This week we go back to basics and focus on the warp prism. Since the increased pick up range and slow warpins from pylons in lotv the prism has become an incredibly important fundamental ingredient in Protoss play. I have previously written about other control groups but the prism control group is one of the most important ones. This weeks build will allow you to improve your prism control using several different tactics. The idea is to open with 1 immortal and 1 prism for safety and pressure at the same time similarly to 1 oracle, and to use hallucinated phoenixes instead of a real phoenix to scout the back and entirety of the zerg base. In other words this is very similar to a sg opener in terms of safety, pressure and scouting. In order to stay extra safe and to force ourselves to use the prism to its fullest potential we also open with a full wall at the natural, and use the prism to elevate a probe out to take a fast 3rd and transition to the macro style of your choice. The most unconventional part this week is the full wall at the natural, something which can be more of an inconvenience that something out of a playful dream. In order to better learn the prism micro and of course in order to enjoy this week we want to be thankful for the stuff the prisms achieve, so instead of thinking “gosh it’s annoying to get an immortal to my main army with a full wall” think “yay look how I just microed my immortal out of a trap”. (That said I’m not trying to convince you that full walling is great fun, I admit this week was not ideal in terms of unconventional fun play and will try to do something more crazy next week :p). In my opinion this build is still great fun to play though, it allows scouting with free hallucinations instead of fragile oracles / archon drops / phoenixes and you can get a really fast robo bay if you want to. The lack of creep also allows for some interesting tactics although I leave those for you to experience yourself.

Build overview

Phase 1:

Open with a 1 gate expand. When the expand finishes take 1 probe out of each gas in the main. The idea is to then get 1 adept and warpgate followed by a robotics facility when it can be afforded. After the adept you start a sentry and fullwall the natural before lings can get in. The sentry will give you your first good scout of the opponent with an hallucination at 75 energy. The robo builds 1 immortal followed by 1 prism. Get a forge and +1 attack when you can afford it, and gases at the natural a bit before getting oversaturated.·

Short phase 2: - when prism finishes

As soon as the prism finishes start an observer and rally it to the prism. Here things get interesting, we use the prism to let 2 stalkers kill highground overlords while the prism elevates out the immortal and probe that clears the 3rd base location and expands. The prism then goes into warp-in-mode to place a shield battery followed by a pylon at the 3rd creating a nice wall-in for your sentries to defend the 3rd. All of this should take at most 10 seconds, try to do it as fast and accurately as you can, this is a lot of useful techniques already.

Phase 3:

The prism then picks up the 2 stalkers and immortal and goes across to kill any overlords or creep or anything else it can find while scouting the opponents front. You also scout here with an hallucination that determines if you have to cancel the 3rd base and recall your units back or if you can defend with more warpins or if you can keep the base and continue adding infrastructure and tech of your choice. If you want to go all out with this build then imagine that the prism will never be sent home, it will clear creep the entire game. This will give you an opportunity to do regular prism drop micro continuously and the more you do it the more you are rewarded. This also lets you practice jumping to and from the prism quickly. If instead the opponent goes for an all in the prism serves a very different purpose. Then it stays at home and decides which units are on which side of your full wall using elevating, and it is also used for dodging the opponents attacks or saving weak units like it is normally used in micro situations. Against zerg attacks with no anti air it can become quite interesting also. Against banelings you can try to drop a stalker onto them and pick up clumped up units last minute to dodge attacks. Against a zergling flood you can re-possition all your units at the 3rd to make sure that all units are in chokes between pylons or mineral patches as well as save weak units. Against ravagers you can last minute save units that are surrounded by lings and corrosive biles and sometimes make the ravager shots deal damage on the lings instead. Against pure roach you can kite better with stalkers since they can be saved by the prism after walking themselves into a corner and drop them somewhere else and start the kiting all over. And against all of the above except 3+ ravagers you can re-power shield batteries that lost pylon power using the prism as well as warp in units at the 3rd even after all pylons are down. A lot of these all in situations come down to the execution of the protoss. As always watch replays if you struggle against all ins and see how to micro better in theory from a paused situation, then try to implement it next time. But also remember that sometimes the 3rd should be sacrificed; losing the 3rd doesn’t necessarily mean that you’re behind.

Phase 4 - Followups

Here I leave it to you to do anything you want. The quick robo allows for a fast robo bay, and the fast +1 and +2 gives you more options with army compositions. Just prepare in some way for mutalisks since you don’t have a stargate which will prevent you from quickly adding a fleet beacon and starting phoenix production. The new cheaper blink can be used.

Handling the full wall at natural

You might think this is a huge problem but it also has some advantages, and the disadvantages disappear if you have perfect prism micro. Ling runbys that are supported by ravagers or banelings can be an issue in the mid game stage if you don’t have a full wall, and roach runbys can also be an issue often combined with a lurker attack or swarm hosts. The full wall makes it easier in this situation. The more bad scenarios are nyduses/drops/mutalisks,. Against these you can simply kill a structure in the wall, it already served its purpose, or use prism micro and a prepared composition/static defence.

Transfering probes can be done with a prism or with recall. I recommend recalling probes to the 3rd after scouting with the hallucination and then using a prism for any later transfering. The recall with also reduce the probe travel time! :D

The robo might cause some units to not rally to the 3rd because of the full wall, rally them to the wall and use a prism to elevate them out. This is actually not more micro intensive than getting it out of a hold position zealot if you are used to it.

It is recommended to get a 2nd prism as soon as possible to deal with this as soon as possible since the first prism is supposed to never come home again. If you want to play it safe get 1 immo then 1 prism after the obs finishes.

Recommended control groups for prisms

As mentioned briefly the prism should definitely have its own control group. However this can be insufficient, a lot of strategies are more powerful when you combine a main attack with side/back-attack. Both of these attacks are much stronger if supported by a prism, but 2 control groups is a lot. This could be done by simply not control grouping the counter attack prism, but that makes it harder to save. My suggestion however is to have a “single prism/prism with army” group and to gain comfort with a “counter attack prism” group. This second group can also be counter attack units of any sort, but ideally have this group be separate from the counter attack group introduced in week 2. To recap with all the previous weeks control groups we have achieved the following setup (in no particular order):

  • main army (taken for granted)

  • prism (week 9)

  • counter prism (week 9) (could be omitted and share purpose with the counter group, but will reduce mobility)

  • counter group (week 2)

  • trash group (week 2)

This is starting to be a lot and I recommend you choose a complete system and go with it, this is a good opportunity to do so. My recommendation is the following:

  • nexi (and upgrades if you want)

  • robo/sg (don’t need to control group gateways anymore now that they auto-transform to warpgates)

Only 3 more to go. If you’re like me and like mixing in a couple of disruptors while playing with multiple other units then that is extremely difficult if you do not always use the same group for it.

  • disruptor (I also put these in the main army group)

Similarly oracle micro in the mid/lategame becomes easier if that is consistent

  • oracle

And now we only have 1 group left. I suggest a “secondary army” group which usually consists of units in the main army that are microed separately, for example ht or stalkers or immortals vs lurkers or tempest or phoenix. This group can also be used when you have to split your army to defend 2 locations, in this situations it is too difficult to have the main army consist of 2 control groups anyway so there is no overlap in necessity of these 2 different purposes.

  • “secondary army / split army in half to defend 2 places”

Take the time to invest in loss and find some good keys that you don’t use a lot to use as control group, for example tab, q, (key left hand side of 1), caps lock. Also choose the purpose for each group smartly, aka make sure spells of the oracle are easily pressed after pressing the oracle group key etc.

Why I think 10 control groups are helpful and relatively easy to learn

The massive advantage of 10 groups is you're allowed to use a setup which has each units in the same group every time. That means that with the cue "I want to select oracles" you're always pressing the same key, in other words you don't have to think about in which group the units are before pressing it. The first advantage with this is that it speeds up the process of selecting control groups (after you have gotten used to it, expect a 1-3 months to get used to that). The second advantage is that because you're using habits instead of short term memory to keep track of which group things are in there is no confusion with using many different units in a single army which would otherwise create a massive brain lag. This is great for late game armies.

The most important thing when trying to learn to use 10 groups is consistency, you want to first go in unit tester and make sure that you can do it all mechanically without misscliks, both selecting units and control grouping a set of units correctly. After that force yourself to always use your system in ladder. If you always put oracles in the same group for example the brain will quite quickly catch on and get faster and it will be less and less brain intensive to do. Expect this to take several weeks though, or even months. I actually think that learning to use 10 very strict and consistent groups is easier than learning to be flexible with much less groups, it just takes longer to get started and it will temporarily make you very bad. If you want to do this 1 group at a time or all 10 at once is up to you.

Of course you might settle with a worse control group system since the above is quite a lot of work, in that case I still recomend using clear-purpose groups rather than trying to rely on flexibility, and perhaps ommiting groups that you find not necessary, for example the 2nd prism group and the disruptor group and putting disruptors on the "secondary army" group instead. Of couse this will make using tempest + storm + disruptor + force fields impossible, but maybe you are okay with that.


Replay - FBW#9

Build Order

Get SALT Encoding   

  12  0:00  Probe  
  13  0:12  Probe  
  14  0:18  Pylon  
  14  0:24  Probe  
  15  0:29  Assimilator  
  15  0:33  Probe  
  15  0:41  Probe  
  16  0:48  Gateway  
  17  0:51  Probe  
  18  1:03  Probe  
  19  1:15  Probe  
  20  1:26  Nexus  
  20  1:35  Cybernetics Core  
  20  1:38  Probe  
  21  1:44  Assimilator  
  21  1:50  Probe  
  21  1:53  Pylon  
  22  2:12  Warp Gate, Adept  
  26  2:28  Robotics Facility  only mine gas with 2 probes in main gases after you reached 100 gas for this
  28  2:40  Sentry  
  32  2:46  Gateway  
  34  3:01  Gateway  these make a full wall at natural, on some maps you only need 1 at the front. If scouting lings come you might need to cancel probes to get this wall in time. After this you can chrono more workers.
  36  3:09  Immortal  
  36  3:15  Phoenix  hallucinated at 75 energy
  40  3:25  Pylon  
  41  3:29  Assimilator  
  42  3:38  Warp Prism (Chrono Boost)  
  42  3:39  Forge  
  44  3:55  Pylon  
  48  3:57  Assimilator  
  48  4:00  Sentry x2  
  53  4:07  Pylon  
  54  4:14  Observer  rally to prism, make this right after the prism. After the 3rd base the robo can make immo-prism.
  56  4:24  Stalker  
  61  4:33  Nexus  
  61  4:36  Shield Battery  at 3rd using prism
  61  4:40  Sentry  
  61  4:41  Stalker  enter phase 3, load up 2 stalker 1 immo, the prism is ideally never coming back.
Spawning Tool Build Advisor. Get it on Overwolf


Learning how to do it

This one is a simple one, just do it over and over and do your best. If you don’t know how to control the prism do some replay analysis and watch vod where I give some suggestions. There are also some suggestions in the phase 3 description. Force yourself to control the prism nonstop throughout the game and remake observers if you lose it. Recent patch should make observers a lot more durable though.

  • As always start by executing the build against ai, phase 2 is tricky so play from replay against ai to allow yourself to perform this 10 second phase 20 times in less than 5 minutes time. You can also practice clearing creep or poking with the prism in a similar way.

  • You will probably lose a lot of games to different all ins due to missed scouting opportunities or keeping the 3rd when it should be canceled or not prism microing to its fullest potential. Don’t be too bothered, keep trying. The trick to learning these micro tricks that usually are not needed is to put yourself in a losing position against all ins where you are forced to micro your heart out, so even if the scouting is off and even if you’re not even close to holding it can be a good learning opportunity as long as you keep fighting to the best of your ability to the very last army unit. If your scouting and macro is spot on however this might not happen, then try to be as greedy as possible with your defence to make this happen anyway.

Build simplifications

  • Start by doing the same opener but keep the prism at home instead of harassing with it, this will train all the defensive prism moves and make macro and defence easier. Get extra sentries instead for scouting. Likewise 50% of the time instead don’t make a full wall and use a prism only aggressive. This will allow you to only make 1 prism each game, which can sometimes be a good start.

  • One easy build would be to stay on 2 bases and do the same build otherwise. After the prism keep building immortals and prepare for a 2 base sentry immortal all in. This is quite weak but can be a good way to make macro easy and defence easy to allow you to practice the offensive prism usage.

  • Skip +1 to make defence easier initially and always follow-up with more immortals to stay safer.

Additional difficult techniques/improvements

  • Get 2 prisms that constantly deny creep instead of 1. (simply follow up the observer with 1 more prism and 1 more observer). This makes it much easier to achieve that purpose since zerg can no longer just put their queens where the 1 prism is and chase it off as well as instantly re-spread any lost creep tumors. But of course 2 prisms are harder to control.

  • Make a macro follow up that allows for researching prism speed to allow the prism to be useful after hydralisks and mutalisk enter the field.

Go to:


  • Created by: FightingFrog 
  • Published on: Jan 26, 2019
  • Patch: 4.7.1
View Real Time Build Order
Export as new Build Order

Comments (0)

There are no comments. Be the first!

Create an account to comment!