Loading...
  OR  Zero-K Name:    Password:   
Title: MatchMaker 915
Host: Nobody
Game version: Zero-K v1.6.1.6
Engine version: 104.0.1-141-gae5ba3c
Started: 6 years ago
Duration: 2 minutes
Players: 2
Bots: False
Mission: False
Watch Replay Now
Manual download

Team 1
Chance of victory: 51.6%

USrankFealthas
Team 2
Chance of victory: 48.4%

CHrankAdminDeinFreund
Spectators
ATrankATOSTIC
NLrankWesley

Show winners



Preview
Filter:    Player:  
Page of 2 (28 records)
sort
A quick list of today's games:

Here are the games that managed to get past the first engagement (>10 minutes):
Multiplayer B483540 2 on Lonely Oasis v1.1
Multiplayer B483527 2 on TitanDuel 2.2
Multiplayer B483524 2 on Intersection v4.1
Multiplayer B483509 2 on Ravaged_v2
Multiplayer B483511 2 on Adansonia v4.1
Multiplayer B483560 2 on Trojan Hills v03
Multiplayer B483554 2 on Fairyland v1.0
Multiplayer B483550 2 on Vittra v2.1
Multiplayer B483543 2 on Red Comet v1.3

Here are the games that concluded during/before the first engagement, lasting less than 10 minutes:
Multiplayer B483530 2 on Ravaged_v2
Multiplayer B483513 2 on Red Comet v1.3
Multiplayer B483512 2 on Intersection v4.1
Multiplayer B483566 2 on Lonely Oasis v1.1
Multiplayer B483564 2 on Onyx Cauldron 1.8
Multiplayer B483556 2 on Intersection v4.1
Multiplayer B483555 2 on Obsidian_1.5
Multiplayer B483551 2 on Barren 2
Multiplayer B483545 2 on Obsidian_1.5

Yes, that's half of the games being decided mostly by blind RPS.

Multiplayer B483540 2 on Lonely Oasis v1.1 was the longest game, lasting a whole 20 minutes. Zero-K's comeback mechanics aka reclaim and mex rebuilding don't even get a chance during these games. They both require the player to have a stable energy income which is not affordable for such short matches. I'm asking for double com in 1v1 not because it's impossible to have good games with the current mechanics, but because they make having a good and interesting game so hard. If the balance tips just a little during the first few minutes the game quickly snowballs into a victory. Teams is popular because its increased starting economy means you can at least put up a fight.
+4 / -0

6 years ago
IMO the problem is just terrible factory balance and terrible maps(isle of greif, avalanche, barren, vittra, badlands, intersection). There is a huge lack of vehicle maps in the MM pool too. Where is desert needle small? Alien desert? CCR? Taking these out in favor of the porcy, rush, RPS determined small crap maps makes no sense to me.

Cloaky is OP. Too versatile. What you can do with a bunch of warriors requires 2+ units from other factories. Rocko is OP.

Spider units are just meh in general. Expensive and low hp.
Shield bots have nothing to deal with skirmishers(unless op racket spam).
Amphib is too slow. Speed buffs across the board.
JJ- pyro sucks, probably keeps the rest of the factory from being very op. Sumo/placeholder/jack/firewalker are all stupidly strong.


Skirmishers in general are too good. The optimal strategy is to escalate to higher ranged skirmishers or units(or heavies). I would like to see all of them get nerfed in some way. Perhaps reduced movement speed while reloading or something like that so they can't just kite endlessly with 0 support.
+4 / -0


6 years ago
Reduced movement speed while reloading is an interesting idea, although it may be hard to convey.
+0 / -0


6 years ago
CHrankAdminDeinFreund I am not convinced by what you are implying.
+1 / -1
My problem is that it's exceedingly difficult to have a balanced 1v1. Not in the way that units are in perfect balance, but the game itself. In order for a game to be fun, there needs to be some back and forth. Every player should have the feeling that he was close to winning at a point.

Of course zk should remain competitive so this only applies to players closely matched in skill. I had a win rate of around 50% during these games so I would expect there to be some rather close games. What happened was the opposite: each game was decided during the first conflict after which the economies snowballed disproportionately.

The reasons for this to happen are:
  • You start with very little metal and one factory, losing just one unit early on is a huge hit on your economy
  • Your factory/initial units might be badly matched to your opponent. Rebuilding is not affordable without ruining your economic growth.
  • Maps are tiny, meaning there is little time to respond and you are forced to make unfavorable decisions in order to defend yourself timely.
  • Scouting is expensive:
    • You can't react well to what you find because you already used your starting metal.
    • The enemy might have chosen a strategy where you would've better invested the metal in economy/defense.
  • Defenses are weaker than ever. Defender used to keep skirmishers in check. Now, the only way to defend yourself against a skirmisher ball is more skirmishers or unaffordable mid-game units/strategies. This means you are forced to always match your enemies army. This makes switching factories while on the defensive hard to pull off.

In the end, each player needs to follow very strict rules in order for a balanced, fun game to happen. (Assuming your opponent knows the game) Winning without your opponent having any chance to fight back isn't very fulfilling while it feels even worse for the opponent. The game should be designed such that it lends itself to interesting games that are hard to decide in the first few minutes.

There are many ways to achieve this:
  • Boost the starting economy. This makes attacks on the base harder and early losses snowball less.
  • Buff defenses. This slows down the game and allows it to progress into mid-game by requiring heavies/artillery. Alternatively nerf skirmishers as these are the usual early-game counter to defenses.
  • Play on bigger maps. This allows the game to progress further before each fight due to the increased travel times. It also magnifies the defender's shortened supply lane bonus.
  • Make maps more defensible. Following Licho's idea of clustering metal more towards each starting position, this would make it easier for the metal income to remain balanced and lessen the impact of taking early map/mid control.
  • Although I'm not a fan of it, revealing the starting factory can simplify defense against some rushes. I think it nerfs some strategies much harder than others though.
+3 / -0
quote:
Yes, that's half of the games being decided mostly by blind RPS.

I've only seen those of these games which i played. Of those, two were mirrors, and one was with different factories, with you complaining that my pick was "lucky". The one that ended in two minutes was a mirror and not necessarily blind RPS - but you deliberately delayed scouting in face of cheese. The other mirror "lasted longer than 10 minutes" but was realistically decided when i suicidally dove in 3 scorchers - first engagement! - failing to achieve anything (despite myself having scouted!)

I find this casts severe doubt on validity of your selection: it feels suspiciously cherry-picked. I do somewhat agree with the theoretising that follows ("games are often decided early"), even if i assume it naked and not supported by any evidence, tainted or not.

However:

quote:
boost the starting economy. This makes attacks on the base harder and early losses snowball less.

The most straighforward ways to do this also lead to more expensive cheese availability. Which means more games decided early.

quote:
Buff defenses. This slows down the game and allows it to progress into mid-game by requiring heavies/artillery. Alternatively nerf skirmishers as these are the usual early-game counter to defenses.

"Nerf skirmishers" wins here. Mainly, skirmishers beat defenses by either outranging them, or out-alphanig and out-lanchestering the simultaneously. A quant-ish approach would be to nerf their health across the board.

quote:
Play on bigger maps. This allows the game to progress further before each fight due to the increased travel times. It also magnifies the defender's shortened supply lane bonus.

Bigger maps with same metal density greatly reward exponential expansion and have higher incidence of runaway economic victories. Bigger maps with same metal value - and thus lower density - barely exist. The practical way to implement this is to get mapping then and have enough new high-quality sparse-metal maps to fill out the MM pool.

quote:
Make maps more defensible. Following Licho's idea of clustering metal more towards each starting position, this would make it easier for the metal income to remain balanced and lessen the impact of taking early map/mid control.

I don't think "make expansions defensible" was the point of Licho's idea. I think the sparseness ( described above) was, so that maneuver means more and lines solidify later. But some core expansions being defensible in this paradigm may be good. Again, this requires new maps.

quote:
Although I'm not a fan of it, revealing the starting factory can simplify defense against some rushes. I think it nerfs some strategies much harder than others though.

It is an ugly solution. You are now feeling an ugly feeling inside of you for having mentioned it.
+1 / -0
In Multiplayer B483545 2 on Obsidian_1.5 I didn't go for a scout because it would have severely weakened my strategy, which was supposed to be a surprise raider attack from behind, using the mountains to cover my approach. You countered that with your warrior rush. Both of these are common strategies, yet they decide the game before one sees the other.

If I am to scout, I would have a hard time defending against the raider rush strategy due to it being invisible to scouts. If I don't scout you might have any build that counters mine. Whatever I decide, it will have a major impact on the game that follows by being a purely blind decision.

In Multiplayer B483543 2 on Red Comet v1.3 you lost the game because you blindly chose the "rush raiders" strategy vs my "rush llt and eco" strategy. I might not have classified the game correctly, but this doesn't change the fact that this was yet another case of blind RPS.


quote:
The most straighforward ways to do this also lead to more expensive cheese availability. Which means more games decided early.

Adding double com leads to:
  • More metal that can be used to rush things
  • Halves relative cost of llt
  • Halves relative cost of radar
  • Halves relative cost of scouts

+0 / -0
quote:
On Red Comet you lost the game because you blindly chose the "rush raiders" strategy vs my "rush llt and eco" strategy. I might not have classified the game correctly, but this doesn't change the fact that this was yet another case of blind RPS.

I had scouted it when i had only 2 scorchers done. If i judged that your base was unassailable i could have switched approach (like i did after losing the scorchers) while keeping my mobile force to threaten your expansion and defend mine. The error that lost the game was not blind.

quote:
In Multiplayer B483545 2 on Obsidian_1.5 I didn't go for a scout because it would have severely weakened my strategy

You would have had 1 glaive less in your raiding squad if you had 1 glaive scouting.
+0 / -0
I'd suggest giving the featured map pool some stirring and nerfing skirmishers before doing anything more structural.
+3 / -0
Here's some stirring for you.

Feature:
  • Valles_Marineris_v2
  • La Isla Bonita v1.1
  • Iceland_v1
  • Desert Needle Small 3.2

Those are not only nice maps, they also historically made for the least rush games. Additionally they're viable for vehicle play, which comes short in the current pool.

Remove from pool:
  • Avalanche v3.1
  • Badlands 2.1

These are the two worst offenders for short games dictated by pre-game RPS.
+1 / -0

6 years ago
IMO, La isla bonita is very favored for the western player. Not the best choice.
I would also include CCR and Alien desert, inculta, archers valley. All classic 1v1 maps that dont see play anymore. SAD!

ZK plays better on larger maps in my opinion because you have time to react and utilize the strengths of different units in a non-cancerous way(skirm spam on small maps = cancer, its too easy to porc up and raiders become useless).
+0 / -0
I find Valles Marineris visually oppressive and tiresome while also quite porcy and center-dominated.

Desert needle has weird blocky terrain that's also aesthetically jarring to me, and is speedmetal.

The other two are alright. I enjoy the weirdness of Bonita's starts.

This is also reasonably fine since you can't just remove two, add four.
+0 / -0

6 years ago
So we should be able to measure the effect of fac matchup RPS by comparing a non-cherrypicked sample of mirror matches to non-mirror matches right?

How do we expect them to differ?
+0 / -0

6 years ago
quote:
IMO, La isla bonita is very favored for the western player.

How so? The map is symmetrical and starts are supposed to be randomized.
+0 / -0
I expect Fealthas means the clockwise or counter-clockwise player.
+0 / -0

6 years ago
It was not randomized before. The western player had easier access to the bottom hill and just more favorable terrain or something like that. Although the map is radially symmetrical, it is not symmetrical from perspective of the players.
+0 / -0

6 years ago
Boosting the starting economy would mostly lead to you being able to do cheesier all-in rushes. Boosting commander income might help though, as that comes in gradually, but also puts more eggs in that basket.

Lowering buildpower would also allow you to respond more reactively with your resources and make rushing harder, we did a global cut from 6 BP to 5 BP at one point. This was to slow down the possibility of a rush and make it easier to go early con without stalling. This took 4 BP out of the early game with a con start- 2 from the com, 1 from the factory, 1 from the con. We then immediately upped the factory buildpower to 10, adding 5 buildpower straight back into the game, totally destroying those gains. I was strongly against this and it means most players stall or have to put their factory on low priority during the early stages of the game- this is precisely when a rush hits.

You will note that Intersection has 4 safe mexes, 2 pocket mexes, 2 raidable but near mexes, then 3 mexes in the middle and 3 on each side. That's a total of 8 fairly safe mexes each (16 total), and 9 contested mexes split between both players. This I think is much better than the 3 starting mexes, and then clusters of contested 3 mexes.
+0 / -0

6 years ago
As a sidenote, I think this map would play better if the corner mex count near the geo was cut to 2, the central hills were cut to 2, and the center was increased to 3. It seems a little bizarre the hardest to hold(lowground and water slowing) area yields the smallest award.
+0 / -0


6 years ago
So CHrankAdminDeinFreund are you are effectively saying that you feel forced/encouraged to execute blind all-in rushes? Is there no decently solid play that fares well against such rushes? Double commander is super ugly and micro-intensive so it's not happening. Just to pin down an example, what do you think about doubling commander income?

Big maps (and vehicle maps) tend to have issues in 1v1 because they tend to have the same metal density of smaller maps, leading to a runaway expansion-style game and an economic victory by one side. However there are a few ok ones that could be added.

PLrankAdminSprung the map is symmetrical for 1v1 only in the sense that all maps are symmetrical, that is, that your start position is randomly assigned. The relationship between the start positions is not symmetrical.

AUrankAdminSaktoth I never figured out why you treat not stalling in the early game as a end goal and I don't think that was why BP was reduced.
+1 / -0

6 years ago
Is "runaway expansion style economic victory" a bad thing? Sounds like a healthy way for games to be decided. Small maps are dumb to play exactly for the reason that this sort of thing is not possible.

The downside is ofcourse, that this sort of thing tends to magnify skill differences.

On small crap maps which I've listed 10 times already, since you can no longer contest a significant amount of metal, the game becomes about efficiency above all. This is where skirm spam and porc comes in.

ZK community seems to have a huge mental block about the idea that a map can be bad or mess with balance and good gameplay. Starcraft2 has accepted a long time ago that maps have an impact on balance, and there is an accepted idea that some map features or designs are just BAD.
+2 / -0
Page of 2 (28 records)