TvP: A Dog Chasing its Tail

I have been asked about the current state of TvP frequently this week, and I think it is simply a dog chasing its tail. Let’s take a look at how things turn out to be how it is now.

The Phoenix Adept composition in TvP has been heavily debated in the community, as it is something warrant attention. At the end of the last post, I have mentioned that I scrapped an article about Phoenix Adept, because Artosis’ recent article echos what I wanted to say. Instead of repeating the same points, I will start the discussion by going back to the middle of 2016 and look at how things shape up to what they are today.

Mid 2016

Protoss were ahead in the metagame in the middle of 2016. Terran had problem dealing with the standard macro build – One Gateway expand into Robotic Facility then Twilight Council, and that was not a good basis of a solid match up. This was evident when Terran players used a wide range of sub-optimal builds, while Protoss players were able to stick to one solid macro build.

The struggle could be attributed to the fact that both sides were still using the old framework from Heart of the Swarm, and this outdated approach was in favour of Protoss in Legacy of the Void. Terran had problem adjusting to the poor reliability in scouting in early game with Reaper, while Protoss still had a good range of trick builds just like they did in Heart of the Swarm. Further, Protoss could play the same defensive style that they used to, but they could have an earlier and safer third based than Terran.

This asymmetry in the understanding of the match up led Protoss to further get ahead in the same period by cutting corners to abuse Terran’s uncertainty in the early to mid game transition. Consequently, Stargate opening became more popular, as Terran had problem scouting it. During this time, Terran had been exploring different options like Cyclone and Tankivac in the early to mid game transition, but the results were mixed.

Late 2016

Terran started to catch up in the match up by developing solid builds that put them on par with Protoss’ builds. The builds could be broadly categorised into one of the two:

  1. Bio-oriented: 16 Marine double Medivac Stim
  2. Tech-oriented: Drop and Tankivac timing

The bio-oriented build is essentially the famous 5:00 2-1-1 Stim build used in TvZ. The tech-oriented build is an improved version of some trial and error builds in middle of 2016, and this build usually involves a Widow Mine drop and a Tank push follow up. I consider the match up to be 50-50 at that point of time.

Patch 3.8 then kicked in with many major redesigns. There were quite a number of interesting changes to the metagame immediately, for example, Cyclone timing push was one of the most common builds. More importantly, there were two phenomenons that eventually became the centre of balance discussion, and had a huge influence in the current state of the metagame.

  1. Mid game Siege Tank push buff
  2. Cost effectiveness of Widow Mine and Liberator

Tankivac was removed, and it was compensated with stronger attack damage. It was hard to say whether the change was a nerf or a buff in the TvP match up at that point, and Terran just continued to use the Widow Mine drop into Tank push build. This later became a heavily discussed issue in the early 2017, which I would discuss later.

In comparison to the Tank push, the cost effectiveness of Widow Mine and Liberator did not receive as much attention. I have this belief that INnoVation’s builds are usually a good starting point to understand what are consider the metagame builds, as he almost always plays the “best” builds (you can argue that he defines the metagame anyway). INnoVation was using an interesting 3-1-1 convergent point with three Barracks (two Tech Labs and one Reactor), one Factory with Reactor, and one Starport with Reactor. Despite 3-1-1 is a normal convergent point since Wings of Liberty, INnoVation’s use of add-ons suggests he was favouring tech over bio. Typically, Terran do not have an add-on on the Factory for 3-1-1 in TvP, because the production focus is on bio and Medivac. The Factory only has an add-on after Terran have reached the 5-1-1 convergent point, which again emphasises on bio production (read more about it here). The continuous production of two Widow Mines in each production cycle delays the two additional Barracks to reach 5-1-1, and this delay is accentuated with the production of two Liberator in each production cycle. Therefore, Terran simply cut the number of bio units in favour of Widow Mine and Liberator, and that was the evidence of the cost effectiveness of these two units. This new composition also changed Terran’s tactical approach to make a deathball-ish style in the mid to late game.

Early 2017

The Tank push in the mid game was considered to be too rewarding for Terran, and it became the talking point of the match up. Blizzard had stated that they were aware of the concern for the match up, but they wanted to allow the players to have more time to make adjustments.

Tournaments have started going this year, so we should look for any “absolute must” balance changes that need to happen. We definitely see the concerns such as early/mid game PvT right now, but we’d like to strike a good balance between giving players enough time to adjust to new strategies vs. stepping in to make balance changes.

As you can see from the quote above, Blizzard did not explicitly identify the issue, but had only highlighted the temporal factor of the situation. That means, Blizzard did not narrow down and attribute the concern to the Tank push. Consistent with this, their subsequent community update did not suggest to make a change directly involve Siege Tank, and instead they singled out Liberator.

Liberator Concord Cannon damage changed from 85 to 75.
While there has been a lot of great feedback regarding suggestions for PvT, this change looks to be the solution proposed by many players out there. However, we wanted to confirm with you before making this move. Because a change like this targets the specific problem and is a very small change outside of that, we wanted to get your thoughts on possibly pushing this out to the live game as early as next week.

… We are currently still seeing discussions around… Widow Mine strength in PvT…

Thus, Blizzard had made it clear that they were targeting the second of the two phenomenons I stated earlier in this post, and that is the cost effectiveness of Widow Mine and Liberator. Incidentally, the above community update was posted ten days after I had made a post about the state of TvP here. In that post, based on the observations I had made about the professional games, I concluded that the match up was on the verge of a metagame shift. This was because, similar to what happened in mid 2016, the situation was that one side continued to do what they had been doing previously as it was in their favour, while the other side was trying to test new ideas to overcome the unfavourable situation. The difference was that the coin had flipped, and it was Protoss that had more questions than answers instead of Terran. Interestingly, the adjustments were mainly targeted at the first of the two phenomenons, which was the Siege Tank push. For example, two base Colossus into a later third Nexus was a common build. Therefore, the Protoss players had identified the Siege Tank push as the main issue they had, while Blizzard had identified the cost effectiveness of Widow Mine and Liberator as the priority of balance concerns. A balance patch to nerf Liberator’s ground damage from 85 to 75 was live shortly after the above community update.

In my opinion, it is this antithesis that leads to the current situation of “a dog chasing its tail”. I agree that Liberator was too cost effective in TvP, and the nerf appeared to be appropriate. However, the change mainly affects the mid to late game of the match up, while Protoss players were concerned about the early to mid game push from Terran. If I was right about the match up was on the verge of a metagame shift, it was only a matter of time that Protoss would come up with an innovation to tilt the scale in their favour until Terran had something to one-up it. Then, the nerf on Liberator could potentially “over-corrected” the match up once Protoss figured the early to mid game issue out.

Current situation

Protoss’ answer to the Siege Tank push was Phoenix Adept, which was heavily discussed in the community recently. I wrote a post about it after I had witnessed its success in GSL and IEM. I considered it as an anti-mid game aggression build, and its timings matched up against the Widow Mine into Siege Tank build very well (I won’t discuss about it here, read that post). My proposed solution was to adjust the mid game plan to be more passive, and add Ghosts into the unit composition. It was also this period of time that Blizzard further nerf the cost effectiveness issue with another patch, and reduced Widow Mine’s splash damage +shield bonus from +40 to +25. All of a sudden, the early to mid game push problem was resolved by the Protoss players, and Terran had to adjust to this innovation and the two patches at the same time. Essentially, the two phenomenons turned from favourable to problematic altogether all of a sudden for Terran. More importantly, Widow Mine is the best answer to Adept to begin with, so the timing of this nerf could not have been worse.

Then, while Terran were trying to figure things out, Blizzard swung the hammer again. This time the target was Adept’s health, and this decision was made rather swiftly as the patch went live today after Blizzard had announced its intention to nerf Adept in a community update post six days ago. When you look at it retrospectively, both Protoss and Terran got nerf, and Protoss came up with something to neutralise the early to mid game push. Zerg must be laughing now.

The intention of this post was not to blame Blizzard’s balance team. I understand they have a tough job, and they cannot please everyone. They got criticised for making changes too quickly, and also for doing the same too slowly. Thus, I will not go down that path, but rather I want to highlight three things that caught my attention.

First, Blizzard identified the problem lies in early to mid game for TvP, and they wanted to give players time to figure it out (posted on 6 January 2017). They then nerf Liberator three weeks later (patch went live on 31 January 2017). It makes me wonder whether my understanding was wrong about the early and mid game issue we previously had in the match up, as it occurred to me (and many others) that the issue was the Siege Tank push and not the Liberator. If my understanding was correct, then why Liberator was nerf to help with the early to mid game problem that was correctly identified by Blizzard? It seems like the intention was to balance from a standpoint of an overall win rate instead of the issue of concern directly. Then, this leads me wondering why Siege Tank push was not addressed, and how it could be addressed. The Siege Tank changes in patch 3.8 was mainly targeted at TvT and also little at TvZ, so no one foresaw the problem back then. I think it was fair, because Siege Tank was basically an unit that takes the back seat in the match up. Thus, making changes to Siege Tank directly to address the TvP issue may indirectly affect the other two match ups, so it was a risky direction. Therefore, the Liberator change was arguably the best option for both short and long term.

Second, the Widow Mine nerf was made five weeks after the Liberator nerf (patch went live on 7 March 2017). While I understand the rationale behind Liberator nerf, I do not agree with the Widow Mine nerf. This is because it is the same period of time we see the rise of Phoenix Adept, and it is an ideal situation for Blizzard’s stand of “wait and see” philosophy to balance changes (something I agree with). As a result, the Widow Mine nerf appears to be poorly timed at best, and unjustified at worst.

Third, the Adept health nerf patch was live today. The community appears to have a consensus that Adept needs a nerf for cost effectiveness reasons in PvT and mobility reasons in PvZ, but opinion is split regarding the change of decreasing Adept’s health by 10. At face value, I believe it is too big of a nerf, and it can potentially lead to another loop of nerf to balance the previous nerf (i.e., a dog chasing its tail). You must consider the total health lost in a battle whereby there are multiple Adepts (see a quick comparison). My intuition tells me that the other reason behind the change is to let Zealot fills the tanking role, but the difference still does not justify the decision to pick Zealot over Adept (see an argument about this on reddit). Therefore, I believe the recent Widow Mine and Adept changes are rather sloppy by Blizzard’s high standard. I am inclined to have both changes reversed, and the current Phoenix Adept situation should resolve by itself as Terran players improve. This also gives Blizzard more room to focus on Adept’s mobility issue in PvZ and not have to worry about “overnerf”.

Advertisements

2 thoughts on “TvP: A Dog Chasing its Tail

  1. Great article, as always! Thank you, Max! Waiting for more!
    Sorry to disagree but I think Adept nerf was needed. Massing Adepts was way too good in both PvT and PvZ. There wasn’t a situation where making other Gateway unit was better than making Adepts.

    1. I actually think that stats wise, Marine may be up there at the top if you want to mass a type of unit. But the issue with Adept is that it’s a basic unit that fills too many roles, and it can do it all by itself. The only “weakness” is it can’t shoot up. Nerf-ing its stats is not that bad, but Zealot needs a buff and it is on the agenda.

What do you think?

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s