[b21] Macro ballance - lowest networth challenge

[Cisz]

Ok, let's get this bay ballanced.

I tried a few times to no-leak wave 60 with different builds and the lowest possible networth.

For example:

My lowest 122222 was a all but light lv2 build, and I had one support each and two periodics only, for the cost of about 60000 gold. I played at 4-5 and did not leak.

My lowest 3332 (one dual support only) was 83800 gold. Can you do it lower?

If not, I assume that 6 ele has a 40% advantage in lategame over 4 ele with 1 support dual.

[DDRKirby(ISQ)]

great topic! (b^^)b

[Twilice]

You know, that is really not my style

I play with as high net worth as possible. Always

Gj btw.

[dvdlesher]

Hmm... This challenge will be hard for me also

I usually play exploiting as much interest as I can... But this challenge is the opposite of my playing style (and Kirby I guess)

[Cisz]

Jolin allready hinted that he has doubts about the 3332, so maybe he'll find a better sollution.

It's easy to try out. Use -gold and -lumber, start with xsm, but after summoning the elements for velocity, change to -level 1. Then summon what you need and build your defenses. Save game, and go to -level 58, so you can play wave 59 and 60. That's important, because smith and well will need 2 waves to power up all towers.

Now add or remove towers until you no leak 59 and 60, but don't cheat: Try to keep a mix of dmg that could win you a game.

I'll provide an example for this soon. While you wait, listen to some light music.

Edit: And here it is, 60k gold, no leak, 222221 (all but light on 2), and yes, i microed the trickery.

1676 Etd(4b21 vh pick) testing 222221 60k gold nl.w3g

[Cisz]

I did a second test, and compared a 33 to a 222 build, as jolin predicted that 33 > all. Afaik it's bad but not as bad as the other issue.

I tested 6 refined nature and 2 waterfalls (33) vs 4 fanatic and 1 forge (222). Notice that 2/6 is a worse ratio that the ideal 1/4 of the 222, so 222 has an advantage there. (I made sure all towers were boosted.) Notice also that the 33 towers cost about 19k, while the 222 towers cost about 21k gold. Again in fovour of the 222.

I tested both builds on wave 45, at 4, and only the 33 did a no leak, while the 222 leaked a bit. So I assume that the effective output of a 33 with maxed support is at least 12% better than that of a 222 with a support dual.

4 fanatic + 1 forge

16093 networth displayed with towers built => 21457.33 gold

wave 45

few leaks

6 refd nature + 2 waterfall

14334 networth displayed with towers built => 19112 gold

wave 45

no leak

=> Gold ratio is 1,1227, or +12% dmg/gold/time for the 33.

1694 Etd(4pb vh) 33 vs 222.w3g

[jolin012]

Let me put it this way.

The lowered dual boost from 240% to 200% was just wrong and pointless. it made clearer contrasts between t3-sup-dual-builds and non-t3-sup-dual-builds. It also hides the early t3 sup dual problem a bit, but does not remove the issue, because we will have to reestablish the support from sup duals t3, that was already perhaps a bit low (240%)

So - looking at the early dual problem now, will show a bit but not all of it, the contrasts will go stronger as sup duals of t3 will have to increase power. We need a solution such as the variable dual support strength.

Anyway, I have also been thinking about trying to test the balance, but it's hard to draw conclusions other than what we need to buff/nerf. it's hard to tell how much. because, the best way to do this is what cisz suggests, testing net-worths for builds. it's the best way, but it's not good, in fact it sucks. it's just that it's the only possible way in game to test how much to change. (the case would be different if there was a command for summoning a wave with in game chosen hp or armor by command).

Why does "check which gold can win" suck?

first of all, the lategame is too easy with at least the best builds. and for no reason the pures increased from 7k to 15k, so the cover all the gold in the best builds. and therefore, the 4ele can't show it's advantage over 5ele, having t3 damage dual (instead of extra t1 support), because it beats wave 60 w/o any other than pures and support.

So this is how easy it is to beat lv 60 noleak now: 10 lumber spent, 32k net(43k gold spent). 2 pure water, one monsoon, one supernova, one well tier 3.

Note That This Replay Does NOT Prove the Map balanced. It is a combination of several abused imbalances of the map and the way of testing that Makes 4ele Look Stronger than It is

Remember that it is also the cause of one more problem. slow and splash still pwns, as shown in the latest draft. I tried the same thing but with monsoon and supernova replaced by enchantment t2 and polar t2 and well3 for support, and pure nature x2 for damage. it leaked two thirds at area 4 and one third at area 3. both because it doesn't have slow and splash that the other build has, but also because the ST-support problem was never solved. some claim that one is enough, of polar and enchantment, but it's 2 needed. you can't enchantment and polar cast one every other creep and hit those very creeps with the pure natures. it would perhaps be possible but even i fail, it's too hard to see which one is struck by polar, and polarstruck is in fact very easy to see.

Edit: the replay doesn't seem to work, at least not for me, but the text above explains it all anyway

To me, the slow and splash is a great problem that I dunno how to solve... a problem that causes other problems, it currently makes 4ele look really strong with it's 43k gold to beat wave 60 no leak with all dmg on pure water.

I dunno what we should do about the slow& splash problem. mech can cause a few leaks, but those are solved by not 50% ST damage, but by 20% STLongrange damage, or accepting a few leaks, also bosses do't prevent much slow% splash abuse, other than the same as mech. Perhaps, if Mechanicals with the buff on both remove slow and ignore slow? or is that llready how it is? Perhaps make them have no collision-size when buffed?(like I saw that on phase boots in dota I think)

Beta21 NNNWWWFFLLx 32k.w3g

LastReplay.w3g