|
EDIT2 : http://www.teamliquid.net/forum/viewmessage.php?topic_id=83287
+ Show Spoiler + The searchword was WORKER SATURATION. I am a search function noob indeed
EDIT: What we got so far: 2.5 , as an 'ideal' multiplier X. Is that 100% saturation or the one close to it?
Also we have the number 85 for planned 4base play.
What is the ideal number of probes per mineral line?
What I used to do now is constant probe production, because of future expansions and harrash, and I stop just by the feeling of it. But this leads to too many probes and problems with PvT especially, when T decides to go for timing with scv cut push and I have too much money in probes.
Is there an exact formula saying X * Y =Max expo potential ?
X=number of mineral patches
Y=number of probes
I know that the law diminishing marginal utility applies, and any probes above X*1 will bring less extra minerals, but did anyone try to construct the curve for marginal increase of probes from x*1 until max potential?
+ Show Spoiler +Thank you for help or forwarding me to a thread that deals with this.
Probe counts are the only reason that holds me from staying at C+ level on Icc
|
It's generally about 2.5 per mineral patch for optimal mining. There's probably a lot more keeping you out of C+ than just probe count.
|
On January 28 2009 21:33 Tropics wrote:
It's generally about 2.5 per mineral patch for optimal mining. There's probably a lot more keeping you out of C+ than just probe count.
I can reach C+ comfortably with aggresive midgame strategy.
My multitasking is better than vast majority of C players. But when I try to play all out full macro game when I go to 200/200 I will be stuck at C-, because of my poor mining distribution.
I always end up with too many probes for some reason, its my bad bad habit.
My strategy for ages was based on aggresive early game, gaining destructive advantage and mass expo against any race. But people are getting better at defense and thats why I am trying to build my mid/late game which is HORRIBLE
|
Isn't the saying "If you don't know when to stop producing probes, never stop"?
|
On January 28 2009 21:39 SteveNick wrote:
Isn't the saying "If you don't know when to stop producing probes, never stop"?
That goes to early game/mid game. And is the reason why I can win easily in this period because I almost never stop hitting 9/0 and make probes. But beyond certain point the extra probes bring much less extra minerals, until they don't bring any.
Thats why I need to know how the curve works and to know the saturation levels in order to manage my mineral patches properly TT
|
Valhalla18444 Posts
if you think you have too many probes, it's a clear sign that you are not expanding as relative to your economy as you should be
this means you aren't effectively battling for map control, or have a sub-par knowledge of when/where you can expand. this trickles down into other things like gateway timing and messes up your macro hardcore (though not in a way you might notice if you are capping at 200/200 at varying times and sitting on the army)
|
On January 28 2009 21:46 FakeSteve[TPR] wrote:
if you think you have too many probes, it's a clear sign that you are not expanding relative to your economy
this means you aren't effectively battling for map control, or have a sub-par knowledge of when/where you can expand. this trickles down into other things like gateway timing and messes up your macro hardcore (though not in a way you might notice if you are capping at 200/200 at varying times and sitting on the army)
Correct. I have problems beyond my third base (97% of my games were finished off of my 3bases before) and that is what I am working on. But I am completely oblivious to the probe count, and thats why I have problems beyond 3bases (when I simply mass probes until my 3rd is running and then manage my gates according to the inflow)
I need to know the curve to know when to stop the nonstop probe stream, to be able to plan my probe count with my planned expansions
|
Valhalla18444 Posts
On January 28 2009 21:53 LemOn wrote:Show nested quote +On January 28 2009 21:46 FakeSteve[TPR] wrote:
if you think you have too many probes, it's a clear sign that you are not expanding relative to your economy
this means you aren't effectively battling for map control, or have a sub-par knowledge of when/where you can expand. this trickles down into other things like gateway timing and messes up your macro hardcore (though not in a way you might notice if you are capping at 200/200 at varying times and sitting on the army) Correct. I have problems beyond my third base (97% of my games were finished off of my 3bases before) and that is what I am working on. But I am completely oblivious to the probe count, and thats why I have problems beyond 3bases (when I simply mass probes until my 3rd is running and then manage my gates according to the inflow) I need to know the curve to know when to stop the nonstop probe stream, to be able to plan my probe count with my planned expansions
i mean that probe count isn't what you should be worrying about. with a stronger grasp of expansion timing, your excess probe problem will solve itself.
|
I believe you stop probes at around 85. This number is enough for 3 running expansions.
And from what I see in games if you have the opportunity to make safely more than 4 Nexus that can mine at the same time, the game is pretty much over and your opponent is dead.
|
Valhalla18444 Posts
On January 28 2009 22:04 hymn wrote:
I believe you stop probes at around 85. This number is enough for 3 running expansions.
And from what I see in games if you have the opportunity to make safely more than 4 Nexus that can mine at the same time, the game is pretty much over and your opponent is dead.
nahh, i get into 4 base terran vs 5-6 base protoss games all the time
there is definitely no threshold that says 'if protoss has x number of bases he will win', there are only advantages and disadvantages
|
Hot do you get to the number 85?
No really my question stands and only thing I got so far is 2.5x mineral Patch is the Ideal (Is it maximum? ) amount. There are many situations when you know you have the right amount of bases, and adjusting probes accordingly would be nice.
I would be really surprised if those theorycrafters out there didn't analyse this before
|
On January 28 2009 22:13 FakeSteve[TPR] wrote:
nahh, i get into 4 base terran vs 5-6 base protoss games all the time
You mean you play games where you have 5-6 Nexues mining and saturated at the same time or you mean games where you have 3-4 Nexus mining and 2 that got no minerals?
And I got to 85 probes cause I remember I read a translated article or an interview with a progamer and he said that he gets like 85 probes cause they saturate 3 mineral lines (yes, the formula is 2,5 probes per mineral patch) and by the time he gets another expansion he needs to transfer probes to it from an expansion that is left with almost no minerals (like the main) .
I got the feeling he times his fourth Nexus to be done around the time when his main is going very low on minerals. So when he transfers from the main to the 4th Nexus he still has 3 mining bases even though he has 4 Nexus.
And also when I watch reps say PvT that go to very late game where both are 200/200 I sometimes pause and check the army of the Protoss player. He usually has 4 groups x 12 units (48 psi), 6-7 observers (6-7 to 10 psi) around the map (sometimes even 10) and a shuttle full of HTs (10 psi) and say 2-3 Arbiters (8-12 psi). So that makes for a army of 96 + 7 + 10 + 8 (or 12) = 122 (or 126). This leaves a bit less than 85 psi for probes (leaves 78 psi or 74 in the latter case) but still it's close enough.
Just look some replays and check army : probe psi and you will see that 80 to 85 probes is the number progames aim at.
Yes, they can time a fourth Nexus to be done when the main is depleting (and we can't), yes they will transfer probes at the right time and they won't be idle for too too long (and we can't) and so on. But I think that's the number that progamers make and we should also make.
BTW making like 100 probes wouldn't be that big of a deal at the level of the OP (mine is currently C-) cause 15 more than 85 would be like 3 goons and 4 zeals less (bad if he's missing the Arbiters and the Shuttle with the Hts). Cause I think 3 goons 4 zeals won't make a difference in a 200/200 battle (and 3 Arbiters and the shuttle will make the huge difference as far as the Protoss is concerned).
|
There are quite a few threads on this topic, and unsurprisingly a lot of bad info as well. But I suppose people don't know the history and so some things are bound to come up over and over.
see: http://www.yellowchrome.org/1com/sclegacy/final_review.pdf
As you can see it depends on the why the mins are laid out in the map. But the rule for most popular maps is closer to 3 per patch than 2.5.
Of course, playing the game well is a more complicated matter than producing probes consistently until optimal. For example in pvp keeping up in unit production is very important, and it is often too risky to make probes all the time. So the trend in pvp is more along the lines of making enough for 1base, then expo. Then make more or cut based on situation.
I also suggest watching replays to see how these dynamics play out so you get a better idea of how everything relates. One thing you may notice is that while very good players tend to follow this, they hardly follow it exactly, as you might expect of very good players. My guess would be that these players have some internal understanding of when to build and how many they have, but its hard to be exact without planning it all out beforehand.
Err come back later and say if this improved your game any.
|
United States43989 Posts
Watching a group of reps with Carlo I noticed he stopped probe production at 65-70 every game (that adv loader plugin was awesome). Was kind of surprising that he did it with such precision, even when he'd lost probes before that should mess up his timing. That was mainly PvZ though.
|
if you cant stop pressing 0p9p, change what u have hotkeyed to 0,9 when you feel you have enough
|
Didn't someone make this UBER GOSU probe/scv/drone saturation thread for his 10k posts?
|
On January 29 2009 04:32 Ziph wrote:
Didn't someone make this UBER GOSU probe/scv/drone saturation thread for his 10k posts?
Yea, I'm looking for it, but I can't find it. I'm actually looking for this again, if someone could find it.
I think his conclusion was that the rate at which you stop gaining a linear benefit from each probe doesn't level off until way off into the future, but I can't find it so I can't look it up.
Edit: Ugh, I'm never going to get over the the fact that I'm a DT.
|
|
|
I guess this is your ICC account:
http://www.iccup.com/starcraft/gamingprofile/doc.lemon.html
@Topic: 2.5 Probes per patch are alright until you hit 4th expo, at that time I like to stop probe production, if there wasn't any major loss like reaver drop, successful mutaharrass, lurkdrop, vultharrass and so on. It also depends on the map. On maps with low expocount, Blue Storm for example, you might want to stop probe production after your 3rd expo.
|
That and
http://www.iccup.com/gamingprofile/brutal[pain].html
Just to show you how massive difference is when I play safe and my normal/aggresive style.
There is no clear answer to the probe saturation, and the curve is fairly linear, but the ration of 2.5 with constant adding of expansions sounds about right for protoss.
3.0 with low base play.
I was just curious about this, and when I knew I had 4 bases and my opponent 3, and I knew there was no safe expansion point I just didn't know how to maximise my mineral inflow in order to outmacro him ( before I believed that the more probes the better)
But obviously for early/mid game expert like me the gateway count, expansions timings and large army control becomes a big problem after 3 bases, but forcing myself to play like this should improve my game overall. And I finally understand what are people talking about in the strategy forum by not cheesing/ doing something unorthodox every game.
Better players beat me now almost every game, but on the other hand i find it very very hard to lose to a C-/C max player. Before I could win or lose virtually against anyone at least remotely near my skill.
|
|
|
|
|
|
EPT
