Forged Alliance Forever

Re: 3619 focus.

2012-08-24T09:49:31+02:00

More feedback about the Tempest. Yesterday, I realized in a game that the biggest problem of this Experimental is the Torrent Missile ship. Similar stats and use, but the Torrent's missiles are twice more efficient.

Tempest: range = 150 / Damage = 6000 every 10 seconds / AOE = 5 / Inaccurate / lead target
Torrent : range = 200 / Damage = 10000 every 20 seconds / AOE = 5 / Pinpoint accurate / doesn't lead target
Torrents shoot 10 missiles at a time, meaning they can wear off shields much better than a single shot (shields can't recharge while the missiles hit). TMD cannot protect against several Torrents, because each missile has 2 HP. 10 missiles require 20 TMD to be completely nullified. Do the math for several Torrents.

Torrent ships cost 8000 mass only. 3 of them are as efficient as a 63k mass Scathis (I am not exagerating). They need a nerf to their AOE, so that they have a different role from the Tempest. They can't have the same AOE, better range, and almost the same DPS for a fraction of the price. Their AOE should be no more than 1.5 or 2, so that you can snipe surrounded mexes and t3 pgen, but not blow up 1/4 of a base. Big AOE damage should be the role of the Tempest.

And for the incredible range the Torrent has, each missile should deal less damages (around 700-800 damages rather than 1000). 8000 damages every 20 seconds at 200 range are plenty enough to annoy any player. Or alternatively, each missile could have only 1 hp, so that it's actually possible to defend a little against several of them with lots of TMD.

If the Torrent weapon is not modified, the Tempest will never be interesting. The Tempest's role should be clear : big AOE, and frontloaded damage; Torrent's : crazy number of missiles, but small AOE.

To sum up, I would advise to adjust these two units relatively to each other, for instance like this :
Tempest:
- Health and Max Health = 60 000 HP
- Mass cost = 24 000; Buildtime = 14400
- Damage per shot = 8000 (kills any destroyer and depletes a shield boat in one shot, like the UEF Battleship does)
- Muzzle Velocity = 35 (from 30)
- Firing Randomness = 0.2 (from 0.3)
(AOE and rate of fire stay the same)

Torrent Missile Ship:
- AOE = 1.5 (from 5)
- Damage per missile = 800 (from 1000) / or Missile Health = 1 HP instead of 2

Ze_Pilot, I attached here modified bp, in case you agree at least partially with my analysis, along with the modified t3 subs energy cost reduction (80 000 energy instead of 800 000). I will be away the next 3 days, that's why I'm preparing this now before I forget all my brainstorming, in case you update the 3619 balance test mod this week end.

Statistics: Posted by pip — 24 Aug 2012, 09:49

Re: 3619 focus.

2012-08-23T08:24:47+02:00

Additional feedback to the new Tempest:

Faster muzzle velocity means that inaccuracy is greatly amplified. In that case, 0.3 firing randomness has not been changed, so the weapon is almost 50% more inaccurate than before (it will shoot much more often behind or next to the target, which is a problem for a 10 seconds reload weapon). The highest the weapon velocity is, the more inacurrate the weapon becomes if it has a firing randomness superior to zero.

In other words, if the point was to make the Tempest more accurate, as Lu Xun wrote, what happens is the opposite. The value that needs to be changed first is the firing randomness (like 0.2, or rather 0.15 instead of 0.3 if the muzzle velocity is increased to 40).
Also higer muzzle velocity will mean the trajectory of the shoot will be more straight rather than arced. It can, in some rare cases, make the weapon shoot into cliffs rather than over them when firing at shores, but it's map dependant, and i don't think it affects the official maps (40 is still low enough).

Statistics: Posted by pip — 23 Aug 2012, 08:24

Re: 3619 focus.

2012-08-22T22:08:38+02:00

I just tested the new Tempest in the new 3619 test :
- Max health was not changed, so its health is 50000 / 35000.
- DPS remained the same (changelog says +30% DPS but it's not changed, so it's still one shot every 10 seconds).
If you want + 30 % DPS, you need to set rate of fire to 0.13 (from 0.1) = 780 DPS. I'd say 0.15 rate of fire would be more appropriate (= 900 DPS), considering the huge cost compared to Battleships (especially UEF battleship, that has similar range and 450 DPS).
- Mass cost is back, but buildtime is still 25200 (quite long for an experimental, you may want to align the buildtime to the same as the Atlantis (14400), for consistency.

I would personnaly prefer an OC of 4 seconds but i guess it's just personal .

Statistics: Posted by pip — 22 Aug 2012, 22:08

Re: 3619 focus.

2012-08-22T20:56:52+02:00

Ok, new balance test mod online. Same speed for the t3 land (I feel like it still need more datas), but most of your propositions are in (except the fire priority, don't have time to look into it now).

Statistics: Posted by Ze_PilOt — 22 Aug 2012, 20:56

Re: 3619 focus.

2012-08-22T20:52:40+02:00

replay 1 is not a mistake, it shows you a timing push just before the ennemi's T3 land.
Hehe i like how you have an explication for each replay ZeP

Statistics: Posted by Lu_Xun_17 — 22 Aug 2012, 20:52

Re: 3619 focus.

2012-08-22T20:39:54+02:00

Dismissed replay 1, not T3 land

In replay 2, you are stalling in mass. The buildtime are irrelevant here, as you build slower than the normal new BT

Same in replay 3, but zock was not stalling. But he had lot of t1 engys that would be used to spam T3 more quickly.

I will look at replay 4 later, but I'm still not convinced. Maybe T3 land is too easy to get, but I don't think the BT are a problem here.
Maybe indeed T4 BT must be tweaked..

Statistics: Posted by Ze_PilOt — 22 Aug 2012, 20:39

Re: 3619 focus.

2012-08-22T20:38:13+02:00

I watched all games between Lu Xun and Zock. And I thank them very much for really testing the balance testing patch. They provide very interesting info, and that's what this dedicated balance testing for 3619 is all about.

My opinion is more mitigated now about t3 land units 20% buildtime reduction. I don't think the buildtimes reductions are a bad thing for t3 land per se, it does, however, bring an additional reason to skip t2 land alltogether. Balance between t3 and T4 is definitely better IMO, but between t2 land units and t3 ones, it's a bit worse.

This leads me to think that :
1) it may be better to adjust t4 buildtimes rather than t3 ones.
2) something needs to be done to make t2 units more interesting (t2 structures are of course very good)

I think that the speed nerf that t2 units recieved in 3603 was uncalled for. It makes raiding with t2 units very slow. I don't see why a Rhino should have an inferior speed to a medusa. 3.3 speed was very nice for Rhino and Pillar, 3 speed for Obsidian, and 2.9 for Ilshavoh were nice too. Now, they are really slow, by the time they arrive at a base, your opponent can set up t2 pd, or even tech to t3 and streamroll you.

Zock is brooding some more precise thinking about t2 land units, I think it's important.

Finally, that Open Palm game was freaking epic guys, than you so much for the show!

Statistics: Posted by pip — 22 Aug 2012, 20:38

Re: 3619 focus.

2012-08-22T20:21:24+02:00

45min intense Open Palms vs Zock :
(couldn't upload 4 replays in 1 post, sorry for double post)

Statistics: Posted by Lu_Xun_17 — 22 Aug 2012, 20:21

Re: 3619 focus.

2012-08-22T20:20:21+02:00

Here are the 1vs1 i made with zock this afternoon, focusing on T3 land.

My opinion hasn't changed. Even if it's always super fun to play some T3 land wars, with this -20% buildtime, T1 and T2 are even more squized.

Here are the replays: (game 2 is quite epic, but the OP in my next post is even more cool )

Statistics: Posted by Lu_Xun_17 — 22 Aug 2012, 20:20

Re: 3619 focus.

2012-08-22T17:05:34+02:00

Firebases wars isn't what happens in every game- there is always a way to avoid it. The reason it works however is also because of how maps like confrontation have like 1 chokepoint. Of course this will favour pd creeps. What you're saying is like playing a thermo and then complaining theres too much ecoing, or playing a setons and complaining that you end up with too much asfs.

Statistics: Posted by Softly — 22 Aug 2012, 17:05

Re: 3619 focus.

2012-08-22T16:55:25+02:00

PIP = The game is going to a firebase war frequently with this nerf ACU.
Missile lauchers is sooo easy to hold, spent a lot of time to build, and with 2 or 3 Shields t2, u can hold a lot of them.
Bombers t2 are soo expensive and goes down with a couple of AAT2.

I think everyone here is looking to 1x1 games... But a team game is going boring.
Firebases war, Experimental Racing, no attacks after 10-12 minutes (they awaits t3 or experimentals), many players on the main base.
Is no more advantage to send your ACU to battlefield.
Is no more advantage and dangerous to use ACU in other side of the map, because they don't use, i don't use, now it's better to make a turtle game on the main base.

3 seconds to OVERCHAGE is more than perfect to avoid firebases WAR and turtle gammers, RTS must be a RISK GAME.
Good players know how to defend and make a inpenetrable main base with a couple of shields and some PDT2.

Wanna nerf RAMBO ACU???
Let's change another things, not the time to use overcharge, he is the MAIN WEAPON to kill firebases, to make some people to back down. even a PDT1 it's hard to kill now.

OVERCHARGE COST: 15000 E or 18000 E.
(it's avoid RAMBO ACU on the start of the game)

Still 5 seconds to make overcharge? Sooo let's see these compensations!
OVERCHARGE DAMAGE TO FIXED UNITS: 1200
OVERCHARGE DAMAGE TO SHIELDS: 5000
OVERCHARGE DAMAGE TO ANOTHER ACU: 850
OVERCHARGE DAMAGE AREA: BIG AREA

Imagine the situation...
It's a 3x3 game on the "confrontation"
XUCRUTIS and PERNILONGO planned to make firebases in front of TRIPASECA, they upgrade just t2 each one.

TRIPASECA TEAMS will spent a lot of time to came with their ACUS to help.
Units T1 and T2 don't break a FIREBASE easily.
Even with T2 ACU, GUN, and a couple of units TRIPASECA can't break a small UEF firebase with 4 PDT2, 2 PDT1 and 2-3 shields t2.

Statistics: Posted by Destructor — 22 Aug 2012, 16:55

Re: 3619 focus.

2012-08-22T15:33:09+02:00

Lu_Xun_17 wrote:
-shild 3 sec to wait before recharge
- civilian focused bug fixed
-T4 get a priority focus fire on other T4

agreed

Lu_Xun_17 wrote:
- soul reaper 34000 mass
-Novax 28000 mass
-tempest back to 28000 mass
-tempest speed of shoot faster (would make it more precize and buffs dps +30%)
-tempst hp 50000 hp (there we would get a T4)

this all needs to be tested, but i basically agree to these changes

SR is too powerful atm, having air superiority

Tempest should indeed be more expensive (it's a strong t4 unit) but therefore be more powerful, although Lux' changes also look like random numbers, it might be the better deal

Novax has too high dps for it's invincible, it can attack everywhere and also break shields now, also useful for scouting or killing units like a fatboy which have ''low'' hp but high range
i suggest to let the costs stay as they are (else it's never used) but reduce damage by 25% (to 150 dps)

Lu_Xun_17 wrote:
-OC reload time back to 3.3sec

i also think 3 to 3.5 seconds would be best, reason:
in early game, you can't spam OC anyway, because you don't have about 1.5k e/sec to spam it each 3 to 4 seconds anyway
(not a big threat against t2 land) and later, when fighting t3 or t4, it's fine because the acu should be able to defend a bit against these units

Statistics: Posted by Myxir — 22 Aug 2012, 15:33

Re: 3619 focus.

2012-08-22T15:01:22+02:00

Lu_Xun_17 wrote:
-SACU + graphics+ sera colormod sounds good
-shild 3 sec to wait before recharge
-stealth fixed
- civilian focused bug fixed
- soul reaper 34000 mass
-Novax 28000 mass
-tempest back to 28000 mass
-tempest speed of shoot faster (would make it more precize and buffs dps +30%)
-tempst hp 50000 hp (there we would get a T4)
-OC reload time back to 3.3sec
-T4 get a priority focus fire on other T4
-ACU vet : 20-40-65-90-120

ACU vet will be fixed.

SR will cost more. Probably not as much as before. Need tests.

OC, tempest, novax, SCUs and T3 land need more test.

Problem is, if I don't get more replays during the test phase, when the deadline for 3619, it will be pushed in his current state, for better or worst. So you know what to do...

Statistics: Posted by Ze_PilOt — 22 Aug 2012, 15:01

Re: 3619 focus.

2012-08-22T14:53:55+02:00

About T3 land buildtime :
Your tests in parangon conditions only show that the T3 land buildtime is well decreased by 20% in 3619 ^^

What we need is ingame conditions. Anyway you have your point of view about this change, i have mine and it looks like it won't evolve on any side.
I really think this change is a mistake, a decline for the gameplay. T3 land is very well measured currently.

I think all the last "tries" to balance FA got made with too impulsive and random numbers. (Soul Reaper, Novax and Tempest are absolutly overpowered now. Well too cheap at least.).

There are over things to improve, which are probably more important (for exemple a fatboy still chooses to focus a T1 engi instead of a T4... or often some units are locked when leaving a T3 factory (don't tell me T3 land build time can be a solution to this problem lol)).

Here is the changelog of my dreams for 3619 : (of course everyone will have a different one)

-SACU + graphics+ sera colormod sounds good
-shild 3 sec to wait before recharge
-stealth fixed
- civilian focused bug fixed
- soul reaper 34000 mass
-Novax 28000 mass
-tempest back to 28000 mass
-tempest speed of shoot faster (would make it more precize and buffs dps +30%)
-tempst hp 50000 hp (there we would get a T4)
-OC reload time back to 3.3sec
-T4 get a priority focus fire on other T4
-ACU vet : 20-40-65-90-120
-grace period when losing radar coverage

Statistics: Posted by Lu_Xun_17 — 22 Aug 2012, 14:53

Re: 3619 focus.

2012-08-22T12:50:21+02:00

@Lu Xun: I ran some tests to check if the 20 % buildtime reductions were that big a difference.

A parangon was spawned, so only buildtimes, in ideal conditions, are measured here.

- A t3 land factory assited by 20 t1 engies in 3619 produces in 3 minutes (of game time, not real time):
8 Harbinger or Othuums (7 in 3618) / 12 titans or loyalists (10 in 3618) / 6.3 Bricks, Percivals, artilleries and snipers (5.5 in 3618)

- A t3 land factory assited by 30 t1 engies in 3619 produces in 3 minutes (of game time, not real time):
10.3 harbingers and Othuums (8.8 in 3618) / 15.2 titans or loyalists (13 in 3618) / 8.1 Bricks, Percivals, artilleries and snipers (7 in 3618)

Without Parangons, of course these results will be less, because your income fluctuates during a 3 minutes time span.
You can see that the number of additinal units produced is not extreme. If you compare the 2 scenarii like this : 3619 buildtimes with 20 engies and 3618 buildtimes with 30 engies, you will see that the number or units produced is quite similar : around 8-9 / 12-13 / 6-7 for the different units.

So, in order to really make a real difference, you need to have plenty of mass, 30+ engineers otherwise your production will not be hastened significantly.
What I want to show is that you can roughly build the same number of t3 units with around 8 less t1 engineers than before. You cannot reasonably go beyond 30 engies assisting anyway, because of pathfinding problems making units bumping into one another, and nullifying the buildtime reductions anyway.

In any case, please provide replays where it's obvious t3 land becomes clearly messed up with 20% less buildtime because I fail to see how 1 or 2 more unit every 3 minutes radically screw the gameplay. However, one more Percival, or 2 more Harbingers when a Monkeylord is on is way is still nice to have.

EDIT: this test also confirmed that SCU buildtimes are set too high in 3619 because you can produce a lot of t3 units that don't need to be upgraded during 3 minutes and maybe only 1 SCU + 1 upgrade, so they are probably not worht it currently (see my previous post about them).

Statistics: Posted by pip — 22 Aug 2012, 12:50