Scouting vs. Random analysis - Page 2

On April 21 2010 09:32 Chronopolis wrote:
The word Analaysis has turned to anal in the sidebar countless times lol. @ the OP, while the mineral loss for early scout is small @ 100 seconds, if it delays a key building it will slow down the entire build by that much.

On April 21 2010 08:47 junemermaid wrote:
Don't want to be a debbie downer here, but you didn't need integrals and fancy equations to figure this out...

Regardless, cool implications and well done.

On May 10 2010 02:42 OPSavioR wrote:
this is why random is OP

On May 10 2010 05:30 Evoke wrote:
Last night, I scouted a random opponent, and as soon as my SCV got there, he called me a noob. Later I asked him why it was noob and he said it's a waste of time and resources. /confused

On May 10 2010 02:07 ilnp wrote:


you're an egotistical asshole and besmirch others for their research and willingness to share because not only can your tiny brain not even equal a task of the OP's doing, but it can't even figure out how to use this information or that it's even useful at all.

regardless, thought provoking post and i'm glad we can share the same internet space

On May 10 2010 09:25 igotmyown wrote:


No, unfortunately I have to agree. It was an overcomplicated way of showing that
minerals lost = mining rate x ( 100 - time sent out ).

The step of after the integrals
A(t) = f(t) - g(t)
= K([t/20]+1) - K ( [t/20] )
=K
that is, the difference between the rate of mining being down 1 scv is ...
the mining rate of one scv

The integrals are never used, and furthermore, we could simply have added together an arithmetic series if we ever did want to calculate the total number of minerals mined.


What would be more useful is calculating how fast an expansion (gold or not) will pay for itself, which represents your window to be rushed. Or if you called supply instead of a mule to get an expansion faster, how many scvs would you have to maynard to pull ahead of the mule (assuming your main is fully saturated) in income, if it is possible. Or by map how many minerals do you lose maynarding.

Or for scouting, how much money do you save by sending a marine/zealot/reaper scout instead of an scv, assuming they will die. Of course you won't scout cheese...

On April 21 2010 09:15 Koltz wrote:

yeah there would be a couple ways.

the "easiest" would probably be empirical data, but a 30 minute write-up along with half an hour of testing would take much longer than just a 30 minute write-up

another way you can do it is by using differential equations, but thats more advanced than just integration, especially with floor calls

I realize i made it seem very complicated, truth is its just one small integration problem, then using relations between the functions to get what you want. im bad at writeups when im in a hurry.

Don't really know another way i could've made these conclusions without using these methods.

I chose my method because I thought itd be the easiest to express, and this way we get pretty graphs where we can pinpoint things and see relations

not to mention, by using functions its easier (at least for me) to see some of the connections

either way, thanks ^^

On May 10 2010 08:51 Hold-Lurker wrote:
Excuse my ignorance, it's been years since I've done any integrals, but why does the integral of Kt/20 + K equal Kt^2/20 + Kt rather than Kt^2/40 + Kt?

Also, are you using game time or real time? The time to make an SCV should be 17 seconds of game time and roughly 12 seconds of real life time I think. I don't know how that impacts any of the numbers, probably by some constant.