[G] 2 gateway 1 stargate PvP Phoenix build

I am a mid-level Master's league player, and my ambition has been to break into the top 200, and while I've come close, I've never managed it. As such I am not an amazing player, I feel I have a very good understanding of the game, but lack the coordination to pull it off. Because of this I have not been able to test this build at the absolute highest levels of gameplay, but I suspect the strategy will be more viable as the skill of the player increases, as it requires a high degree of multitasking

Furthermore, this build relies on forcefields to block off any sight of your base's high ground, and while the build order provides the sentries, this is still a difficult task. As such, I still lose approximately 40% of my PvP games outright due to a single misplaced forcefield or slow reaction. It is also near impossible to pull off on Scrap Station due to the wide ramp size, and maps with back entraces (Jungle Basin, Delta Quadrant) make it slightly more complex (walling off the back entrance with buildings is recommended)

- Hold off any 4gate push without cutting probes.

-Secure an economic advantage against 4 gate builds through constant probe production and phoenix harass

- Force the opponents army to cluster up by picking off lone units with phoenixes, and pin him in his own base by harassing the probe line whilst securing further economic advantages.

-Phoenix harass keeps the opponents army in his base early and allows easy scouting of colossus timing

-Phoenixes are faster to tech to, allowing a push before a colossus is built

-Robo builds often depend on immortals early to deflect aggression which are exceedingly vulnerable to phoenixes, similarly sentries are almost entirely nullified by phoenixes as they can be picked off before a battle starts with relative ease.

-Succesful colossus production still allows for easy void ray production. Since colossi reduce the number of stalkers in an army due to gas requirements, it is a very common situation to have enough phoenixes out to lift up every single enemy stalker, allowing voidrays to quickly decimate the colossi (Note colossi are unable to escape from void rays, as both move at the same speed, 2.25, and can path over cliffs.)

9 pylon
10 Chronoboost 2 x probes
13 Gateway
14 Assimilator*(see note)
15 chronoboost 2 x probes
16 pylon
Keep up constant probe production without further chrono boosts from this point
@100%gateway, Cybernetics Core
@75 minerals immediate second gas (~18 supply)
@100 minerals Zealot + chronoboost it out**(see note)
@100% cyber core, Warp gate research + chronoboost it out
@100% Zealot, Sentry + chronoboost it out

-If scouting revealed a 4 gate strategy (Indicators: single gas after ~20 probes, very high energy count on the nexus, 10 gateway [compare his gateway timing to yours to determine this])

Chronoboost out a second sentry and get a second gateway as soon as possible. Order 2 more sentries from your 2 gateways and chronoboost where possible, your 4th sentry should finish shortly after warp gate research. Get a stargate as soon as you have 150 gas after ordering your 4th sentry.

-If scouting revealed a build other than 4 gate

Sentries can be cut and the second gateway delayed for a faster stargate.

*Note: Delay your second nexus Chronoboost, this is done to allow a slightly earlier assimilator as the build is gas heavy.

**Note: This build is timed out such that if it is done perfectly your zealot finishes 3-4 seconds after your cybernetics core, and literally AS you get 100 gas for your sentry. Chronoboosting this zealot allows the earlier sentry, whilst still providing a zealot.
Since you will not get out a stalker, and since sentries are slower than workers, you should try to micro the zealot to pick off the enemy's scout. This is unlikely but can happen.

-Even against fast warp gate rushes, such as a 10gate 1 gas perma-chronoboosted research 4gate build, you have 2 (or at least very nearly) sentries out when the push occurs, A chronoboosted sentry takes approximately 30 seconds to build, and you have 2 forcefields available to use, each lasting 30 seconds, hence the timings are tight, but with practice it can be accomplished.

-4 sentries are enough to permanently forcefield a ramp, FF costs 50 energy and lasts 15 seconds, hence 3 sentries are enough to FF a ramp for a very long time if the sentries start with lots of energy (or if you can buy 5 seconds inbetween forcefields). 4 sentries allows for completely permanent FFs.

-Correct forcefield placement (at the very bottom of the ramp) denies sight ontop of a ramp, preventing units from being warped in on your high ground at all.

-Any tech to flying units (such as robo for observers / hallucination) to allow sight up the ramp delays the push significantly, and as only 4 units can be warped in at a time in your base, and your ramp will still be perma forcefielded, your army will be enough to crush the 4 units warped in quickly.

-Teching to blink stalkers is an extreme delay, and correct forcefield placement will still deny the majority of blink stalker plays.

-Phoenixes are light units, and as such stalkers do not get a damage bonus against them. A phoenix costs 25 minerals and 50 gas more than a stalker, and has 120hp and 60 shields. Stalkers have 80 hp and 80 shields

Maths
+ Show Spoiler +

Hence phoenixes are just as cost effective as stalkers in terms of soaking up damage from enemy stalkers and assuming sentry dps to be negligible this is the only unit your enemy posses that can destroy phoenixes

- By being selective about which targets your phoenixes attack, ie light units, you can make use of the phoenixes (frankly insane) 16 dps (18 damage in a 1.11 second cooldown) Stalkers by comparison deal 9.8 dps to armoured targets, and cannot be as selective in their targetting (dropping to 6.8 dps against light targets). In the 3:2 ratio this still leaves phoenixes at 16 dps vs 14.7 of stalkers. (Note that phoenixes 'true' dps will be lower, due to the fact that at least one phoenix cannot be firing at the same time)


-Phoenixes become more effective in larger numbers, as the disadvantages brought about by at least one phoenix being unable to fire while lifting becomes negligible. Furthermore space becomes an issue for large groups of ground units, with the possibility of stalkers being unable to fire, this does not pose a problem to phoenixes.

-Phoenixes give a great deal of map control, and provide essentially unlimited scouting of the enemies base due to their unmatched speed. Phoenix harassing is extremly effective, even more so than mutalisk harassing due to the nature of shield regen. Where a mutalisk requires a long time (or a transfuse) to heal, essentially meaning that every point of damage taken by a mutalisk is a cost, phoenix shields regen to full extremely quickly. Hence as long as care is taken to not lose the shields on a phoenix, harassing is almost costless, as stray stalker shots (even from blink stalkers) are often not enough (even with focus firing) to deal sever HP damage to a phoenix.

-Phoenixes exploit a weakness in protoss ground armies It is a nice symmetry that the units phoenixes kill quickly, zealots and sentries, are much slower than the only units that can significantly damage the phoenix, stalkers. As any army moves out stalkers will naturally tend to pull ahead of the rest of the army. With the scouting power phoenixes give, it is a fairly simple task to swing around behind the stalkers and kill of sentries and zealots. This forces the enemy into the micro intensive task of keeping his entire army in a tight cluster, and allows for VERY easy picking off of any stray units.

-Constant forcefielding of the ramp + harassing of the enemy mineral line keeps the enemy's army in his own base.

-Robo builds generally depend on immortals for defense before a colossus can be produced. Since phoenixes are a faster tech, and provide complete scouting of the enemy base, a push shortly after completion of the enemy robo bay, combined with phoenixes lifting up all enemy immortals and low sentry count (due to high gas requirements of the colossus) generally produce an easy victory.

-Phoenixes bypass all forcefield stalls by harassing the mineral line, requiring stalkers to be pulled back, sentries at the ramp can then be picked off by the much faster phoenixes. Essentially, if the enemy pulls his entire army back to the mineral line, you can bust up the ramp, if he separates stalkers from sentries to cover both the mineral line and the ramp, his sentries die, and if he's placed cannons in his mineral line to prevent both he has wasted a great deal of money in stopping your harass, and so cannot attack you, allowing you to expand and win that way.

-Phoenixes usefullness increases dramatically once 5 are on the field. It requires 4 Phoenix shots to kill a sentry, and 8 to kill a zealot. Hence a group of 5 phoenixes can 1shot a sentry (provided no guardian shield is up) and 2 shot a zealot.

-In general I only stop phoenix production after 10 phoenixes, allowing 2 groups of 5 on two separate hotkeys to patrol the map independantly. This requires a huge amount of multitasking, and since I'm not a capable enough player to pull it off without slipping up (though I would love to see it in the hands of a proffesional) I have found that even at my level it is worth splitting them into two groups, Scouting the map extremely effectively to pick of anything that can be picked off.
This coincides nicely with the time you need to start void ray production to deal with mass colossus transitions.

-Dealing with mass colossus balls relies on void rays, but if you fail to kill the enemy before his colossus get out, immediately start void ray production. Instead of lifting single sentries and zealots in a straight up fight, instead lift as many stalkers as you can, using every last phoenix. Get your voidrays to focus down the colossi before cleaning up the stalkers, Phoenix voidray compositions are EXTREMELY effective against colossus balls

-The immobility of colossus balls can be exploited in the same way Terran do: focusing on highly mobile Guerilla tactics.

-The only other transitions the enemy can make is are to:
Dark Templar: Scouting from phoenixes provides plenty of time to get a robo + observer out

High Templar: extremely ineffective as storm is wasted on phoenixes, and feedback can be negated by spam lifting units to keep energy low. (I have never been able to test this scenario as it has never happened to me, esentially because it is such a poor choice)

Stargate: This is now playing your game, and essentially just becomes a resource war (Again I have never been able to test this scenario)

-Blink Stalkers (see note below)

-Mass Gateway Units with upgrades
This is the trickiest response to deal with, if your enemy ignores the colossi all together. At some point you will have to expand, so forcefielding the ramp can not be counted on forever. Furthermore your ground army is substantially weaker. The only way of dealing with this approach is to harass well enough for an economic lead, picking of zealots and stalkers where possible. Then, when a fight occurs, immediately use phoenixes to take out enemy sentries. Without Guardian Shield and Force Field your army can win through good forcefield placement of your own.

Furthermore, zealots are easy pickings for your phoenixes, allowing your own stalkers to target enemy stalkers. Furthermore, while phoenixes prevent large stalker counts, you do still end up with an approximately 3:1 split of zealots to stalkers.

While ordinarily your zealot-heavy army could kited by enemy stalkers, note that phoenixes prevent this to some degree. By lifting off 1-2 stalkers and marching your zealots forward, enemy stalkers must either stand their ground and tank damage from the zealots (your sentry advantage helps with this) or retreat and give up the stalkers that are lifted.

As such, when played correctly phoenixes are an amazing addition to your ground army, and can dismantle much larger ground armies.

*Note* I have not had many encounters with blink stalkers on the ladder, but on the 2-3 occasions that I did I was surprised at how well phoenixes dealt with them, I was expecting blink stalkers to dominate phoenixes.

However:
-Stalkers are much slower than phoenixes, even with blink, and deal very little damage to phoenixes. This combined with quick shield regen means that running away from blink stalkers usually results in very little damage taken.

-Phoenixes actually serve to counteract some of the mobility of blink stalkers. As above, by lifting 1-2 stalkers, and marching zealots forward, stalkers can be individually picked off by your own phoenixes / stalkers.

Alas:
-A large army of pure blink stalkers becomes almost impossible to deal with, void rays are a bad transition because of how well blink stalkers kill them off.

-Blink stalkers overcome the disadvantages of ground units in large numbers, making the stalkers more effective in a large ball than phoenixes, they also bypass most forcefield plays.

Way out:
-The only solution to this I have found is to tech to blink stalkers of your own, cut phoenix production and use the phoenixes you have to harass. Try to deal enough damage so that you can catch up in the blink stalker game.

-As a generally rule, try to kill off an enemy protoss before he can amass a huge blink stalker ball.

*Note* while I have been playing this build for a while, I only had the idea to create a guide after iEchoic's TvT build was published. After looking through the temporary replay history, I could only find one decent example. As such I will link two replays now, and as my laddering produces more games I'll edit them in later.

Replay 1: Custom game against a computer only, this is simply to demonstrate the opening and how the timings work out so neatly.

vs Computer, only for demonstration of the opening timings.

Replay 2: Example of holding 4gate aggression. Note that I make several mistakes (build order mistakes, macro slips horribly during the initial 4gate push, phoenix control was terrible) And still pull out a win. This is simply because 4gate transitions so horribly. Upon scouting robo bay I even begin to get voidrays, but he skips out on colossus all together, when it comes to the final battle my micro was again terrible, but none of it matters due to the large lead

2gate 1stargate holding 4gate