Release 2022.1.5

Entelect Challenge 2022 Announcements
1

Our main focus for this update was on tweaking configs for balancing. As well adding more details to the gameStateDto for your bots to have all the information they need.

  • added a max resource cap for each tier. This sets a limit on how much resources your bot can hoard.
  • increased the tick rate to 250ms. To give bots more time to think
  • added extra tiers
  • made changed to the regen rate of farm, wood and stone resources to even out the odds

Bots now have access to

  • the full game engine config
  • player actions - this includes all the information you need about your actions; the type of action, when your units were sent out, when they will return

  • this is meant as an enhancement to the farmingUnits, miningUnits, lumberingUnits, etc

  • Optimized logging!! you’ve asked for it and we have delivered. We have reduced the number of redundant code in the logger substantially so you can follow the game through the logs. We also understand that some of you have created visualizes so this is an additional GameStateCondencedLog that can be toggled off in the appsettings in the gamelogger project “CondencedLoggingToggle”: “false”.

  • Fixed various bugs with population calculations

  • View the appsettings.json for the updated configs

2 Likes
2

For the lazier amongst us :slight_smile:, will the starter pack be updated?

3

Good morning @PhoneticallyUnstable :wave:

Sorry we haven’t made plans to update the started pack :pray:

4

Morning :slight_smile:

All good. :slight_smile: Thanks for the feedback.

5

Haha, there goes my hording strategy…

Back to the drawing board I guess?

@Jenique , just to make sure. The source is v1.5, but the Starterpack will remain v1.3 but the tournament will be run on v1.5?

1 Like
6

I was having a bit of fun with my bot as well :slightly_smiling_face:

But yeah, once you have the processing down, its all the same.

7

Just my thoughts,

We all participate under the same rules, so its likely irrelevant.
But it could be cool if the first tournament runs on V1.3.

The first event was extended by 1 week,
The below scenarios make this complex:

A) Some bots need time to train and diagnose.
B) Some users might have completed their entries for the first event and might not have the available time to change it in time.

**Some bots like mine could get away with some minor tweaks, but in some cases this warrants a complete rewrite.

Knowing that this is the environment for the First Tournament would be good.
Then changes can be made.

Also a small thought (and i admit Im in this group) given the time constraints, I believe getting a starterpack up on github for this could be very nice if anyone can build one. It could help save us a bit of time with ide and dependency issues that we might face,

Just some thoughts, I have not looked much deeper into my own bots,
But will likely need to rebuild a lot of my logic to account for data changes.

**No longer need to track my units manually if the engine does it for me, for example.

Still fun though.

Part of me just feels that introducing these changes into the second event could be worth considering.
Probably saying that cause I just finalized my bot for engine V1.3 and im so happy with it. :rofl:

And I have never been completely happy with an entry before.

8

Yeah the same here.

I was ready to rock and roll but now with these changes it will need a mayor re-strat and rewrite and with only a week to do if, the pressure is going to be on.

9

I like the changes, thank you for the hard work!

10

There goes my 1 million score … …

11

I think 1 Million score will still be viable after engine changes.

it will just make res last a little longer,
Depends on a few things though

12

Jip, ill probably compile the v5 engin tomorrow and see what it does, and if needed post a compiled version for others that dont have knowledge of publishing c#

13

https://drive.google.com/file/d/1OE2-eS29wlaZpBJl7LN-XYEbQkAQShJC/view?usp=sharing

here is a compiled version of the game engine, logger and runner for version 5

1 Like
14

Thank you for this one.

I will test it out early tomorrow.

At least in theory if you wrote Heuristics, you should be able to make minor changes to your bot.
They did however mention that the JSON data is much smaller, so some things might be broken.

Will only know when I can delve a little.

Running my bot in engine V5 gives me 20 points for the 2500 ticks so I have some reparations. Reference bots fine though,

15

After some adjustments to my bot and running a few tests I do feel the changes were great.
Part of me does feel that the population growth rate is a little slow for the size of the map / amount of res.

Though this call is made with a older version of my bot. And im still adjusting.

There are things thats kind of bothering me with this version…

In version 1.3 there was some cool aspects where A) the best bot had resource advantage, B) You had a great blend of population play VS stockpiling.

In version 1.5 theres so much resources that its a race to see who grows the fastest only.
Now im okay with that. but the problem for me lies with the margin of error.

We have much more units than we need for upkeep.
So Bot A can be say 70% efficient compared to Bot B and still have a large enough buffer to not get messed around when it comes to population.

My concern is that the margin for error is way too large here. As in Version 1.3 you had to be precise, a 10% deviation meant 100K - 200K off your endgame score.

Im only mentioning this so that we leave enough room within the engine to give bots room to flex.
Again, this is on an earlier bot of mine, its not nearly tweaked for the current environment. For all I know I can hit a large enough population that I do end in my opponent’s territory.

I know for event 2 things will change drastically.
Im having a lot of fun though.

Just feel Version 1,3 had a lot more depth than Version 1,5 because of the population starving out scenario.
It meant people had to think about that.

This is my largest concern with a the new engine engine. Along with:
"is the storages large enough to keep up with max population growth, or will I end with units that I cannot use for anything, adding to my concern of a too large “Margin of error”

I feel the engine needs a dynamic of scarcity, it needs that looming: oh man, foods almost up.
I essentially had to dumb down my bot for version 1.5.

Im only saying this because it could help with balance.
**I have looked at a lot of number balancing the last while. I’m super sensitive to numerical imbalances at this stage of my life.
Just some thoughts

1 Like
16

Hey everyone, the starter pack has been updated. I apologies I forgot to publish the release.

I do thank each of you for your commitment and feedback. Rest assured we are trying to make the game as fair and competitive for everyone. We felt the changes were a step in the right direction and there will definitely be balancing changes in the future (after tournament 1).

This year has been difficult for the team as this game is far more complex in terms of balancing. So we want to say thank you to our awesome community.

We will be running the Tournament with 2022.1.5 so please make sure you have updated your bot accordingly.

We also promise to not make changes so close to the tournament date.

Happy coding and good luck everyone

1 Like
17

No problems,

We could easily assess the results of the first tournament to pick up possible problems.

I can personally be quite paranoid when it comes to these things.

Its great stuff.

18

Just something to take into account playing a perfect game in terms of hitting max population growth each cycle you can only achieve a max population of 432k in a 2500 tick game.

So its impossible to reach population tier 7

19

Hey there, Sorry to bother,

Just want to follow through. My earlier comment about population not going into enemy territory is nonsense. I just had some limitations left over from my old bot., In terms of scaling the endgame looks more or less like it did in the last engine after the population died out around round 400 or 500 or so. The only difference is it now takes the full 2500 ticks to get there.

This is amazing for certain

So in terms of population I feel we are good.

And about my concerns about the storage and capping it. Early Game it is a little troublesome and it does feel somewhat restrictive. But lategame theres not really enough res to sustain a 4V4 match.

**Resource Scarcity does play a role unlike I initially thought.

So based on this I must back down and say, this looks much better than the last version and Im really impressed by the numbers. I believe if this is pared by some “Warfare” **Maybe you can build an army, guard nodes and any units thats not your own gets killed (giving you a point boost) or something could be interesting. Then “having some expendible units” could mean that theres not even issues with storage early game.

I just didnt see the bigger picture and wasnt near the endgame of my original bot. There might still be some things Im missing though. But still feel good, and I didnt need to change too much at least. So was paranoid for nothing.

Still testing but my entry is about done.

Just have 1 or two minor tweaks left.

I am afteral still optimizing on my end to account for the limitations.

2 Likes
20

Hi we are aware that the latest tiers are not reachable and this is intended.