|
That is the question. Or is it a question at all? While Stimpack is considered to be the most important upgrade for Terran bio, one might wonder, is sacrificing HP to gain DPS always worth it? As LoneYoShi asked,
On August 22 2014 22:44 LoneYoShi wrote:
(might be stupid, but I want to make sure): is there a situation where it is better for a terran to not stim his bio units (extra health vs extra DPS) ? In all games, we see pro terrans always stim, but are there situations where this is the wrong choice ?
As it turns out, the question is not stupid at all, in fact very valid. To answer it, we once again use Lanchester's Square Law as a model. If you are not familiar with it, check it out here, otherwise the calculations below may be hard to follow.
Open the spoiler tag for detailed calculations of Stim effectiveness, or view in pdf here.
+ Show Spoiler +
TL;DR: The core idea is to calculate if the Marine/Marauder will trade better with Stim than without. We can do this calculation for an 55, 54, 53, ... HP Marine, and so we have a function of how "good" stimming is at different HP values. It turns out that there is almost always a point where using Stim actually hurts more than it helps, so it's not beneficial (well, except Marines vs. Thors). Where this point is depends on the upgrades and enemy unit type. At maximum health, though, it is always advisable to Use Stim, except against Banshees and DTs, appearently.
So there you have it, LoneYoShi  . As always, I am open to criticism, opinions and new ideas for new projects.
Check out my previous works as well:
+ Show Spoiler +
|
Terrans you heard it. DONT STIM
|
Stim doesn't just let your units have more dps, it also increases movement speed which makes kiting possible in an increased number of scenarios. Stimming bad vs banshees and dts? Not if they're running by a turret and will make it by if you don't stim.
How is this calculated btw? Is it one marine vs one of these other units?
|
Sweden33719 Posts
Why the distinction between ht archon and dt archon? Your posts are awesome and make me wish I had a stronger mathematical grounding.
|
I just hold down t, works for me.
|
On September 28 2014 01:32 -HuShang- wrote:
Stim doesn't just let your units have more dps, it also increases movement speed which makes kiting possible in an increased number of scenarios. Stimming bad vs banshees and dts? Not if they're running by a turret and will make it by if you don't stim.
How is this calculated btw? Is it one marine vs one of these other units?
Of course, but the effect of movement speed increase (which is obviously important) is neglected here. This calculation only focuses on battles with big unit numbers (and anyway, you are not likely to encounter 10+ Banshees or DTs). It is calculated as army vs. army, so not just one unit vs. one unit (see the calculation in the spoiler tags, though).
On September 28 2014 01:30 TheBloodyDwarf wrote:
Terrans you heard it. DONT STIM
Oh, no, you revealed that my plan is to secretly manipulate Terrans into not ever stimming and defeat everyone on the ladder...
|
|
|
|
|
On September 28 2014 01:34 Liquid`Jinro wrote:
Why the distinction between ht archon and dt archon? Your posts are awesome and make me wish I had a stronger mathematical grounding.
There is no difference, obviously. The reason they are treated separately is that the program I used to draw the plots and make the calculations was originally made for calculating cost-efficiency, and there it did matter what the Archon exactly cost. Thanks for the compliment, by the way, I am really honored.
|
stim mostly used for running, splitting the army and killing banelings.
edit: maybe next time you should look about bunker. if marines or marauders or hellbats or whatever the best is vs protoss all-in.
|
1001 YEARS KESPAJAIL22272 Posts
you should show this to flash, he apparently thinks stimming twice makes his marines stronger
|
Thanks for taking the time to answer my question ! <3
So there are indeed situations where stimming is bad. I'm always wondering if I should re-stim a second time or not when my units aren't fully healed (I tend to overstim a lot). Good to know there are situations where I shouldn't ! (marine vs marine fights in TvT for example) !
|
We used to say among friends:
- What to do if a Terran opponent presses T?
+ Type GG and out.
|
These mathematics don't take into account player skill and kiting micro allowed the by the increased speed of your units. In 99% of real SC2 games you will usually want to stim your marine/marauder.
|
From a pure mathematical and theorical point of view this looks awesome. I admire your math knowledge
|
You really deserve that Tl+ by now, have fun with it :D
( I would have bought you a beer but thats harder to do)
|
On September 28 2014 01:39 hellokitty[hk] wrote:
I just hold down t, works for me.
I once won against a terran because while his death force was marching into my natural and all i had was a tiny number of muta ling bling, but his T key got stuck so all his marines had 5 hp, and he lost the whole army
|
On September 28 2014 03:02 Tufas wrote:
You really deserve that Tl+ by now, have fun with it :D
( I would have bought you a beer but thats harder to do)
Thanks, man!
|
Another interesting topic!
|
It also depends on army ratio: If you got 50 marines vs a probe, I am pretty sure stimming is not worth it (well unless you need to stim a few to chase down the probe).
|
On September 28 2014 01:50 lichter wrote:
you should show this to flash, he apparently thinks stimming twice makes his marines stronger
You never miss an opportunity to take cheap jabs do you? Cheers Lichter.
|
Not very practical data since a large part of stimming is the mobility boost, plus it's not taking into account potential kiting. But I guess if you like maths it's fun?
|
On September 28 2014 05:29 ROOTiaguz wrote:
Not very practical data since a large part of stimming is the mobility boost, plus it's not taking into account potential kiting. But I guess if you like maths it's fun?
But once an engagement is about to happen where you don't want to kite, you have the choice to activate Stim, or not, don't you. I guess you know it better, but I think it can be applied to some situations. And anyway, I like maths, so it is fun.
|
On September 28 2014 05:45 Sholip wrote:Show nested quote +On September 28 2014 05:29 ROOTiaguz wrote:
Not very practical data since a large part of stimming is the mobility boost, plus it's not taking into account potential kiting. But I guess if you like maths it's fun? But once an engagement is about to happen where you don't want to kite, you have the choice to activate Stim, or not, don't you. I guess you know it better, but I think it can be applied to some situations. And anyway, I like maths, so it is fun.
Yeah, the way to use this information is to look for places where it is useful, not look for places where it isn't
|
On November 04 2014 07:39 phantomfive wrote:Show nested quote +On September 28 2014 05:45 Sholip wrote:On September 28 2014 05:29 ROOTiaguz wrote:
Not very practical data since a large part of stimming is the mobility boost, plus it's not taking into account potential kiting. But I guess if you like maths it's fun? But once an engagement is about to happen where you don't want to kite, you have the choice to activate Stim, or not, don't you. I guess you know it better, but I think it can be applied to some situations. And anyway, I like maths, so it is fun. Yeah, the way to use this information is to look for places where it is useful, not look for places where it isn't
Agreed. Knowing this information will never heard anyone.
Although the second post in this thread is really funny
On September 28 2014 01:30 TheBloodyDwarf wrote:
Terrans you heard it. DONT STIM
|
Don't know if this has already been said or if it has been taken into account but I'm thinking that the big reason that stim is "always" better is the range of marines and marauders.
Many times against both toss and zerg bio can attack in a bigger number faster. Vs Z that is because Z is mostly melee and vs P that is mostly because bio is much more compact that most protoss armies. Archons and stalkers for example really fatens the army up.
Not thinking mano a mano so to speak but army vs army I think stiming is basically always better.
Imagine marines fighting stalkers for example, the ball of marines stims and runs close to the P army. Not all of the stalkers will reach to shoot and those that do will many times overkill the closes marines. There are many reasons for why stimming is better.
|
Could I may ask politely for an update for LotV?
|
|
|
On July 24 2016 07:33 propagare wrote:
Could I may ask politely for an update for LotV?
Keep holding that t.
|
|
|
On July 24 2016 07:33 propagare wrote:
Could I may ask politely for an update for LotV?
why would you want an update for LOTV? the calculations are purely theoretical and of no practical use.
The reason being the formula holding only for very large armies.
In the standard scenario encountered in a game you will find x<20 marines against 1 banshee or 2, not 10'000 marines against 100 banshees. Therefore, mechanisms come into play that are not considered in the calculation. E.g. stim might make the banshee get only 1 shot off instead of 2 which obviously makes a huge difference.
(the fact that increased movement speed, to close the distance to the banshee, is not considered has been pointed out by others already)
|
Old thread, but let me comment anyway....
This calculation ignores:
1) difference in range. Stim allows you to shot down lings before they can close distance, making the lost HP less important. Also the fact that marines can close distance faster vs longer ranged units.
2) Healing. marines have medivacs over them in almost all situations where you have stim. Which obviously makes stimming much more valuable, as medivacs get more targets to heal earlier.
3) micro advantage: splits vs banes, stutter step in to get all your marines to shoot, hunt down units, everything.
Seeing how these three are huuuge effects, it is nothing short of ridiculous how you present the number with 5 digits accuracy, as if that's how precise your model is.  I mean... you pick a model that essentially doesn't describe any battle in sc2, and then it doesn't matter how much maths you do, your results will still be nonsensical. The calculations are cute, but please don't claim that they have any applications in a real game.
It's a reason other people don't bother with this kind of calculations. Because they don't give you anything of value. To be able to apply them to a real game, you need to take all the relevant factors into account, which is incredibly messy.
|
there are definitely scenarios in which you don't want to stim especially during smaller skirmishes but when both main armies clash you should ALWAYS stim.
|
As a protoss, i'd recommend you to not research stimpack. It's not very good most of the time, just forget it (well I mean, you can use it against zergs, but just dont use it when you're against a toss, it would be considered bad manners).
|
|
|
|
|
|
EPT![[image loading]](http://i.imgur.com/t4sxq7X.jpg)
![[image loading]](http://i.imgur.com/0RAF73X.jpg)
![[image loading]](http://i.imgur.com/OYU0Apc.jpg)
![[image loading]](http://i.imgur.com/fdmkba2.jpg)
![[image loading]](http://i.imgur.com/VrfbuZx.jpg)
![[image loading]](http://i.imgur.com/bYnopHB.jpg)
