Laddering on Korea Part 2: The All-in

After playing thousands of SC2 games and watching hundreds, I’ve seen almost every variation of all-in used in pro play and on ladder. Many of them are failures and some are successes. There are many factors on why an all-in has failed such as bad execution, wrong build order or being scouted too early.

Here I will write down the three types of most successful all-ins so that you can discern exactly how stronng an all-in build actually is when used in competitive play and when on ladder.

1. The most efficient build possible

The first type of effective all-in is the most efficient build possible. This build relies purely on execution and control. Currently the best form of this type of all-in is the WonWonWon or the Soul push. This build relies more on proper control of units and a very clean build order to bring out its full potential. It doesn’t matter if it gets scouted or not, as long as the protoss controls properly, he can win.

2. The Reactive all-in

This type of all-in is predicated on scouting information and is generally done completely impromptu. The best case example is the game between Oz and Squirtle on Squirtle’s first run through GSL.

Oz opted for a Nexus first, a never before seen PvP build. Squirtle had started with an anti 4-gate build with double Stalker rush. Instead he changed it to a double Stalker pressure and then transitioned into a very fast 4 gate that killed Oz. Other examples include catching Terran go 3 Orbital mass Marines and doing a baneling bust or a Cannon rush if Zerg goes Hatch first.

3. The Deciet

This type of all-in is based on giving the opponent false information about your build and timings. For instance, the best example of this is in ZvP when a Zerg goes triple Hatch, but then follows up with either a Ling all-in, a baneling bust, or a Roach-slow ling all-in. The protoss scouts all three Hatches and then skips on the Sentry to tech faster or do more Zealot Stalker pressure, but then dies to an unscouted all-in.

4. The Player Snipe

While the above three are the main types of strong all-ins, there is a fourth type. It’s only possible in a professional setting when both players have time to prepare for one another. It is a build specifically designed to kill one player for one time in one series and would probably never work against any other player of the same race. The first case is Nestea vs Anypro.

Anypro was the first Protoss who made the Forge first build into a 7 Gate Blink Stalker timing that would become 1 of 2 key factors that would buff the Infestor. Anypro had used this build and had beaten every Zerg player of the day with this Stalker timing. Nestea in the first game did a Lair rush into a Spinecrawler attack on Anypro’s natural on Dual Sight. This would break Anypro’s natural and his mind for the rest of the series as he was too afraid to ever make another forge expand. This build would incidentally make mapmakers put the natural on the high ground so that Protoss could actually fast expand in games. Also the map Cloud Kingdom would become home to Bly’s variation of the same build against Protoss (which he used against Sase) and would later on be used by Curious in the GSL.


These are the four types of really powerful all-ins used in SC2. In some cases, a build can be 1 or 2 of these things such as a proxy 11/11 rax which is both an efficient build order as well as an all-in based on deception. In some cases a reactive all-in becomes an efficient all-in like MVP’s double Starport pull SCV timing against Protoss. He did that reactively against Parting in his GSL Code S run, but he later used it again as a standard build against Rain.

I hope this has been informative to all of you so that you can recognize more of the reasoning behind why some all-ins are stronger than others.

Post nav practicePost nav contentPost nav coaches