Well we may as well move the thread from http://zero-k.info/Battles/Detail/861594 to here.

[Spoiler]
In the early game ATOSTIC spends 525 metal (7 Bandits) to keep 500 metal of Darts and Fencer locked on the other side of the map.

[Spoiler]
By ATOSTIC is tying up 910 metal in Rovers with the remaining three Bandits that originally only cost 525 metal. Behind this he has expanded super-greedily with no turrets. His only other combat units are an Outlaw and Engineer Commander.

The Rovers keep chasing until 3:20, to the mid-left start position.

[Spoiler]
When the Rovers get back to the middle they sit around for a while until a Dominatrix reinforces them, then blindly assault an expansion. By this time ATOSTIC has built another 590 metal of army (2 Thugs and an Outlaw) and is starting a HLT.

[Spoiler]
The Rovers find a completed HLT, do a bit of fighting, but they don't want to enter range and ultimately nothing dies on either side.

[Spoiler]
The Rovers now blindly attack a different expansion, only to find it empty. The Shielbots that came down to defend the HLT expansion attack a barely defended expansion which they spotted on radar. Another 525 metal in Bandits are sent to raid. These are spotted by the Rover army in time for a Mason to make an incomplete Stardust.

[Spoiler]
The roaming Rover army of 1960 metal find a relatively weak Shield army of 1280 metal. The Rovers attack piecemeal. The Shielbots are aided by the very short reinforcement distance of having their factory just off-screen, and their commander.

Two Rippers and two Ravagers are lost without inflicting any losses on the other side.

[Spoiler]
The rest of the army swings round back to the middle to fight off the south Shieldbot army. Dregs switches to pure Recluse but in doing so has an even slower army and is behind on economy (once ATOSTIC fixes his).

As an aside, ATOSTIC queued a large block of Wind nanoframes in the top right and may think they they are generating energy. They are not. Perhaps the nanoframe queue was accidental.

ATOSTIC makes a Fusion shortly afterwards and outproduces Dregs until the game ends. None of his mexes die.

I should start by saying that I don't know how to play Banana Republic. The start boxes encompass the entire side of the map and there are many viable looking start positions. I have not played it enough to figure out the pregame RPS relationships of choosing each spot, or to identify good expansion patterns. For all I know ATOSTIC just picked a better start position.

Anyway, assuming that Banana Republic is being played properly, I think ATOSTIC is getting better at outplaying Dregs. His raids always tie up the opposing army or do damage. He only ever defends where he needs to. I don't think he ever loses a mex and almost all of them are left undefended all game.

Looking at the stats limited to loser elo at least 2500, Rover has won this matchup seven times while Shield has won five. I had a look at a few of the Rover wins and the difference is that Shield made Rogue in response to Dominatrix. We've seen Felon in two games. It looks to me like if the Shieldbot player does something different then the Rover player had better do something different too.

quote: Tried a bit of fencering the ball, didn't manage to kill a single thug in early game. Just wasted time. Also opted not to try raiding because outlaws/bandits, especially in chokepoints on that map would easily have stopped me from doing so and I don't like donating metal where I can avoid doing so. Maybe my best option was to go mostly ravagers but even still, bandits, rogues, outlaws and snitches would have done me in. I decide to switch to recluse spam and ATOSTIC simply makes a good play with cloaked snitches, GG.

I've watched the other games and you do basically the same things there as you do here. The difference is that they don't work here. Also I don't know why you scout and raid in other games but barely do so here. Your first serious assault is six minutes into the game.

In other games I see you fight the same type of spread-out piecemeal battles. That works against Rogue but not against Felon. Sending units in piecemeal will just get them picked off by Felon and you will lose on attrition. A Felon is slow, expensive, and requires support. When your opponent has slow expensive armies you need to attack everywhere it is not, and only attack head on when you have an advantage. Rover mobility gives you greater choice over how battles start.

I think not raiding is a big mistake. As ATOSTIC showed this game, a raid can be very useful even if it loses you more metal than you kill. Fencer is barely slower than Bandit, so what happens if you poke around the sides with pairs of Fencer and Bandit? I have seen you do this in other games. The shieldbot player has to spend more metal to push back this combination and it puts their units out of position.

quote: If playing SBs myself (or fencer spam or dagger spam on other maps) is the answer, then I lose appetite for ZK because I'm not really playing a game with options, I'm following a ruleset to the letter instead. Factory RPS and metas where certain factories are nearly mandatory are the most limiting things to gameplay here.

Nobody is saying that playing Shieldbots is the final answer, and that games will just be Shieldbot mirrors for all time. You are taking it to the extreme. It is the answer to the questions "how do I show that this is OP?", or "how do I learn how to beat this?".

One of the best ways to learn the weaknesses of a strategy is to try it yourself. What ATOSTIC is doing to make enough room for Felon is not trivial. Refusing to play one side of a matchup hampers your ability to learn it. You'll have to put in more work to get the same out, such as looking at a replay in detail like I have done these two times. By 'work' I don't mean to say it is boring. It is the kind of work to extract the "learning the intricacies of matchups"-fun out of playing 1v1.

quote: For clarification, my view on OPness/balance that Dregs is referring to is that I try not to complain about balance or something being OP, I much prefer to use what I believe is OP so I will either a) win a lot and others will start to see what I see, or b) I will be shown how to play against it or get ideas myself.

I agree, but not in all cases. I particularly advocate for playing what you think is OP because it is a good way to learn how a strategy works and gives people who disagree with you a chance to prove it.

This approach fails when the meta decided upon a matchup and stops playing it. For example, if everyone played Shieldbots instead of Rover there would be no opportunity to explore Rover strategies or to gather data on whether there is a problem. We don't have the thousands of hardcore 1v1 players required to rely on someone like Bisu coming along and upending the meta with dedicated experimentation and testing. If we reach a point where a lot of people are playing Shielbots instead of Rovers then it is time to consider changes even in the absence of data. However, we currently have very little data and I expect people to keep playing Rovers, so I don't see that it is time to make a change.

quote: Postgame chat Randy says he doesn't agree with my perspective but I honestly can't see what options I have as rover here. The fact that so many people can argue against this for so long has frustrated me to the point where I'm honestly confounded. Really pissed off that I'm having to expend so much energy on something that to me, appears glaringly obvious.

ATOSTIC has developed a strategy that beat Dregs twice. Previously Rover was dominating Shieldbots with Dominatrix. This strategy beat Shieldbots more than twice. If I were to apply the onus of proof then I would have nerfed Dominatrix a week ago. It may well be the case that ATOSTIC has developed something that demonstrates shieldbots being a little overtuned, but part of the fun is in figuring that out and then figuring out how to counter it. I'll be more interested when others repeat this strategy.

Looking at the very early data I'm expecting Rover to need a buff for the Hover and Tank matchups before shields need tweaking.

UltraGodzilla was opining at length not long ago that rovers were overrunning the meta on pretty much all MM maps outside Intersection. There have been balance changes since then, but I'm having trouble squaring all this with your assertion that shieldbots have been OP for the last year.

Although perhaps you are not counting the blip caused by the Ripper buff? idk. I am, to be honest, having difficulty seeing past your frustration to what the details of your argument are.

quote: Have you ever considered WHY people are picking shieldbots more frequently on what were rover maps now?

I have been toying around with https://zkstats.antihype.space/# and am failing to find evidence to support this statement; if anything the opposite is true compared to some months ago. Perhaps it only applies at high enough ratings that data is difficult to obtain though.

I'm not at all in tune with the 1v1 meta these days, but it would not at all surprise me to find that shieldbots are overtuned. It's still important to be careful about knowing exactly what's going on.

It's not me you have to convince, but it does take a lot of data points (in this case, games) to be sure about this kind of thing in my experience.

(And yes, a shieldbot player beating a rover player on Trojan Hills isn't tremendously shocking to me.)

It super duper wouldn't surprise me personally to find that Outlaw is still overtuned, for what that's worth.

quote: Meanwhile most of my contemporaries have already said they're willing to just play what's OP.

Are the rest of the MM top 10 in fact playing a lot of shieldbots? esainane's stats tool does not seem to think so but perhaps I am using it wrong.

quote: You guys don't trust me by now or what?

Even at points in time when I was much more of a 1v1 player, I wouldn't have trusted myself to have opinions about factory balance that were free of personal bias.

Randy donated: B863383 2 on Intersection v4.1 to the cause.

I remember this happening with Cyclops, and indeed switching to the dark side promptly demonstrated the absurdity of the situation.

I think shields is a strong choice vs rover as it counters a good deal of the fac very efficiently indeed however that doesn't change the fact as while it is great vs some factories it is not so great vs others such as jump and spiders.

Factory RPS is most certainly still a thing I feel like and it is not necessarily easy to balance a factory particulary towards a single factory without possibly causing significant changes in how it plays vs other factories.

I would personally ask the following

How closely should we stray from allowing factories to have their advantages on their "Favored" map type and non favored map types?

Is there data to support whether one factory is winning vs others too much?

What particular units seem to be the main issue and why?


Personally I find factory rps to be moderately annoying however I have found rover will generally play on a pretty much even playing field vs shield granted both players play properly/use the appropriate counters.

I think shield most certainly holds an edge on smaller maps that have choke points vs rover that can be abused but rover has powerful raiding tools that can be difficult to catch on maps that have the space.

I think each factory has distinct advantages over the other that still provides them a chance to one up the other.

Not much can stop ravagers from running around a shield ball and killing back eco for example.

Rippers kite thugs/outlaws forcing rogues.

In which case the only answer here is dom. Which is not very cost effective vs rogue in many situations unless you are tediously dodging all the shots.

I think this is possibly a situation where the fac who has the better skirm receives a notable advantage such as when recluse fights shields or moderator fights tank ect.. so perhaps its actually the skirms which may need to be looked at as a whole since they potentially lead to some rps styled gameplay






Also after watching this replay I feel like that corner is in my opinion very inferior compared to the top version that has an easy to access geo right at the start of the match. I always found the geo corner and mid spawn near super mex to work the best.

A lot of what you're saying is either late to the discussion or has already been answered UG. Hell, I think you already know the answers to some of the questions you're asking.

In an effort not to loop back in circles over already covered angles of this ongoing discussion, I'll just re-summarize: Outlaw is too effective and thug is too tough and available. You, as a proponent of shieldbots already know why.

Map based fac rps though, yes, will always be a thing. Not much to be done there. We're now in the territory of shields taking any choked flats from rover with little to be done about it though. Eye of horus, fairyland, BR. Maybe otago, mercurial.