Season 7 Config Discussion

Announcement
1

This post is a work in progress and will be released on monday

The season 7 config has been released. This post discusses the philosophy behind how we came up with it, and how terminal should feel on this config.

Changes
Encryptor:

  • Removed

New Unit: Factory

  • 30 health
  • 4 cost
  • Generates 1 MP per turn
  • Upgraded: Generates 1 SP per turn as well
  • This unit is the centerpiece of the new season. All subsequent changes are based around it

EMP:
With more ways to generate bits, it is possible to reach the critical number of EMPs needed to destroy an entire base much earlier. This keeps them from going to crazy

  • Renamed Demolisher
  • Damage reduced from 8 to 6

Filter:
The health of structures has increased across the board to make them better able to stand against larger groups of both pings and EMPs

  • Renamed Wall
  • Health increased from 60 to 75
  • Upgraded health increased from 120 to 150

Destructor:

  • Renamed Turret
  • Health increased from 75 to 90
  • Damage cosmetically adjusted from 16 to 15. This makes no difference with encryptors out of the game, and is just a cleaner number.

Scrambler:
While unchanged, Scramblers will be much more useful as there will be many more units for them to intercept, sending them out will be less punishing as you will have more bits to work with, and saving resources by skimping on turrets will let you invest more into factories. This season is hopefully the scramblers chance to become the intercepting all star it was always meant to be

  • Renamed Scout and Interceptor

Ping:
The ping is unchanged, but should the king of closing out games when the enemies defences are crippled

  • Renamed scout

General resources
With units relying on sheer numbers rather than encryption, players take more damage. Players have slightly more health to compensate. The number of starting cores has been reduced so choosing to skimp on your initial defences to build factories is much harder.

  • Player health increased from 30 to 40
  • Starting cores reduced from 40 to 20

Philosophy
When making a new config, we have to weigh the size of the change with the cost to make sure it feels ā€˜balancedā€™. If the changes are too small, the new season is not fresh and exciting. If the changes are too large, balance can become a problem. Because both players have the same resources, a balance situation doesnā€™t manifest as ā€˜unfairnessā€™ as you are not pitted against powerful cards, weapons, or characters as in other games. The real balance issues, as we have discussed in the past, manifest as lack of diversity in strategy or in extreme snowballing, where a small win can have massive consequences. As well as other negative things like that.

In the past, we have leaned towards focusing on the ā€˜pristineā€™ balance and strategical diversity we have enjoyed through most of its lifetime, and have not taken much risk that might jeopardize our design goals. However, in recent seasons, we feel the config has not changed enough and the game has felt too stale.

Other games embrace the far-opposite side of this spectrum. When the new ā€˜Demon Hunterā€™ class was released for the card game hearthstone, a dozen or so of their cards needed to be nerfed in the first month. League of legends notoriously consistently releases champions in an over powered state, and reduces their power over time.

As part of the design philosophy of this season, we will be trying out this approach, and seeing how it feels for terminal.

Season Focus
This season should feel focused around the new unit, the Factory. This unit replaces the Encryptor, and fulfills a similar role: Allowing you to invest SP to increase your attack power by generating MP. Shielding is out for this season, but with a large number of factories, you can generate enough bits to demolish your opponents frontline, turn after turn. Interceptors (Formerly Scramblers) may be the only way to defend against these massive onslaughts in the mid-late game, and they will be less punishing to send out since you should have more MP at your disposal.

Without encryptors, attacking units will feel weak in the early game, and can be used to try to chip away at opponents walls, earning small advantages. Both players will try to hold off these attacks with as few defences as possible while trying to build up their factories, enabling larger and larger attacks faster than ever before. With enough resources, players will be able to send huge waves of pings to charge through enemy defences, or EMPS to destroy their opponent completely.

Ideally, while we expect most strategies to use factories, we hope to see aggressive strategies that focus on non-upgraded factories to quickly send out many units, and control strategies that build up huge bases with upgraded factories that help build more upgraded factories.

Risks
Our main concern is the durability of the corners, which is something we will consider patching sooner rather than later, likely by increasing destructor cost and strength. Besides this, there are a dozen different ways this config can ā€˜go wrongā€™ and create a low-diversity meta. We will keep an eye on things and are planning a rebalancing around mid October, if it seems like it will be needed. Ideally, it will be as small a change as possible to restore diversity if some single strategy seems uncounterable.

Some specific things we are looking out for:

  • Snowballing: Do early wins have a disproportionate impact on the rest of the game?
  • Diversity: Are there multiple reasonable base configurations and strategies? Does each one have strengths and weaknesses? Are games all lasting basically the same amount of time?
  • Units: Does every unit have a clear use?
6 Likes
2
3

does factory have range anymore or is that just unfixed, also it still gives encryption

1 Like
4

What an interesting idea. I have no idea how this will play out, but we are going to see a very new mindset with this Iā€™m sure.

The classical strategy game concept is attack/defend/economy where defend beats attack, economy beats defend, and attack beats economy. The trick is to balance your game out so you pick the winning idea. Of course it is not black and white so algos who can balance this out could be quite successful.

I do think there will be a major snowball effect and mostly onesided battles, but that might not be too bad. The concern I have is that algos play rock/paper/scissors on their first turn and the winner of the first turn wins the game. If that is not a problem, this new change could be successful.

2 Likes
5

@RegularRyan, I think the range and the encryption still needs to be removed from the config of the factory.

2 Likes
6

From my first few tests I can say that Structure Points earned from scoring seem a bit overpowered ā€¦
Previously this point were a balance for a more aggressive play style, where you donā€™t build much, and hope to score and then reinforce your position.

But now a heavy economics build, gets rewarded with extra MP (and SP) and then when they get high enough MP, scoring gives them EVEN MORE SP ā€¦

So if you just spend first 4-5 turn on ECO and manage to defend with Interceptors, it is an easy snowball game.

it is really hard to destroy Factories for the first few terns, if they are protected b y 2-4 interceptors
And the alternative just to free scoring 5-20 points ā€¦ might not be enough to counter 10 upgraded factories on turn 5.

So may be a good point for balance will be to revisit the SP reward from scoring. In the current config, it is not really needed.

Prove coming soon :slight_smile:

1 Like
7

From playing by hand against the bosses, the factories certainly seem very strong, but itā€™s hard to tell since they also work as encryptors which makes them ridiculously overpowered.

8

I made a new algo (The_Stick_1) that builds only factories, walls, and pings. Currently it is 28-1 with no game lasting longer than 7 turns, since there is not really any way to stop a 35-ping stack on turn 6, regardless of whether the pings are encrypted.

My current opinion is that the factories are quite overpowered, and Iā€™m curious if anyone will come up with a strong enough counter to The Stick to make games interesting with the current parameters. I suspect that factories will have to be nerfed to offer a worse ROI (either higher cost or lower generation of resources).

That said, I think it will be interesting to leave the config as is for a week or two just to see what sort of meta develops.

Example game: https://terminal.c1games.com/watchLive/6965957

5 Likes
9

I have a suggestion (maybe for the far future)
The return of the factory is proportional to the y value (just like the upgraded encryptors previously) this incentivices risky factory placemeny for more resources. For example: 0.1*y per turn.

2 Likes
10

This is something we are considering for future seasons. For now, we are expecting them to be placed in the back and are fine with this.

We will consider this. Im curious about what you guys come up with in the first week or so.

Ill make sure this is fixed asap, definitely unfortunate that this snuck through. Probably made a mistake when copying over the finalized list of changes to the new main config and didnā€™t notice during final testing, or something similar to that where it snuck in during the finalization steps.

Similar to my response to demorf, lets give it a few more days before I start bumping the cost up

11

The Stick vs Glass Factory:

Mass factories with strong pushes after 4-5 turns vs
Upgraded factories, with constant simple attacks,
https://terminal.c1games.com/watchLive/6966220

This is the baseline for now, simple strategies that crush everything else.
Can some figure out an actual GOOD algo please ? :smiley:

12

Those heavily scoring ping waves could have been stopped by just a few scramblers, really curious how scrambler focused defences work out. Also, is this a fresh replay? the encryptor issue should have been fixed this afternoon, let me know if you are still seeing the encryption happening in matchmade games

13

What about a scrambler focused offense? :thinking:
I threw together a quick algo called Scrambler that just spams factories and scramblers, designed to counter stick which it wins against on turn 4: https://terminal.c1games.com/watchLive/6966459

It still loses against a few algos from last year once they can stop the scramblers, it obviously needs to transition to something else after they stop working, but it was able to reach #2 on the leaderboard. Mattxai started off with a stick copy, but took my scrambler spam and added a couple destructors to win the mirror and other minor optimizations which got him to #1 as of today: https://terminal.c1games.com/watchLive/6966977

Most people are still in the testing stage which is why there are so many dominant simple strategies. I think the best strategies will be more complex, but I want to see what simple strategies work well before putting in the time to make something complicated. Itā€™s only two days into the season so no one has had time to create a complex algo that is optimized for season 7.

The encryptor issue is fixed but the old replays have it.

2 Likes
14

I may have to take back that statement about the best strategies being complex unless someone can find a way to beat factory spam + scrambler death ball. This is what peak terminal looks like, folks.

peak_terminal
peak_terminal1301Ɨ1265 141 KB

Part of the problem is the reduction of starting cores, since you canā€™t set up an effective defense against interceptors until they murder you (not to mention falling far behind on econ).

2 Likes
15

I am a bit concerned about attacking units.
Early game, Interceptors can easily protect vital structures, and Scouts and Demolishers are rather useless.
Only death ball of Interceptors or lucky Scouts can SCORE a bit ā€¦ but not really destroy anything.
And this get countered later, with minimal static defence.

Mid game, even 5-7 Demolishers look pathetic ā€¦ they are 4 time nerffed:

  • lower damage (-25%)
  • structure have higher health (+20/25%).
  • (hits to destroy changed for Wall from 8 to 13 ā€¦ and for Turret from 10 to 15. Around 33%)
  • cheep lvl 1 Turret that now can actually hurt them
  • they are on the clock to do something, before the carpet of Interceptors mops them out.

Late game,
Structure damage is not relevant any more, as long as the factories are alive.
10 Demolishers somehow kill your left flank worth 60 cores, ? no problem ā€¦ just put double carpet of Interceptors, and rebuild it in 2 turns ā€¦

I really think Demolishers got nerfed too much.
I was counting on them to be able to destroy factories ā€¦ but this is almost impossible, with the ā€œcheepā€ Interceptors, and the obvious thing to protect.

Also now you have good incentive to stay passive, and just defend and ECO.
even if you donā€™t do structure damage to your opponent, you are getting stronger.
Factory protection is simple and efficient.
To counter 2-4 Interceptors, you need to attack them with more Interceptors FIRST, and then to hit the buildings behind them with the Demolisher/Scounts ā€¦ that are 2-4 times fasterā€¦ so you will need advanced maze ā€¦ to counter 2 interceptors ?!

2 Likes
16

Im going to setup the auto-competitions so that the first one plays sunday night so I can start looking into making a balance update sooner rather than later. Lets see how everyone stacks up on sunday night and ill put out an initial balance patch early next week

5 Likes
17

And just to clarify, I just share my impression so far ā€¦
Probably there are simple solutions for some of this ā€œproblemsā€ that I donā€™t see yet.
By laying it out here, I hope people will confirm same concerns, or prove me wrong,
and we can isolate any actual problems

3 Likes
18

From what iā€™ve tried, pure defense does not seem to be the play. The SP you get from damaging the opponent is crucial to winning the factory race. I think there is still a lot more room in the current meta for an offensive play, it seems like most people on the leaderboard are doing a similar strategy right now which is mostly defensive. I agree though that a Demolisher buff is needed, its very hard to beat a factory protected by interceptors now.

3 Likes
19

It seems like interceptors are the best units for both defense and offense right now. Theyā€™re the best for defense since you donā€™t waste SP you could be pouring into more factories, and theyā€™re the best offensive unit too since everything else dies to opponent interceptors without much of a fight.

2 Likes
20

Agree that with nobody building many turrets they are running rampant. What happens if some turrets are built? Are the ineffective? Do you end up too far behind because you didnā€™t put all your resources into more factories?