|
So... after reading some threads on when to stop worker production I couldn't find any calculations on Protoss units, so I used ix's data to come up with this spreadsheet:
https://docs.google.com/spreadsheets/d/1q3gHinLLnWLIFbfKi7-sRiI-uLa22akyrPA7mElUisk/edit?usp=sharing
(edit) I put together calculations for the other races as was asked by Cryoc
(edit2) Corrected major mistakes in my timers (mixed both normal and fastest game speeds). Everything should be fastest now
(edit3) Corrected spawn timer for Larva thanks to Cryoc. Liquipedia article had it wrong
(edit4) Put the morph timer in parenthesis for Lurkers and Guardian/Devourers, as tested by Cryoc. Liquipedia was confusing
Hopefully it helps some people.
Threads I read before coming up with this:
+ Show Spoiler +
|
Nice work. Could you maybe add the units for Terran and Zerg too? I know in that one thread they talk about it, but there is no organized information like you did for Protoss.
You should probably also include pylons in your calculations as your units need supply.
|
On September 04 2014 21:34 Cryoc wrote:
as your units need supply.
*Psi
Just sayin'
Have a good day
|
love the thread title. sounds like a newspaper advertisement for new labour
|
|
|
On September 04 2014 21:34 Cryoc wrote:
Nice work. Could you maybe add the units for Terran and Zerg too? I know in that one thread they talk about it, but there is no organized information like you did for Protoss.
You should probably also include pylons in your calculations as your units need supply.
Good points. Also, lol @ above :D
|
Added the other races and made a bunch of corrections/improvements.
|
|
|
Nice, thanks.
I noticed some mistakes in the Zerg section. The spawning time for the larva is wrong. It should be 20*0.65. The larva Liquipedia article had a mistake there. Also if you include the build cost of a muta into the guardian/devourer you have to include the muta build time as well for the calculation, same thing for the lurker.
|
United States2948 Posts
Very interesting spreadsheet, Scarbo. Would you mind discussing your analysis? It looks like the workers needed column are the total (not per mineral patch or gas geyser) number of workers necessary for each unit to be produced, taking into account the minerals/gas consumed during their production process?
Just some random thoughts, not aimed at anyone in particular:
How do these numbers compare with ix's and others saturation/wandering numbers? For example, with one base and saturated workers on ~9 patches (or whatever standard number), how effective is that -- how many of each unit can be produced per given time? What would be a sample ground force composition look like?
How does the saturated worker model compare to production with fewer workers (like the ~1.5-2 workers/patch I sometimes see pro-gamers use)? And with some expansions, one base vs two vs three, etc. bases, how close does a player get to enabling the constant production numbers calculated in the spreadsheet?
|
On September 05 2014 23:32 Cryoc wrote:
Nice, thanks.
I noticed some mistakes in the Zerg section. The spawning time for the larva is wrong. It should be 20*0.65. The larva Liquipedia article had a mistake there. Also if you include the build cost of a muta into the guardian/devourer you have to include the muta build time as well for the calculation, same thing for the lurker.
Corrected the timer for larva. Thanks!
About the build time for morphing units, well, for one they aren't in the calculation because I'm using the larva spawn timer and morphing an unit from another one doesn't interfere with that so it doesn't apply. As for the number in parenthesis I was going to include both timers but as stated in the notes section below the liquipedia articles are confusing so I decided to let it be.
On September 06 2014 04:34 prech wrote:
Very interesting spreadsheet, Scarbo. Would you mind discussing your analysis? It looks like the workers needed column are the total (not per mineral patch or gas geyser) number of workers necessary for each unit to be produced, taking into account the minerals/gas consumed during their production process?
Just some random thoughts, not aimed at anyone in particular:
How do these numbers compare with ix's and others saturation/wandering numbers? For example, with one base and saturated workers on ~9 patches (or whatever standard number), how effective is that -- how many of each unit can be produced per given time? What would be a sample ground force composition look like?
How does the saturated worker model compare to production with fewer workers (like the ~1.5-2 workers/patch I sometimes see pro-gamers use)? And with some expansions, one base vs two vs three, etc. bases, how close does a player get to enabling the constant production numbers calculated in the spreadsheet?
My analysis was a starting point for Protoss players to calculate how many workers they need for constant production of something. I was wandering around threads about this and saw sample ground compositions for the other races so I decided to do that for P.
The number I used for minerals mined per minute per Probe were taken from xi's calculations (like it says so in the notes). It's assuming 1 worker per patch so it will decrease in efficiency per worker added. It may be possible to come up with a better number, it's just that this subject has not that much info about and as I said I just wanted a quick Protoss spreadsheet to work with.
Not sure how does the saturated worker model compares since xi only talks about maximum saturation (zero wandering) which goes from 2.4 to 3.3 workers per patch, depending on the patch disposition. I don't think you should be fully saturated ever in a real game so I'm not going to calculate anything using those. I may look around those threads again and try to find info on that but for now this is the reason of what I did.
|
On September 06 2014 07:08 Scarbo wrote:Show nested quote +On September 05 2014 23:32 Cryoc wrote:
Nice, thanks.
I noticed some mistakes in the Zerg section. The spawning time for the larva is wrong. It should be 20*0.65. The larva Liquipedia article had a mistake there. Also if you include the build cost of a muta into the guardian/devourer you have to include the muta build time as well for the calculation, same thing for the lurker. Corrected the timer for larva. Thanks! About the build time for morphing units, well, for one they aren't in the calculation because I'm using the larva spawn timer and morphing an unit from another one doesn't interfere with that so it doesn't apply. As for the number in parenthesis I was going to include both timers but as stated in the notes section below the liquipedia articles are confusing so I decided to let it be.
Yeah, you are right, the numbers seemed so high, I thought the pre-state has to be integrated. Zerg can really dump a lot of resources really fast. Seems like the cocoon page used a rounded number for ingame seconds (40*0.65~25). I tested it ingame, the basic value is 40 for morphing guardians/devourer, same goes for morphing lurker.
|
i dont get what the excel sheet could tell me. what is the magic number where you should stop probe production? 
|
It is a good estimation for your needed worker count, if you want to exploit some sort of timing window and don't know when to stop worker production for it.
|
1224 Posts
This is actually a resource that has been lacking for ages, so thanks for compiling it all in a great format Scarbo!
If you ever want to design build orders maybe for map specific timings this is a helpful resource
|
On September 06 2014 18:34 Cryoc wrote:Show nested quote +On September 06 2014 07:08 Scarbo wrote:On September 05 2014 23:32 Cryoc wrote:
Nice, thanks.
I noticed some mistakes in the Zerg section. The spawning time for the larva is wrong. It should be 20*0.65. The larva Liquipedia article had a mistake there. Also if you include the build cost of a muta into the guardian/devourer you have to include the muta build time as well for the calculation, same thing for the lurker. Corrected the timer for larva. Thanks! About the build time for morphing units, well, for one they aren't in the calculation because I'm using the larva spawn timer and morphing an unit from another one doesn't interfere with that so it doesn't apply. As for the number in parenthesis I was going to include both timers but as stated in the notes section below the liquipedia articles are confusing so I decided to let it be. Yeah, you are right, the numbers seemed so high, I thought the pre-state has to be integrated. Zerg can really dump a lot of resources really fast. Seems like the cocoon page used a rounded number for ingame seconds (40*0.65~25). I tested it ingame, the basic value is 40 for morphing guardians/devourer, same goes for morphing lurker.
I put that in there. Thanks for testing.
|
|
|
|
|
|
EPT![[image loading]](http://i.imgur.com/yoQnQMS.jpg)
![[image loading]](http://i.imgur.com/XdfvvJL.jpg)
