On the timing of this analysis: I came to the conclusion that the time window between 15th and 20th of a month is ideal considering the monthly seasons and balance patches. It means that the "new meta" has established itself already a bit, but there's still time in the season to use the information to try out a new deck, adapt to it, and grind the ladder before the reset. Let me know if you disagree and what other timing might be preferrable.
On the methodology: I crawled through the Top50 Nod players as of April 19, and picked their most relevant deck. The ideal "most relevant deck" is one that is recently played and has a large sample size. In most cases, this is a pretty straightforward selection. There are some cases where two decks could have been chosen – I tried my best to make good decisions based on winrates, recent match list etc., and even if my choices were not ideal, the impact on the overall analysis should be small.
The data: Below, you can find the main table. You can also access the data in spreadsheet-form, including a full list of the Top50 decks. In the main table, the frequency shows how often a unit/commander occurred in the Top50. The ppt Δ vs Mar 28 shows how the frequency has changed compared to March 28 (in percentage points). The Winrate Δ vs average shows the percentage points that a unit's average winrates deviates from the total average, the Levels Δ vs average shows the same for the unit level.
Reading example: Flame Troopers occurred in 22% of the decks vs 48% at the end of March (-26 ppt). The winrate of decks with Flames was 3 ppt higher than the average, despite the Flame decks having 0.6 lower-levelled units on average.
Additional data not visible in the chart:
- Average winrate of the top50 decks is down from 80% to 77% – likely down to the Avatar, which had above-average winrate in the last analysis
- Average unit levels went up from 12.0 at the end of March to 12.6 now. Seems like a big increase for 22 days, I assume it's because many top players make intense use of the cloning vat to level up some core units.
- The lowest level decks are three L11s, run by RussoMau (Portugal), LaFaf (France) and VictorBansemer (Brazil)
- Overall deck diversity has grown slightly (measured in the average deviation from a situation where every unit is used equally frequently), which is good news (ceteris paribus)
- The average deck has 1.8 infantry, 3.2 vehicles, 0.8 air and 0.2 tech units
- There's only a single deck without a war factory
- There's 14 decks without air, 29 with a single air unit, 6 with two air units, and one with 4 air units
- 2 infantry units is the absolute standard; only 3 decks run no infantry, 2 decks a single infantry, and 2 decks 3 infantry units
2-3-1 with Laser, Scorpion, Banshee
16 decks are the Nod 2-3-1 without Stealth Tank, which always has Laser, Scorpion Tank, Banshee; the other roles are taken in the following distributions:- 12x Militant vs 4x Flame
- 12x Cyber Wheels, 3x Buggy, 1x Chem Buggy
- 11x Seth, 5x Oxanna
Some of the players with the lowest unit levels play it, and from the data, the best deck seems to be the variant with Militant, Laser – Wheels, Bike, Tank – Banshee – Seth.
I'm one of the 3 players who tried to make the Buggy work. And whilst it indeed became more viable, I can testify there's many situations where it's pretty weak and clearly inferior to the wheels, still. The typical being against the two staple units that are run by most decks: GDI's Pitbull and Nod's Tank. They quickly get to the Buggy due to speed & raider mechanic & take it out even faster than wheels.
Yes, there are some situations where the Buggy is a bit better – an example being GDI Drones, which recently became more common. But in the most relevant situations, the Buggy is just worse than wheels, and that's ultimately what counts:
Buggy same as wheels:
- vs Militant / Rifle / Flame / Shock – they both kill reasonably quickly & force the opponent to disengage if at all possible
Buggy worse than wheels:
- 3x the cost
- less vision
- less speed
- less sustain vs Tanks due to the squad mechanics
Buggy better than wheels:
- vs single Laser Troopers/Missile Troopers
- vs light air (e.g. Droneswarm, Talon)
My best suggestion would be to actually flip the vision range of Wheels and Buggy. It gives the Buggy a unique advantage. And Wheels are still viable as scouts due to speed and low cost (you can scout much faster still). Also, it would differentiate it more against the relatively weaker GDI Wardogs (except if you did the flip there as well).
RussoMau consistently plays the Militant-Wheel-Seth version with a very high winrate considering the (for top50) low unit levels
The New Allrounder: Stealth Tank
50% of all decks run the Stealth Tank. The distribution shows that it's both useful in the anti-air and the anti-vehicle role. This makes them viable in a broad range of decks: Of the 25 Stealth Tank decks, we have ...
- 8 Avatar decks
- 14(!) of the 16 decks that don't use Scorpion
- 3 of them Giga Cannon decks, where they serve as the Anti-air
- 19 Oxanna decks (vs 3 Seth, 3 Jade)
What can we learn from this?
- If you don't want to run Scorpion, the Stealth Tank seems to be the only option (besides trying to make a non-factory Nod deck work)
- If you want to run Giga, strongly consider the Stealth Tank; not just 3 of 4 Giga decks use it, but they also seem to work better (winrates vs level)
- If you still want to run the Avatar, run the Stealth Tank as well
- If you run the Stealth Tank, use Oxanna
LaFaf switched from a seemingly anti-Avatar deck (with Stealth Tank) to a Giga Cannon deck (with Stealth Tank), showcasing the versatility of this unit in the hands of skilled players
Avatar's fall from Grace
Only 18% of decks (vs. 44% at the end of March) run the Avatar. This is obviously related to the recent balance patch, which delivered on some of the Avatar nerfs I hoped for.
Last time, Avatar decks had a higher-than-average winrate. This time, it's a bit lower than the average – despite the Avatar users running higher-level decks.
Beyond the patch, a reason certainly is that players adapted to the Avatar: They learned who plays it, learned to look for the timings, and ran more counters (in the GDI analysis, we'll likely see a growing use of Orcas). If you still have trouble with ir, check this article on how to play against the Avatar (it's from before the patch, but the general idea stayed the same).
The most archetypical screenshot I could find: Repsycho switched from the standard Avatar deck to the Standard 2-3-1 after the patch
HusseinTamvan shows that the Avatar is still very strong – for the Top50, Hussein's unit levels are below-average, and his winrate significantly above-average
Inferno – strong, but not too strong
If you read Reddit and Discord, you'd assume more people would play the Inferno. But actually, Inferno usage is at just 16%, having growth by 2 percentage points. I feel this is a good spot for a heavy air unit to be in frequency-wise.
Winrate-wise, they seem very strong – but this might be propelled by highly proficient Inferno users like Magni, VictorBansemer, AliciaDestiny and wwfyin – who all sport excellent winrates of 92% and higher.
Some argue that you need high unit levels to play Inferno. That might be true, but on the other hand, these four 92%+ winrate players I mentioned have L12 decks (one even L11), which is below average for the Top50.
My hypothesis is that Inferno needs a highly skilled user ...
- to get there (1-2 fewer cheap units compared to the 2-3-1, depending on whether it's a deck with or without Phantom)
- to successfully use the Inferno (it is quite fragile and especially when Oxanna-boosted, very quickly dies to a single Pitbull whilst recharging)
This is also illustrated by the fact that Inferno is not a clear single- or dual-harvester unit. Yes, some lower-ranked Inferno users always go dual harvester – but it's not necessary, and e.g. Alicia plays her Inferno deck usually on a single harvester. Speaking of Alicia: Her favorite deck with Flames, Wheels, Bikes, Tanks, Infernos and Phantoms is a great example for the skill needed to get Infernos out. As Alicia's deck sacrifices Lasers, and Infernos are quite expensive, she has to win early Tank battles even against higher-levelled tanks. That's actually quite hard to pull of.
Nod's inferno – expect it to stay relevant
Honorable Mention: ZeroHour's no-Harvester Deck
ZeroHour runs a very interesting zero-Harvester deck, built around flooding the battlefield with cheap units (especially Militant, Laser, Wheel, Bikes). The goals are to keep pad control with superior numbers (harvester goes against the pop cap, so if Tiberium allows, ZeroHour has a unit more on the map) and to ideally score a harvester kill that allows to pay for a Stealth Tank to snipe some Tank or Air unit that the other player squeezed out.
As you can see, ZeroHour's winrate has been incredibly high for unit levels that are – for the Top50 – not that high. This is testament to his skill and that this deck works.
Do you have the data for gdi as well?
ReplyDeleteYes!
Deletehttps://cncrivals.blogspot.com/2019/04/top-51-gdi-decks-in-april-orca-and-apc.html