GalCiv2 Fleet Combat Simulator » Forum Post by kryo

According to Cari all rolls are min 0, the sim has been readjusted accordingly.

Successfully updated karma reason!

According to Cari all rolls are min 0, the sim has been readjusted accordingly.

1...N would lead to deadlocks. Each ship has attack 1 (any type) and defense >= 1 (any type). Each attack would roll a 1 and each defense would roll >= 1, blocking the attack and resulting in no damage. Since battles continue until one side is destroyed you would have an infinite loop.

Successfully updated karma reason!

OK. Where do I click to download? Don't see anything here or on the download page. What am I missing?

Successfully updated karma reason!

Nothing to download. Just use the thing at the top of this thread. Go on... it's interative!!!

Successfully updated karma reason!

Wow.....nice work, so far the predictions are pretty accurate, tried it based on battles I’ve fought, got 7/10, impressed

Successfully updated karma reason!

I get a much smaller width column for the defenders fleet list than the attackers. It might be the valign="center" setting on the defenders td table html. Not sure though. It probably renders fine on IE, but Mozilla is a bit pickier sometimes

Nice idea, thanks for hosting it.

Successfully updated karma reason!

When an armed ship fires upon a ship with no defenses, the game never rolls a zero attack, but the simulator can.

Successfully updated karma reason!

If you simulate two small fighters with good defence and light offense, you get a random result for who wins and remaining HP of winner. I believe this is the same result recieved in-game. That can make for some long battles!

Successfully updated karma reason!

I forget where I read it, but it was an official thing ( but this is off of my memory)
Defense value = 1 + armor value. Then after you get that number (e.g. 1+ 2 Armour value = 3) it roles that number for that turns defense from 0 to the DV (E.G 0 to 3 ).

Then this table come into effect
So lets make an example between two ships. One with laser attack of 3 and one with a rail attack of 3
For this example assume every one rolled max DV ( defense value) and max AV ( attack value )
Vs rail 3 AV VS. 3 DV (armor) = damage 0 damage taken that round
Vs Laser 3AV VS. 3 DV ( armor) : Well here it takes a slight turn fro the worse. since its a laser attack, armor only does a squared root of it's defense witch equal 1.73. I assume it gets rounded down ( that part was never explained)
so you end up with

Vs Laser 3AV VS. 1 DF(armor) = 2 damage done to you.

Def Armour Shields Chaff/ EMP

Laser /def value squared / full /.def value squared

Rail full / def value squared/ def value squared

Missile def value squared/ def value squared / Full

Is that of any help to anyone?

Successfully updated karma reason!

Model Error:

Having played the game for a while and watched the details of battles, I have identified two problems with the current simulation model. The first is that the attack roll is never zero, but always at least one, provided that there is an offensive weapon. And the defence roll is always at least one, if there is any defence. This is as stated in the manual. This does mean that a 1/1 ship versus a 1/1 ship gives an indeterminate answer. The programers probably resolved this only exception with a coin flip, but it would be nice to know what actually happens here. The result of the error is that defensive ships are undervalued. For example, a 2/1 ship will normally defeat a 3/0 ship, but the simulator will show the opposite.

Secondly, the calculation only applies to single ships, and will give incorrect answers for fleets even with the above correction. This is because a fleet does not make a single role as stated incorrectly in the manual. Each ship in the fleet that is still alive does an individual roll against the target ship (largest value of Attack/HP left I believe). This means that the number of hit points on individual ships in each fleet is important because some will be knocked out early in the battle. For example , everything else equal, a 40 HP fleet with 2x20 HP should beat a 40HP fleet with 4x10HP because the latter is likely to lose some of its offensive power before the bigger ships lose any. This gives a big advantage to the larger ships int he ensuing battle after the first small ship is eliminated.

An exact simulation for any two fleets should be possible, but is rather complicated. Any takers?

By the way, Space Empires 4 allows you to do this by a rapid (or slolw blow by blow if you want) simulation of any two fleets (including armed planets as well incidentally ) which can be repeated several times in order to get a rough probability of success. It would be nice if GalCiv could do the same in an expansion pack.

Successfully updated karma reason!

Having played the game for a while and watched the details of battles, I have identified two problems with the current simulation model. The first is that the attack roll is never zero, but always at least one, provided that there is an offensive weapon. And the defence roll is always at least one, if there is any defence. This is as stated in the manual. This does mean that a 1/1 ship versus a 1/1 ship gives an indeterminate answer. The programers probably resolved this only exception with a coin flip, but it would be nice to know what actually happens here. The result of the error is that defensive ships are undervalued. For example, a 2/1 ship will normally defeat a 3/0 ship, but the simulator will show the opposite.

As noted on the official wiki, I've spoken with Cari (the lead devloper) on this before, and she said that all rolls are 0-max, not 1-max. Assumably, zero-rolls are simply not displayed.

Secondly, the calculation only applies to single ships, and will give incorrect answers for fleets even with the above correction. This is because a fleet does not make a single role as stated incorrectly in the manual. Each ship in the fleet that is still alive does an individual roll against the target ship (largest value of Attack/HP left I believe).

The manual states it correctly, though not particaulrly clearly. Each ship rolls an attack in turn against the highest priority target, then any remaining ships on the other side return fire in the same manner, continuing until one side is destroyed. Targetting is based on the highest result from the formula (Attack / (Defense+HP)).

This means that the number of hit points on individual ships in each fleet is important because some will be knocked out early in the battle.

This is a well known fact of the GC2 battle system, and is one of the reasons that the attackers always firing first can be such a great advantage. Personally I use a mixture--masses of high-ATK light fighters that can punch out the enemy before they can fire back, so their low HP is never an issue.

This simulator is as mechanically accurate as the community at large knows. Its current downfall is that it only runs the battle once, giving in-depth details on just that one simulation. As noted by another forumer who developed a Java-based percentile sim, it can take up to 10,000 runs to produce results with the margin of error reduced to an acceptable amount. But even that is no guarantee, as the actual battles in game can be just as random as the sim here--and you only get one chance in game, not 10,000.

Successfully updated karma reason!

Next thing you need is to make it graphic supply a ship type with modules, engines or better still some way to take custom made ships and have them go through the simulator. See the action play out.

Successfully updated karma reason!

I would very much like to see a com bat simulator added to the game complete with importing of ship designs and a graphical display. In fact, this is one addition I would pay extra for.

Successfully updated karma reason!

This might sound like a stupid question but dosent luck and your starting party stuff come into effect for battle

Successfully updated karma reason!

Very nice simulator, but something confuses me. In all the time I've played GC2, I don't recall ever seeing an attack roll of 0 when shooting at a ship with no defense, yet in the simulator it happens relatively often (with low attack values). I'm pretty sure it's not that they are not displayed in the game, because I keep track of all the ships that fire in the game, and this hasn't happened, whether shown or otherwise. From what I've seen, though I may be wrong, is that the defense roll is from 0 to max, while the attack roll is always atleast one. I may be wrong though, but that's what I appear to observe when playing the game.

Successfully updated karma reason!

BTW, has this (or will this) be updated to account for the new attack system? Both sides attack simultaneously instead of 1 after another.

I wasn't sure if this was going to be introduce in 1.2 or already has for the latest 1.11.

Successfully updated karma reason!

The change won't be until 1.2. I'll try and have the sim updated after it comes out.

Successfully updated karma reason!

Ive been messing around with this just thought I'd post what I'd found out:

In all these examples both sides have 500hps: A=attack(laser) D=defence(barriers)

I chose high health to reduce the effect of first strike and somewhat high stats to reduce the effect of the 0 attack rolls

A 10A 10D vs a 20A 0D
close battle, defender seems to have slight edge (the defender being the one that has no defences)

a 20A 20D vs a 20A 0D
attacker loses half health, defender loses

a 40A vs a 20A
attacker loses half health, defender loses, so no special advantage to a matched defence.

a 10A 200D vs a 100A 0D ship
attacker loses, defender loses half health. Too much defence makes it even worse!

A 20A 10D vs a 30A 0D ship
close battle, defender seems to have slight edge

a 10A 20D vs a 30A 0D ship
attacker loses, defender loses about 2/3 health

So never have more defence than attack. And in big ship battles it is very much overrated.

However if you are up against tiny fighters then defence is the way to go!

five 10hp 10a 0d fighters vs a 50hp 50A 0D battleship
the fighters lose about half their ships, battleship is pwned. This is even more the case
when attack is much greater than hps, the fighters lose only 1 ship.

five 10hp 10a 0d fighters vs a 50hp 25A 25D battleship
battleship wipes out the fighters only losing half health.

So in summary to easily beat a high attack 0 defence capital ship use lots of fighters

to beat lots of fighters use a capital ship with balanced attack+defence

To beat a capital ship with balanced attack+defence use a high attack 0 defence capital ship, or use a different weapon.

To beat a bunch of fighters all using different weapon types, use a capital ship with all defence types. However a multi-defence capital ship is easily beaten by a high attack 0 defence capital ship.

If your tech is similar or worse than the enemies, then use fighters making sure some or all of them can penetrate the capital ships defences. This pretty much beats everything but you lose fighters every time. But trading a fighter for a capital ship is worth it. It only really works though if the battle is over in 1 round which is very common at high tech levels since hps don't increase compared to weapons.

if your tech is superior then a high attack/multi defence capital ship beats everything since there won't be any high attack capital ships to worry about.

If tech is inferior or similar then focus on multiple weapons techs, if tech is superior then research a single weapon and multiple defences.

I was getting the impression that a matched defence was supposed to always beat a matched weapon, and was confused when clearly it didn't. But once you understand that defence is designed to be used against fighters, not capital ships it all becomes much more understandable.

Successfully updated karma reason!

Wher is the simulator. I get a blank page at the top of theis post. Is there a link?

Successfully updated karma reason!

It appears to have gotten lost when my server crashed last week--I'll try to get it back together and updated for 1.2 sometime this week.

Successfully updated karma reason!

requested url not found. any suggestions?

Successfully updated karma reason!

Learn to read.

Successfully updated karma reason!

Learn to be cool, and not slam

It seems they are working on an update for the latest version

thanks for all the fish

hitchhiker

Successfully updated karma reason!

Link's broke.

Successfully updated karma reason!

Sim updated (finally), sorry for the delay.

It should now be more useful. Instead of doing a single run and showing the exact results (which can be way off from what actually happens), it now does 1000 runs and gives the resulting win rate for each side, as well as the average ships lost for each and a breakdown of the losses that shows how common each number of ships lost was.

With 1000 runs, the final win probabilities should be off by no more than 2%. It does create much more strain on the server, however, so don't abuse it. If you do, you may find yourself blocked from my server or the sim may simply be taken down for everyone (or moved elsewhere, if someone else wants to deal with it).

And yes, it is also updated for the new 1.2 battle system (ships die at the end of each round instead of as soon as their HP drops to 0).

Successfully updated karma reason!