|
Get the red hot worker calculator spreadsheet off Google Docs here:
Starcraft 2 Worker Count Caluculator
If you watch one of your own replays and fill out this spreadsheet as you go along, it would be totally awesome if you would upload it somewhere and post a link in the thread, along with your current league, so I can put together some kickass charts of typical macro at various league levels.
Background
After reading yet another request for help with replies that said "your macro is bad; go fix it and you would have won this game" without any explanation of "what is macro?", I got annoyed and decided I should do a little something to help. So I made a spreadsheet to help players track their "probes and pylons" worker production.
Why does this matter?
Making workers is obviously not all there is to macro, but it's the basic bench press exercise of your macro workout regimen. In order to keep on top of your worker production, you have to (a) avoid getting supply blocked, which happens at all levels of play, but happens more at lower levels of play, and (b) make sure that you're checking in on your Nexi/CCs every 17 seconds. And to restate the obvious, if you can produce more workers than your opponent, as the game goes on you'll be able to produce a bigger army than your opponent, reinforce that army much faster, have more resources left over for upgrades, etc.
How to use this calculator
Download the spreadsheet as an Excel file, or save it to your own google account. Watch your replay, keeping note of a few things:
The quick version (Terran only) As you're watching your replay, take note of when you start your expansion, and watch the Units tab (U) or the Army Leaderboard (CTRL+A) to see how many workers you have on the field. Follow along and see how close you're staying to the values in the Ideal Worker Count column. When you start building your expo, update the value "start time of expansion build" with the actual value from your game. If you built your expansion in-base, set the "CC float delay" value to 20; otherwise, leave it at 0. If you lose workers due to economic harassment, you're going to be about two minutes behind ideal production for every seven workers you lose while on one base, and one minute per seven workers lost while on two bases. (e.g. if you lose 14 workers after your expo is up and running, at the 12 minute mark you should be at the 11 minute target, etc.)
The thorough version Watch your replay with the spreadsheet at hand, and keep track of the following.
- The number of workers you have at the start of each minute. You can get this from the Units tab (U) or the Army Leaderboard (CTRL+A) Fill in the "Actual Worker Count column" with these values.
- When you start building your expansion. Replace the value under the "Start time of expansion build" with the actual value from your replay.
- The total number of workers you've lost at the start of every minute. You can use the Units Lost leaderboard (CTRL+R) to see the number of worker lost.
- (Terran only) How long it takes to float your expansion to the natural. Do not count any time your expansion sits idle; count only the time it takes to lift, move, and land the CC/OC. Replace the value under "CC Float Delay" with this value
- (Protoss only) when you use chronoboost on probes. Keep a running count of the total number of times you use chronoboost on probes at the start of each minute. Replace the values in the "Cumulative Chronoboost used on probes" column with these values. If you use more than 10 CB on probes, you need to update the value under "max Chronoboost used on probes"
When you've done all of this, the "Difference" column will show you how far you are behind ideal worker production.
After you've gone through this exercise this a couple of times, you will start to have a decent sense of how many workers you should have at any given time (assuming no losses due to economic harassment), and you will begin to feel how close you are to ideal worker macro without tracking all of the details.
I've found that simply watching at least one replay per session with an eye towards worker production has forced me to pay more attention to worker production and avoid supply blocks during every game. If it's been a while since you focused on macro, using something like this is a really good way to see if it needs to improve.
FAQ:
Why is the default target 66 workers?
66 Workers gives you 2 per mineral patch and 3 per geyser on 3 bases. This is "optimal" in the sense that the 67th worker has a much smaller impact on your income than any of the first 66 workers.
Can I change the target number of workers?
Yes! I've heard of people using 72 workers as a target as well; you can adjust the target in the spreadsheet. A Protoss player needs more workers than a Terran player in order to have the same income, thanks to the Imba Mule (TM). Another reason you might want a different target is if you're going for an "all-in" where you stop building workers around 20 probes, and instead use the resources you would have spent on resources to have a bigger army for a brief window of time. The most common example of this is the 4gate, but there are others (2-base 6gate +1 attack, etc.). Unless you're going for one of these specific timing attacks, you will probably be better off in the late game by constantly building workers until you can get a full 3-base economy.
Why isn't there a Zerg calculator?
Because Zerg Is Different (TM). Depending on how the game is going, a zerg player may spend larvae on drones or army at any given point in time. So making a zerg calculator is a bit more complicated. I'm happy to take suggestions if anyone has them.
I'm silver league and I thought I had pretty good macro. But even with a fast expansion I was still 15 workers behind ideal production at the 10 minute mark. That's strange, I thought I was doing pretty well. What happened?
It's impossible to tell without watching the replay. Go back and watch it again. Did your worker production slip during a fight? Maybe you should queue a little at the start of the action. Did you get supply blocked, even for a few seconds, and not realize it? Build that depot/pylon a little earlier. Were you spending money on units/structures instead of workers? In some cases (e.g. if you're under attack and need the units), that might be the right priority, but think about whether or not that was a good judgement call.
What does typical bronze/silver/gold/platinum/diamond/masters/grandmasters worker macro look like?
I'm high platinum, and when I'm playing well, I am about 4 workers behind a 66-worker target. When I'm playing poorly I'm 10-12 workers behind the target. That said, I punch way below my weight when it comes to actual fighting, so my macro is probably good for a platinum player. I encourage others to watch their replays using this spreadsheet, and post how close they are to ideal macro, so that I can build up some league-level benchmarks and put them in the OP. Bonus points if you make a completed spreadsheet and share it out on google docs (or megaupload or whatever), since I can use that to produce pretty charts.
|
Thanks for this. Will give it a try, see how I'm holding up to the benchmark. I feel like im doing ok, bit after a game I see my worker count and cry in silence  maybe this helps me point out easier where I go wrong.
|
whilst a lot of people flame the maths at not being important, I feel this would produce some interesting statistics, graphs and theories if this is followed up with peoples data input. Will be interested to see how this works out!
|
Imo building a good amount of workers isn't the best thing to dwell on, especially for lower level players (bronze - gold especially).
Instead of looking at how many workers you had, why not just find out what/when/why about more important forms of bad macro? For example, instead of counting workers you could find out if you ever got supply blocked or ever stopped building units.
Of course, this idea is nice for people who really like to be picky about their habits. Honestly though, I think practice of any form is better than taking notes about how close to ideal you got with your worker production.
|
On May 24 2011 06:48 noobinator wrote:Instead of looking at how many workers you had, why not just find out what/when/why about more important forms of bad macro? For example, instead of counting workers you could find out if you ever got supply blocked or ever stopped building units.
There's nothing wrong with doing that. But personally I've found that it's hard to get a sense of how much each little bit of idle production/supply block adds up over time. Also, usually watching a replay with an eye towards worker macro causes me to see the other holes in my macro.
|
I've never gone with the whole "constantly" build workers thing... at least not since the beginning. I build workers when its safe, which is most of the time, but im in masters. I dont really care for the mantra and I think alot of people probably lose games by building workers constantly when they need to cut workers and so on, but overall I understand thats its much easier to say "always build workers" than explain the details about when to actually do it.
|
So in an ideal game for Protoss you should accept Terran has more income due to triple MULE (and same workers)? Meh, I like to be ahead of Terran in harvesters otherwise the difference in income is getting greatly out of hand.
This might be good for newer players, but I do hope they do not lose focus of other aspects of the game and simply make workers constantly without responding properly to an all-in by their opponent, for example. I guess it's OK to have some goal and all, but the truth is it is totally different each game 
|
On May 24 2011 07:02 Jayrod wrote:
I've never gone with the whole "constantly" build workers thing... at least not since the beginning. I build workers when its safe, which is most of the time, but im in masters. I dont really care for the mantra and I think alot of people probably lose games by building workers constantly when they need to cut workers and so on, but overall I understand thats its much easier to say "always build workers" than explain the details about when to actually do it.
On May 24 2011 07:03 dani` wrote:
So in an ideal game for Protoss you should accept Terran has more income due to triple MULE (and same workers)? Meh, I like to be ahead of Terran in harvesters otherwise the difference in income is getting greatly out of hand.
Well, P can always spend chrono on probes, and you can set a higher target than the one I have in there by default. Also in the late game Terran is usually saving scans to keep an eye on the battlefield.
On May 24 2011 07:02 Jayrod wrote:I've never gone with the whole "constantly" build workers thing... at least not since the beginning. I build workers when its safe, which is most of the time, but im in masters. I dont really care for the mantra and I think alot of people probably lose games by building workers constantly when they need to cut workers and so on, but overall I understand thats its much easier to say "always build workers" than explain the details about when to actually do it. On May 24 2011 07:03 dani` wrote:This might be good for newer players, but I do hope they do not lose focus of other aspects of the game and simply make workers constantly without responding properly to an all-in by their opponent, for example. I guess it's OK to have some goal and all, but the truth is it is totally different each game
One of the FAQs talked abut the fact that prioritizing army production over workers might make sense when you're under attack, but I fleshed that out a little bit, and added a couple of sentences or two about probe-cutting all-ins.
Also stay tuned to the map forums ... I'm working on a map that keeps detailed macro stats on all facets of macro (idle production buildings, supply block, wasted CC/Nexus/Queen energy, larva injects, etc.), though that may take a little while :|.
|
Don't forget squats and deadlifts in your basic workout regimen
BTW, in a perfect world I'd like to have 4 bases, one of which is a gold.
3 workers/geyser * 8geysers = 24 workers on gas
16 workers on minerals at the nat and third, 12 on the gold (6 patches), 8 on minerals at the main since it will be mined out soon = 52 workers on minerals
= 76 workers target
If you can get a fifth, take 6 workers from your 95% mined out main and put them on gas at the new expo, don't make new workers.
|
On May 24 2011 07:03 dani` wrote:So in an ideal game for Protoss you should accept Terran has more income due to triple MULE (and same workers)? Meh, I like to be ahead of Terran in harvesters otherwise the difference in income is getting greatly out of hand. This might be good for newer players, but I do hope they do not lose focus of other aspects of the game and simply make workers constantly without responding properly to an all-in by their opponent, for example. I guess it's OK to have some goal and all, but the truth is it is totally different each game
The games where I feel like I totally crushed a terran in economy, I'll have 20% more workers than him and expand faster, but when I look at the endgame stats we had the same # of resources mined.
So yeah, you should accept that if both players are equal, the terran will have more income.
Stupid imba MULES 
|
silver league player, i have very good macro (for silver league) but micro sucks
that aside, great thing, just find it almost impossible to get 66 by 17min into a game. This style is if nothing attacks, almost nothing distracting, getting perfectly timed bases and never missing a probe or scv. Chronoes definitely need to be factored in the fact used elsewhere(warpgate, upgrades) and you sometimes need extra workers(building buildings, proxy pylons, usually use probes to take xelnagas) so you should factor that in the spread sheet too. Otherwise this is a great guide to go by and is very helpful
|
Expand faster, 66 probes @ 17min is super easy if you're allowed to expand. Your 4th should be up.
|
spreadsheet looks interesting, maybe include a block of times for cutting workers (say from X to Y minutes) so if your build is supposed to cut for a time you aren't penalized.
Also, chronos add around half a worker, it would be easy to add a cell for chronos on probes by XX time to account for that.
|
On May 24 2011 13:15 friendo wrote:
Also, chronos add around half a worker, it would be easy to add a cell for chronos on probes by XX time to account for that.
I think this is covered by the "Cumulative Chronoboost used on probes" ...
As for cutting workers at specific times in the build, that's not a bad idea. In the short run you can fake it by treating them as "workers lost", but I'll add a column for cutting workers since that's an intentional part of some builds.
Also I'm going to embed a lot of the instructions in the spreadsheet.
|
I dont really understand the use of this spreadsheet. It seems to me, that it would be much easier to just select your CC/Nexus, go through the replay at times 8 and watch if the workerproduction ever stops. Putting all this data into the spreadsheet seems like a ton of work to me.
|
I make 100+ probes in both PvT and PvZ because I expect to lose about 30~ to bling drops/infestors/marine drops/blue flame helions/BC all-ins
|
|
|
|
|
|
EPT
