think it is definitely worth it at high levels. To summarize, here is a list of pros and cons.
- Cuts several seconds off your build
- Only requires actions when you have them to spare (very early game).
- Instead of rerallying and boxing workers, you have to do useful tasks.
- Making mistakes and not correcting them can cost mining time instead.
Q1: Does this work on gas?
A1: No. There is no delay before gas is returned.
Q2: Can't you just spam/hold C to get your workers to return your minerals?
A2: I can not get this to work at all. Unless someone posts a video/replay of this working, I conclude that this method is worthless.
Q3: Pros with over 9000 APM will be able to use this throughout the game to always get +7% minerals.
A3: Not at all. This return only reduces the return time of the minerals, not the actual hands-on mining. When the minerals are saturated, the mining rate is limited by how long each scv takes to mine a patch. The travel time has no effect. Personally, I think if you have over 16 workers, attempting this trick will always cost you minerals instead. However, testing is necessary.
Q4: What is Method 3 for?
A4: Method 3 is for making workers return quickly on their current mining round, while the other methods are for queueing up for future rounds. They can be used in combination for the maximum mining speed.
Q5: How long should you keep this trick up for?
A5: As soon as workers start switching between patches, you can easily lose minerals because workers will keep trying to go to busy patches they are queued for. Additional micro can prevent this, but it will probably still hurt the effectiveness of this trick. I stop around 12 workers, and I caution against trying this with more without verifying you are not costing yourself more than you are gaining.