|
Hi,
I have been in diamond for 5 seasons in a row now with different races. I am trying to commit to terran and I will limit myself to 3 build orders.
Since I've never really bothered learning complete perfect execution, my build orders were pretty much always very sub-optimal which is probably holding me back by allot.
TvZ Bomber's 3 OC
TvP 3/1/1 allin
TvT Thorzain's fe
I've made notes on all of those builds and I have been messing around against the very easy ai on learning how to execute them. However I find it incredibly hard to perfectly remember and then execute for example Bomber's tvz build. It's very lengthy and it's of importance to have good execution.
I can only expect this to be a common obstacle, yet in my first year of rts I have never seen a thread about this. I was wondering how you guys first learn how to remember a complete build order and the rest of the gameplan and what you do to practice the execution part.
Sorry for any grammar mistakes.
|
I don't know how did you write down the build orders, but you should probably focus on timings rather than food supply beyond first few buildings and just remember the order of things. It's much easier to remember and imo a more proper way to learn. For example for a 1/1/1 you'd have something like (numbers from the top of my head):
17 factory, marine
19 bunker
21 gas
after 3 marines: reactor on rax + techlab on fact
starport asap
constant marines, cloak asap
cut SCVs around 30 supply
banshee (ready before 7 minutes) -> harass
2 more rax
start siege before 8:30, after first tank
move out at 9:00 with 2 tanks, 2 banshees, 15 marines (siege done at 9:20)
Or something along that lines. Then just practice in sp or yabot till you have it down and move to ladder or customs.
|
I would not recommend the bomber´s 3 OCs, tried it for a while and i found me in this terrible position, where the zerg has about 20 mutas and dictates the game.
I do not think that an allin like the 311 is a suitable build for learning tvp.
Take a macro opening with a 2-base-timing with stim and medivacs.
My openings atm are:
TvT: cloaked banshee into a tank/marine-push while taking my natural (various Koreans)
TvZ: fast 3 OCs into a quick 16-marine-drop (Kas)
TvP: 1rax-expand into a medivac/stim-timing (Thorzain)
The nice things about the TvZ and TvP openings are that you can switch things up quite easily. For example you can go for a ghostpush instead of rushing to medivacs or in TvZ you are free to do get out your tanks earlier, if you fear any pressure.
All in all i would go for econmic openings with some kind of pressure at around the 10-minute-mark. This is a nice compromise between early pressure and a very economic opening. I personally do not like the very passive styles like Bomber´s 3 OC build, because they give the enemy the freedom to do whatever he wants. I think it is better to pin him down to a few options and do not allow him to play with 100% of his skill.
|
Usually when I construct BO. I start with a mid game goal I want to accomplish. Then I will practice it vs an easy comp. (which does not attack so I can just focus on my BO) and practice the BO over and over again. Making small changes until I meet my goal, and know the BO
|
1. Play a game
2. Watch the replay, look for mistakes ( build order and missed opportunities as well as micro). Ask, "what if he does x."
Repeat
In addition to that, watch or review the day9 daily episodes where he covered these builds. Be attentive and look for stuff day9 may miss.
|
Just saw that i totally missed the point :D
Regarding your question:
Look up a replays until you find a bo, which suits your playstyle.
Search for YABOT, a custom map for training bos.
Play your bo over and over again.
Ask a friend to play some games against him/her, preferably vs different styles.
Train the bo in the ladder: Play just this one build over and over again.
At some point there should be a feeling that you know your bo blindly and how to react vs every style.
Then it is time to switch to the next bo.
glhf^^
|
I started using lenghty BO's since im in diamond and play it saver.
what i do to help me remember all of the steps is i make triggers so something like
(thorzian build) 2nd rax must come before or at 4 min
ebay scv builds factroy thenthe starport
so each time i play a game i learn a new trigger like teh 13 th scv builds the ref.
find out where u are making mistakes in ur bo or play and try to fix em with a trigger.
you can only learn triggers or define em if u play more and analyse ur replays.
|
I think that thorzain's tvt build is very good when your opponent FE...but it isn't that good against one base pushes. I think it is good to have at least 2 builds vs each race...one economical, and one all-inis/defensive
|
Link buildings together, Whenever I throw my fact down in tvz now I grab my second gas. That way I can get tanks out, combat shield and my reacters for my rax without being short on gas. Before I made that link I'd forget probably 50% of the time to drop the refinery at the right time.
Keep a notepad open on each build order and add notes from games, like when should I drop my cc on a 2rax push? What do I need to look for in terms of units to determine it's probably dt play etc. As you answer the questions add them to your notes on the build.
Most importantly only focus on one part of the build at once during gameplay. If you consistently miss scv's start there, if you get up to 700mins before dropping the extra 2 or 3 rax then don't let them happen and as things start to smooth out your build will be a lot better. If you try to simply get the entire thing 100% perfect then you it'll take forever. You might always be short 10 marines but not realize it's because you mess up when you get your second gas but because you're trying to get everything right at once other things tend to fall apart. Simply focus on one thing at a time every game and as you get that one thing right go from there.
Never ever deviate and just go "well I want to do 3rax this game so I will" you won't learn that way.
Terran is the most technical race, missing one part of the build can really throw the whole thing out of whack. Zerg and Protoss can skip a gas and just make more zealots/zerlings and keep a similar army strength. Terran can't the loss of production if you can't afford tanks is huge and your mineral/gas will quickly get out of whack and you'll look at the game and see 1000 minerals and think that could be 20 marines when in reality it should have been 3 tanks and a starport because of a gas shortage. Zerg and Protoss have their own difficulties but Terran have to be nearly perfectly on the technical execution of their builds, thats why korean terrans are so damn good.
|
Small things like knowing which SCV build what and when helps me a lot remembering openings.
|
On February 18 2012 09:58 AoWLuXus wrote:
Small things like knowing which SCV build what and when helps me a lot remembering openings.
I totally agree.
In addition to that you can figure out special benchmarks, e.g. building an engineering bay at 6:20.
But you can do this just in the beginning stage of the game, later your timings might be thrown of, because something like a drop happened.
Same is the case with bos, which are geared to food. You might lose your 3 marines in a zealot/stalker-poke and have to build your next structure "three food earlier".
So be careful with benchmarks in time and food.
|
Ive got a build called the 11 hatch that is really complicated, the way I remember it is 1 thing after another.
10/10 extr trick
11 hatch, 10 overlord, another extr trick, 14 pool, 15 gas, 18 overlord 18 queen, 20 2x lings, 25 ling speed pull off gas, 25 ovie, after first inject run drone to 3rd base (5:15 3rd base at 31 supply). put back on gas after 3rd started, evo at 30 gas, +1 melee first 100 gas, then lair with next 100 gas, start 2nd and 3rd gas as lair starts for +2 melee, baneling nest, baneling speed. at 40/44 do 2x queens, they line up with your first queen who's 2nd energy is creep tumor, then ran to 3rd base.
^ just from memory, havnt done that build in months
Basically using a lot of visual queues like queen almost done, gas hitting certain point, certain building complete, game timer. These will help you remember your build. That and practice it a lot till you remember it easily.
|
Thanks for the replys.
I pretty much write notes down with cues and general ideas that I'm already familiair with. For instance I can already flawlessly execute the start of Thorzain's FE build aswell as Reactor hellion FE so I can skip that part in my notes. Usually I don't use food past like 30 supply and then I just go for a very general order in which I build stuff.
It will be neccesary to take the time to learn builds properly and execute them more then 5 times vs ai to do them well enough for ladder I guess.
On February 18 2012 09:11 maracuja wrote:
I would not recommend the bomber´s 3 OCs, tried it for a while and i found me in this terrible position, where the zerg has about 20 mutas and dictates the game.
I do not think that an allin like the 311 is a suitable build for learning tvp.
Take a macro opening with a 2-base-timing with stim and medivacs.
My openings atm are:
TvT: cloaked banshee into a tank/marine-push while taking my natural (various Koreans)
TvZ: fast 3 OCs into a quick 16-marine-drop (Kas)
TvP: 1rax-expand into a medivac/stim-timing (Thorzain)
The nice things about the TvZ and TvP openings are that you can switch things up quite easily. For example you can go for a ghostpush instead of rushing to medivacs or in TvZ you are free to do get out your tanks earlier, if you fear any pressure.
All in all i would go for econmic openings with some kind of pressure at around the 10-minute-mark. This is a nice compromise between early pressure and a very economic opening. I personally do not like the very passive styles like Bomber´s 3 OC build, because they give the enemy the freedom to do whatever he wants. I think it is better to pin him down to a few options and do not allow him to play with 100% of his skill.
Cause it's slightly off-topic here, I sent you a PM about those builds  .
|
When I practice new builds I do first look at food counts, write them down on a piece of paper and "memorize them"
Once I get used to them the build, I rewrite them with cues instead of food counts.
Once I get used to that, I then base what I build based on what I scout.
For example.
When opening Reactor Hellion FE, I first learned to opened standard--which was hard.
10 Supply depot
12 barracks
13 gas
15 orbital
15 marine
16 supply depot
This was a very hard build for me at first. For example, I wouldn't grab an scv to build the first depot until I was at 10 supply. This delayed my depot by around 2-5 seconds. Which meant I was supply blocked for 2-5 seconds longer than I should have been. Which meant that my mineral line was mining for 2-5 seconds longer than it should have by the time I was building the Barracks. Which meant that the refinery was being started a few seconds too early, which meant I had pulled 2 guys off the mineral line into my refinery too soon--which meant my mineral timings were off and so my factory kept getting built at 20 supply instead of 18.
Then I got used to the build. It turns out, there's a lot more to it than simply the food count.
at 7 supply I begin mineral stacking.
At 9 supply I rally my command center to my ramp while putting a 10th scv on the cue.
when the SCV pops out I click on it, set it to a hotkey, then re-rally the command center to a mineral patch. The SCV will arrive at the ramp with 100 minerals exactly. I then shift command the constructing scv to scout once the depot is finished.
At 11 supply I rally my command center to the ramp--he will build a barracks. It's the same process, hot key him as he leaves, re-rally back to the mineral, double tap the hotkey and build a barracks.
Once the barracks starts, at 65 mineraIs grab an scv beside the command center and send it beside a refinery--build. this will be at 13 food. When you build the rax at 12, you cue your 13th scv the moment you get 50 minerals. But by the time you have 65 minerals, your 13th scv will almost be finished.
When your barracks is almost finished you'll be at 15 supply. At which point your refinery is almost finished (send in two scvs to it at this point) and then select your command center and hold B for orbital command.
When the orbital has started, make a marine at 50 minerals then make a depot at 100 minerals.
This seems complicated--even moreso than food counts. So what does this actually feel like now that I can do it automatically?
Mindset--always build SCVs.
Game opens, make an SCV, send the 6 to mine, start mineral stacking.
By the time I mineral stack my 2nd mineral patch--I rally to the ramp. Go back to mineral stacking.
I hear the SCV sound, click on him and hotkey him, continue mineral stacking.
When I finish stacking my 3rd patch, I double tap 3 (the control group I set that SCV to) and build a depot; then I shift click it to scout.
I hit back space and I work on my 4th patch. When I finish that I rally my command center to the ramp. Same deal, I hotkey, then I jump to the ramp with a control group to start the barracks.
Once the barracks has started I hop to my scouting scv for a second to so to see what he's looking at, then I hop back to the command center to start a refinery (this will be at around 13 supply but there's no need to check--assuming perfect timings)
once the refinery starts hop back to your scouting scv. While your scv is scouting keep tabs on the Barracks. When it is 75~% done, hop back to your base--the refinery will be 80%-90% finished and you can send 2 guys to the gas.
Hop back to your scout for a second or so. When the barracks completes I start my orbital then start a marine.
I don't use the food count to tell me when to build things--but I check on it to let me know if my timings are off or not. Since everything is cued off of what I'm doing--I simply check the supply count to see if I messed up at some point.
Barracks is starting, I cue an scv, check the supply and its at 12... crap I cut an scv/built too early.
The better I get at a build order the more I do stuff based on what is happening and less on food count. But you need the food count to know how well you're doing the build. You don't need it to do the build.
EDIT:
To directly answer the OP's question/request.
I base what I build off of what I am doing recursively. For example, when going for a Reactor Hellion FE opening--I neither build the factory at 100 gas nor do I build the factory at 150 gas (to allow the factory and the reactor to finish at the same time). Instead I build the factory after my 2nd marine and I'm halfway through my 3rd marine.
It goes like this.
I start my third marine.
I hop back to my scouting SCV
I hop back after i see what my SCV sees and build a factory.
Now it may not be a scouting SCV (he could be dead), but I do things such as send one of my marines to scout the perimeter of my base, or run down the ramp a second to check if a ling is watching the bottom of the ramp, etc... It's easier to sequence actions than it is to check food/gas count.
I do this for all my builds. I memorize the action sequences instead of the food counts. But I memorize the food counts to make sure my action sequences are crisp.
|
On February 18 2012 08:46 habermas wrote:
I don't know how did you write down the build orders, but you should probably focus on timings rather than food supply beyond first few buildings and just remember the order of things.
If you want to really refine your build, you should have both the food-timings and the in-game timings line up every game.
|
Is it the build thats giving you the problem or is it executing a build in a game with all that extra info?
The key to that though is fast iterations ... find a map that you can just restart via message and maybe double your effective time. Also time spend practicing an abstraction is precisely that, time not spent playing. I think that playing games is the better approach as you are dealing with real data, just don't have an expectation of winning. Your objective is to practice your build - blame blizzard for the dumb ladders design making that cost you - and analyse your results. My point is balance.
its not been said but day9 does have some good vods on it. But basically he says learn it preferring visual / audio cues to food/time counts. But as above use food and time counts as measures of crispness. That is just repetition, observation and tweaking to hone timings. The other side is analysis but it all depends on being able to execute stuff cleanly.
I find using a program like Freemind (a mind mapping tool) to be really useful as you can easily build branched diagrams and add all kinds of info to a node. Can branch builds with reasons ... mix it with observations that are race and build specific etc. TBH its the analysis that goes into it that counts for me at least ... the map itself just serves as a reminder when i look back a few months later and compare what im doing now to then ... seeing what new stuff im doing and what stuff ive forgotten or rejected.
I should have a lot more to say, but don't do it enough
But i do think its a better tool for the problem than notepad.
|
What makes builds a lot easier for me is by chaining tasks together.
For example I might have an SCV build a rax, then a CC then a refinery. In an ideal world I might be able to get that CC 4s quicker, but by having 1 SCV do all 3 jobs I am much less likely to forget the timings and it's just less effort. Plus you can queue these actions in advance.
Another tool I use is when SCV are produced, for example in one build I know that my first scv after orbital goes on gas, the next makes a CC, then the next 2 build factories, then the 5th builds a rax. Again it means my timings aren't 100% perfect, but it does mean that I never ever forget a vital building.
|
|
|
|
|
|
EPT
