Page 7 of 9

Re: Discussion about game

PostPosted: February 12th, 2014, 3:35 am
by trurle
Igor Kulakov wrote:I see discussion immersed in excessive thinness. The problem of perfectionism. I see the decision to increase the price of heavy and sophisticated weapons and the speed limit for the production of these types of weapons (not increase by 1 every day, such as 1 every 2 days or 1 every 4 days)


If you prefer simple solution, fit to reality ramp-up will be following
1 unit/turn cost<10
1/2 unit/turn 10<cost<40
1/4 unit/turn 40<cost<160
1/8 unit/turn cost>160

In this case, weird amounts of ultra-heavy weapons will disappear.

Re: Discussion about game

PostPosted: February 13th, 2014, 1:53 am
by Khanti
Igor Kulakov wrote:I see discussion immersed in excessive thinness. The problem of perfectionism. I see the decision to increase the price of heavy and sophisticated weapons and the speed limit for the production of these types of weapons (not increase by 1 every day, such as 1 every 2 days or 1 every 4 days)


Great direction. I remember the same thing from SOE1. I could produce tons of tanks DAILY. Something like 100-200 T-34/Pz IV daily was normal.
But in history Germany/USSR could not produce 6.000+ tanks monthly (80.000-100.000 yearly). In fact USSR produced 100.000 tanks in 5 years.
In my game as Germany I produced them all in 1 year.

Solution? Случайные числа.
Random numbers. Please introduce random numbers like this EXAMPLE:
50% - no change in production level every day
25% -increase production level every day
25%- decrease production level every day
Efect = SLOWER increase of industry/production level in facories/cities/ports

Random numbers in level of prodction EXAMPLE:
Production line T-34 Tanks level 10.
25% - production of 10 tanks every day
25% - production of 7 tanks every day
25% - production of 13 tanks every day (высокая продукция промышленности)
25% - production of 0 tanks every day (трудности промышленности)
Efect = SLOWER production of equipment.

Re: Discussion about game

PostPosted: February 13th, 2014, 2:59 am
by trurle
Randomizing factory output will be meaningless because warehouse system in SOE2 will smooth out all variations.
Effect will be the same as from stable reduced production.

Re: Discussion about game

PostPosted: February 13th, 2014, 4:26 am
by dalibor
It's all quite complicated ...
much easier to establish new (higher) prices. D.

Re: Discussion about game

PostPosted: February 13th, 2014, 3:00 pm
by Igor Kulakov
Ok. In version 1.3.3 be corrected prices for superheavy-artillery (and others) and correct speed limit for increasing - "weapons per day"

Re: Discussion about game

PostPosted: February 17th, 2014, 3:25 am
by dalibor
OK. This is good news.
What will be the daily limit for different weapons? (examply as heavy artillery and others) D.

Re: Discussion about game

PostPosted: February 17th, 2014, 8:55 am
by trurle
Daily limit is limited by economy/cost. Of course, i understand economy levels need an extensive clean-up, but it will take months for all countries because economy model is embedded in map.
Simply speaking, if Denmark seems to be too capable with production
1) Costs of equipment will be increased (it is more expensive to produce anything in small country, because full toolchain cannot be produced domestically - therefore tools are imported on premium price)
2) Cities and ports will be downgraded

But this is purely hypothetical way.
With extensive map fix underway (now working through Caribbean islands) balancing economy and price should wait until end of main fix.

Re: Discussion about game

PostPosted: February 18th, 2014, 6:49 am
by trurle
To Igor Kulakov:
Pool of wounded in SOE2 is working too slow.
It must be deciding the fate of soldiers 20% per day, not 1% as now.
See statistic attached for Vietnam/Iraq wars: ~50% wounds are light wounds, resulting in return-to-duty within 3 days.
Also, 100% KIA in naval battle is not realistic. Realistic is 80% KIA, 20% WIA

Also, casualties calculation during attacks on naval transports is incorrect.
1) Seems land units can use their AA artillery onboard of transports - bug?
2) Too small casualties against small transports (<1000 men). Single submarine should kill them off in one attack. For now, seems submarines tries to torpedo every lifeboat (with nearly zero results). But in practice submarines against weak opponents has surfaced and raked them with cannon fire. Also, small transports (lifeboats) tend to surrender to enemy submarines because their chance of being rescued are too slim.

Re: Discussion about game

PostPosted: February 25th, 2014, 3:25 pm
by trurle
I have tried to test AI of SOE2, therefore run super-scenario of WWII with only Germany under AI control and all other countries completely passive.
Conclusions are following:

1) Contrary to the Igor Kulakov claim, amphibious assaults are performed by any unit type (including coastal artillery) which happened to be in the port suitable for disembarcation
2) Units leaving port are moving to the closest unoccupied enemy object and land NEAR it - i.e. if unit have the choice to capture port or to land near it (even with immobile units) it will never capture port on landing.
3) Units ignore undefended areas if even 1 enemy units is closer to them - i.e. priority is shortest range, not the easiness of capture
4) Units are not cooperating - for example if unit attack superior enemy, it will block other friendly units from reaching it.
5) Units can move only if such movement bring them close to target - therefore gathering on the port-less peninsula, trying to swim to target without sea.. (see around 1088,324)
6) Air units are never combined in stacks and are generally used up suicidally.
7) Land units have priority of defense of property over offense - sitting in low-value town and ignoring nearby large undefended cities if they consider their town under danger (see 967,380)
8) Sea units are never stacked and act against closest enemy only - be it land or naval units.
9) Damaged naval units head for nearby friendly port, whenever it has free slots or not. Usual result is stalemate: damaged ships cannot come close to port and repaired ships cannot depart from port vicinity

Overall:
The AI comprise a horde of DEMONS :twisted: - mindless (moving to closest enemy), not cooperating, moving on straight line, cannot cross flowing water (without demonic engineers), afraid of churches (i think the ports includes at least 1 church :lol: )
Why nobody noticed? In this case, showing cross to monitor wins the game automatically! :P

Re: Discussion about game

PostPosted: February 25th, 2014, 4:21 pm
by dalibor
Some wrote it ....
There is unfortunately a lot of other imperfections in the behavior of AI.

But I have a bit of Igor resentment.
Write quality algorithms for AI strategy games is really hard ...
Exceptionally difficult for games, where each state can attack or defend all world continents.Dali