Planetarion Forums - Combat Report does not agree with published Combat Formlae

Thrud · 1 Jul 2007, 00:22

I was shown the following BR as the bcalc didn't get the calc correct;

Combat report for combat at somewhere (x:x:x)
Attacking fleets: 1 Defending fleets: 4
SHIP Att Killed Held Stolen Def Killed Held Stolen
Wyvern 50 0 0 34 50 0 0 0
Dragon 154 0 0 107 200 0 0 0
Leviathan 50 0 0 35 92 0 0 0
Pirate 0 0 0 0 361 185 0 0
Ranger 0 0 0 0 3313 852 0 0
Totals 254 0 0 176 4016 1037 0 0
Values 17680 0 0 12253 72195 14821 0 0

Hardly surprising as there are enough stealers in the defending fleet to cap twice the number of attacking BS, yet they didn't cap it all, they only capped 66%. I have not found any bcalcs so far that agree with this BR (mine,xVx and LCH), and working out by hand from the manual shows all attackers stolen, the same as the bcalcs, whats going on, how did the game get to this result ?

Any help would be appreciated

JonnyBGood · 1 Jul 2007, 00:27

Something completely ****ed up happens when pirates and rangers are defending together, at some point last round I told everyone in ascendancy to avoid taking part in combat with both ships. Clearly it hasn't been fixed yet

Thrud · 1 Jul 2007, 00:34

This is an all new ****ed up, last round I could workout how it had changed and managed to get the calc to match mostly, this just looks like it guessed the answer.

Mzyxptlk · 1 Jul 2007, 01:38

Wasn't there something about Cath and Etd EMP ships as well?

P.S.
It is slightly ironic that the bcalc is better at simulating battles that conform to the formulae used than the actual game is.

SpaceMonkey · 1 Jul 2007, 15:28

From what I could make out the stealing is something to do with relative fleet values similar to roid capping. As JBG said the same thing happened last round.

The roid capping is still bugged as far as I know, so I wouldn't expect the stealing to be any better.

Thrud · 1 Jul 2007, 15:46

No it seems worse than that, Sandman pointed out that when compared to the Detail Log of the calc, the stolen ships were stolen by the Rangers, the ships stolen by the Pirates magically re-appear after the combat, the value of the ships stolen by the Rangers is shared between the Pirates and Rangers to decide what gets suicided. These numbers could just be coicidence given the small sample data but it looks like this is whats happening.

Achilles · 1 Jul 2007, 16:26

The new combat engine code has never worked correctly since it's implementation in R11. When you start adding layers of complexity to code you aren't fully familiar with, that isn't even written correctly in the first place and that ties very closely with the ticker code then you are going to have problems.

I made a similar point in another thread a while back but my posts were "moderated" by Ultimate Newbie. Most were deleted and 1 was shifted to another thread and quietly buried, one assumes at the request of Remy, who in my opinion was made to look slightly foolish by the whole affair. Unfortunately this discussion is now deleted and the truth is that there is unlikely to be a solution unless the entire ticker code and combat engine are rewritten. That said, without access to the code itself I have no way to know for sure. I can only judge from the fact that these problems have existed since the inception of PAX.

Ultimate Newbie · 1 Jul 2007, 17:33

Quote:

Originally Posted by Achilles

I made a similar point in another thread a while back but my posts were "moderated" by Ultimate Newbie. Most were deleted

iirc, i wanted to keep the discussion between what actually does happen seperate to the posts which suggested/demanded changes to the current system in order to try and reduce confusion for anyone reading that thread. It was pretty chaotic. Besides, three posts in that thread were deleted - two by the original authors themselves, and the third by JBG who was just a spammer trolling all forums.

I'm feeling the hate!

Quote:

and 1 was shifted to another thread and quietly buried, one assumes at the request of Remy, who in my opinion was made to look slightly foolish by the whole affair.

I dont recall having ever seen that thread in my life. Its pretty interesting really. I do recall most of the posts regarding "improvements" being split into a thread on PD on my behalf by wakey, but i dont actually remember seeing you thread.

Remy isnt the sort to ask for things from moderators, or at least me. Even if he did, the mods would do that they thought best, and "PA Team" know better than to prod us on forum issues. Thus, your assertion here is, afaik, totally baseless.

Quote:

truth is that there is unlikely to be a solution unless the entire ticker code and combat engine are rewritten. That said, without access to the code itself I have no way to know for sure. I can only judge from the fact that these problems have existed since the inception of PAX.

The example you quoted, remy replied in the original thread and asked for a battlereport as (he believed at least) that it was a bug. Whether it was or not, i dont know as i was not involved in any bug squashing nor informed of any changes to the code. As such, i cant comment directly as to whether that the issue with the roid formula was "solved" or not. However, whilst stealing might be done in a similar fashion, to the extent of my knowledge they dont use the same code.

The best thing to do if you're having problems is to obviously make a post and work it out with people who have access to the code. I'll direct cin et al to this thread now, assuming they havent seen it already.

Ultimate Newbie · 1 Jul 2007, 17:34

Just to reiterate: there is no conspiracy!

Mzyxptlk · 1 Jul 2007, 17:41

Then what alliance is currently #1 then!?

*cue laughter*

Ultimate Newbie · 1 Jul 2007, 17:49

Quote:

Originally Posted by Mzyxptlk

Then what alliance is currently #1 then!?

I heard that's only a theory

.

Quote:

*cue laughter*

I chortled.

...

I lied

Achilles · 1 Jul 2007, 17:52

Quote:

Originally Posted by UN

The example you quoted, remy replied in the original thread and asked for a battlereport

Indeed, which I supplied. This prompted a debate between myself and Remy as regards both the combat engine and PATeam's understanding of same. This entire conversation was deleted by A.N.Moderator, something to which I objected to you about at the time. I hate to paraphrase but your reply was essentially that Remy had asked you to put the thread "back on track" and you had. Which, obviously, completely buried (once again) the debate I have been trying, without success, to raise since R11.

Zirikk · 14 Jul 2007, 10:10

I just noticed this:

http://bcalc.thrud.co.uk/index.php?loadfile=main8196a1
^ In Thruds calc pirates aren't doing anything.

http://bcalc.lch-hq.org/index.php?id...02791184402483
And pirates seem to work fine with LCH-calc.

Which one is correct?
And with correct I mean how PA calcs it.

Wandows · 14 Jul 2007, 10:19

Hard to tell... Rangers and Pirates combined cause some really strange behaviour regarding steal ships dying, actual total ships stolen and how these are spread between different participants. In short, they simply shouldn't be combined in any combat (unless you don't mind being surprised).

Zirikk · 14 Jul 2007, 10:45

Well I hate surprises but I'm still defending.
I just hope that the attacker is using LCH-calc.

Nadar · 14 Jul 2007, 22:06

Bcalcs vs. formulaes hasn't been cooperating so far this round.

(Sorry for mispellingt things, but I've had faaar too many beers tonight!)

Thrud · 15 Jul 2007, 01:41

I have been trying to match the game, but as shown in another thread the game does not appear to be consistent so I have taken that out, so I now match the published formula, mine and LCH calc should agree once more, but will may not match the game.

AdmV0rl0n · 17 Jul 2007, 10:48

Combat
Save 540
Combat report for combat at Avonmore (13:3:12)
Attacking fleets: 1 Defending fleets: 3
SHIP Att Killed Held Stolen Def Killed Held Stolen
Cutlass 105840 55274 26373 0 0 0 0 0
Corsair 40970 0 0 0 0 0 0 0
Voyager 979 0 0 0 9421 0 0 9421
Vendor 778 0 0 0 12804 0 0 12804
Investor 0 0 0 0 558 0 0 0
Tycoon 0 0 0 0 810 0 0 0
Broker 0 0 0 0 452 0 0 0
Rambler 276 0 0 0 636 0 0 636
Baliff 0 0 0 0 68 0 0 0
Totals 148843 55274 26373 0 24749 0 0 22861
Values 243210 77383 36922 0 154293 0 0 77384
Fleet
Save 540
Report of Losses from the NutCracker fighting at 13:3:12
Ship Arrived Lost
Cutlass 105840 55274
Corsair 40970 0
Voyager 979 0
Vendor 778 0
Rambler 276 0
Report of Ships Stolen by the NutCracker fighting at 13:3:12
Ship Stolen
Voyager 9421
Vendor 12804
Rambler 636

That is not all. Apparently, in the above battle, when capping Ramblers, this offset attacking ramblers in some weird way and prevented capping roids.

Thus, if you see an incoming fleet with a make up of:-

Zik RACE
Zik FI
Zik + Etd CO

If you can have some Ramblers home, You lose no roids at all.

The bug has been reported, and PA team seem aware of it, and they did fix my roid take when I asked nicely.

It would seem the combat engine is in need of a full clean build :/ (Or something)

Ad

Appocomaster · 24 Jul 2007, 22:34

:bump:
currently fixing combat engine bugs.
Fixed the stealing pods drops cap bug mentioned ^^ and will also fix the dodgey steal (hopefully) tonight.
Am planning to upload the first of the two bug fixes, as I'm compensating for that when asked to.

Edit: Found the problem with the second bug. It's because the two anti Bs steal ships are different initatives.

Basically, when the pirates steal, they share the steal ships and losses over both pirates and rangers, despite the fact the rangers haven't fired yet. Therefore, the amount of rangers firing depends on the initial amount of rangers, amount of ships stolen by the pirates, and the initial pirate:ranger ratio.

I've fixed it.

Tietäjä · 25 Jul 2007, 08:23

Quote:

Originally Posted by Appocomaster

Edit: Found the problem with the second bug. It's because the two anti Bs steal ships are different initatives.

Also known as "Monroe Special Treatment Bug".

Game^ · 25 Jul 2007, 10:06

Quote:

Originally Posted by Appocomaster

:bump:
currently fixing combat engine bugs.
Fixed the stealing pods drops cap bug mentioned ^^ and will also fix the dodgey steal (hopefully) tonight.
Am planning to upload the first of the two bug fixes, as I'm compensating for that when asked to.

Edit: Found the problem with the second bug. It's because the two anti Bs steal ships are different initatives.

Basically, when the pirates steal, they share the steal ships and losses over both pirates and rangers, despite the fact the rangers haven't fired yet. Therefore, the amount of rangers firing depends on the initial amount of rangers, amount of ships stolen by the pirates, and the initial pirate:ranger ratio.

I've fixed it.

Has the issues with Roach and Vipers been sorted?

The calc's show the following...

http://bcalc.lch-hq.org/index.php?id...03911185354117

Over freezing of the FR, presuming from the roaches and vipers freezing the same ships.

Is it just the calc's wrong or the combat engine too?

Appocomaster · 25 Jul 2007, 10:08

Quote:

Originally Posted by Game^

Has the issues with Roach and Vipers been sorted?

The calc's show the following...

http://bcalc.lch-hq.org/index.php?id...03911185354117

Over freezing of the FR, presuming from the roaches and vipers freezing the same ships.

Is it just the calc's wrong or the combat engine too?

Was this ever an issue with the combat engine? :/

Game^ · 25 Jul 2007, 10:32

Quote:

Originally Posted by Appocomaster

Was this ever an issue with the combat engine? :/

Benneh told me it was last night? :S

JonnyBGood · 25 Jul 2007, 10:42

I never quite knew if that was intentional or not. It works the same for etd/cath emp ships firing on the same fleet.

Game^ · 25 Jul 2007, 10:51

Quote:

Originally Posted by JonnyBGood

I never quite knew if that was intentional or not. It works the same for etd/cath emp ships firing on the same fleet.

It doesn't work the same for kill ships, so logically why should EMP be different I guess?