Matrix Games Forums

Forums  Register  Login  Photo Gallery  Member List  Search  Calendars  FAQ 

My Profile  Inbox  Address Book  My Subscription  My Forums  Log Out

Please I need clarification on this

 
View related threads: (in this forum | in all forums)

Logged in as: Guest
Users viewing this topic: none
  Printable Version
All Forums >> [New Releases from Matrix Games] >> War in the Pacific: Admiral's Edition >> Tech Support >> Please I need clarification on this Page: [1]
Login
Message << Older Topic   Newer Topic >>
Please I need clarification on this - 3/30/2010 10:25:59 AM   
zace

 

Posts: 183
Joined: 3/22/2010
Status: offline
I started with a post in the war room and have brought it here as I have had a lot of runs with the game now.

Setup. I am running the same turns over and over to establish if there is a way that the RND gets stuck or is altered by doing so. This is because of some results I had while doing alternate setups for turn 1 Ahistoric with IJN. Specifically the more I ran the same save file the worse the results at PH were.

Results are as follows.

Results of testing described below
-------------------------------------------------
Ran the turns and here are the results and the comparison

Scenario 2 KB orders entered then saved - all other orders are as start. Replay of save 9 times

Run 1: 5 BB sunk
Run 2: 3 BB sunk
Run 3: 0 BB sunk
Run 4: 0 BB sunk
Run 5: 1 BB sunk
Run 6: 1 BB sunk probably FOW as only 2 torps
Run 7: 1 BB sunk probably FOW as only 2 torps
Run 8: 1 BB sunk
Run 9: 0 BB sunk.

Scenario 2 historical turn 1 no save just loading the scenario and letting it run then doing it again

Run 1: 3 BB sunk
Run 2: 1 BB sunk
Run 3: 0 BB sunk
Run 4: 0 BB sunk

I am now running alternating scenario 1 and scenario 2 with historic turn 1 to see if the alternate scenarios makes a difference.

Run 1 SCN 1: 0 BB sunk
Run 2 SCN 2: 0 BB sunk
Run 3 SCN 1: 0 BB sunk
Run 4 SCN 2: 3 BB sunk
Run 5 SCN 1: 2 BB sunk
Run 6 SCN 2: 5 BB sunk
Run 7 SCN 1: 3 BB sunk
Run 8 SCN 2: 0 BB sunk (checked from the allied side there were actually 2)
Run 9 SCN 1: 3 BB sunk
Run 10 SCN 2:3 BB sunk


It looks suspiciously like the game actually penalizes running the scenario over and over. When I alternated scenarios this did not happen and slowly the offset was corrected back to some sort of average around 3 reported sunk.

I then decided to do the same from the Allied side.

Turn 1 Scenario 2 historic turn 1.

Run 1: 1 BB sunk
Run 2: 0 BB sunk
Run 3: 1 BB sunk
Run 4: 3 BB sunk
Run 5: 5 BB sunk
Run 6: 5 BB sunk
Run 7: 5 BB sunk
Run 8: 0 BB sunk however all have FLT damage over 60 with 6 of them over 90!
Run 9: 1 BB sunk All have FLT over 80 most in the 90's
Run 10: 1 BB sunk 1 has FLT in 60's all the rest over 85

So for the allies I went from having 7 modestly or not injured BBs to having all 8 mauled when I ran this.

So I went back and started with my original save file from the allies side...This save has probably a dozen runs from the IJN side

Run 1: 0 BB sunk (almost all pristine)
Run 2: 0 BB sunk
Run 3: 1 BB sunk
Run 4: 1 BB sunk
Run 5: 2 BB sunk 3 in red FLT damage
Run 6: ongoing.......

So now that it is allied side the results are getting worse and worse......

Is this something funny with my RND or is there some code that makes it so you don't want to setup a game and run it over and over?

I want to do testing and this is making my life very difficult. I understand that you don't want to give away all the secrets but maybe telling me something like Loading Scenario x and running it before running your actual turn in a good idea (once you have setup the turn you want)... Will this affect the start of a PBEM? I don't want to game the system but this doesn't seem right.

Can anyone duplicate this? BTW this took a long long time luckily I had a lot of time sitting at the computer as I have 5k lines of code flowcharts that have to be done this week to meet the schedule for a product launch... Imbedded code so I am not a windows guy at all.

This makes testing very very hard. Can I get confirmation on this? Can someone run a scenario several times in a row with no changes?
Post #: 1
RE: Please I need clarification on this - 3/30/2010 12:03:17 PM   
michaelm75au


Posts: 13030
Joined: 5/5/2001
From: Melbourne, Australia
Status: offline
The first turn resolution generates the random seed for the game.

Re-doing the first turn should give variable results.
If you would need to get past turn one, and start you replays from turn 2.

For PBEM, the seed is created when the Japanese player ends his turn.


_____________________________

Michael

(in reply to zace)
Post #: 2
RE: Please I need clarification on this - 3/30/2010 12:08:05 PM   
zace

 

Posts: 183
Joined: 3/22/2010
Status: offline
Is it possible that the seed generation has an issue?  Is it possible to get a seed created somewhere else?  I ask this because I continued the testing with fake PBEM games against myself.  In doing this I saw the same continuation of results.

Having 0-1 BB lost in PH on average over 10 turns is not realistic is it?

Is it possible that the Seed or RAND functions used could be processor dependent.  It appears that combats go both ways with noone doing damage to anyone when this is going on.  IE the allies can't hit the broad side of a barn either...

I guess the specific question is.......

When the seed is generated does it have anything to do with past games or saves?  Even if not directly could there be an indirect relation?

(in reply to michaelm75au)
Post #: 3
RE: Please I need clarification on this - 3/30/2010 12:10:43 PM   
michaelm75au


Posts: 13030
Joined: 5/5/2001
From: Melbourne, Australia
Status: offline
The game seed is generated once, on the first turn only.

Following turns, use the game seed plus a known variable (constant for the turn) to ensure each replay is consistent for PBEM.


_____________________________

Michael

(in reply to zace)
Post #: 4
RE: Please I need clarification on this - 3/30/2010 12:13:49 PM   
zace

 

Posts: 183
Joined: 3/22/2010
Status: offline
The seed generation on the first turn.  Is that one dependent on previous turns? 

Does it use a processor dependent advanced algorithm that my particular clock could be harmonizing with (IE one of the advanced random functions that use clock harmonics or background processing to create the number....Way too slow for each turn but for the original seed it could be used.)

Trying to figure out a way to return my first turn seed generation back to random.  As it stands it tends to lock up as shown above.

I appreciate you clarifying this for me.

(in reply to michaelm75au)
Post #: 5
RE: Please I need clarification on this - 4/4/2010 11:55:15 PM   
Disco Duck


Posts: 520
Joined: 11/16/2004
From: San Antonio
Status: offline

quote:

ORIGINAL: michaelm

The first turn resolution generates the random seed for the game.

Re-doing the first turn should give variable results.
If you would need to get past turn one, and start you replays from turn 2.

For PBEM, the seed is created when the Japanese player ends his turn.



I am not convinced that re-doing turn one gives variable results. I have tried it a few times and gotten pretty much the same results. But if I change something in the first turn setup, like adding a couple of new convoys, things change drastically.

(in reply to michaelm75au)
Post #: 6
RE: Please I need clarification on this - 4/5/2010 2:31:12 AM   
michaelm75au


Posts: 13030
Joined: 5/5/2001
From: Melbourne, Australia
Status: offline
quote:

ORIGINAL: zace

The seed generation on the first turn.  Is that one dependent on previous turns? 

Does it use a processor dependent advanced algorithm that my particular clock could be harmonizing with (IE one of the advanced random functions that use clock harmonics or background processing to create the number....Way too slow for each turn but for the original seed it could be used.)

Trying to figure out a way to return my first turn seed generation back to random.  As it stands it tends to lock up as shown above.

I appreciate you clarifying this for me.


Seed does not rely on past turns.
It is initialized based on current time just before the pre-turn save (#2 will have the seed if it is used from turn 1) for AI games and before the Japanese player sets his password in PBEM games.


_____________________________

Michael

(in reply to zace)
Post #: 7
Page:   [1]
All Forums >> [New Releases from Matrix Games] >> War in the Pacific: Admiral's Edition >> Tech Support >> Please I need clarification on this Page: [1]
Jump to:





New Messages No New Messages
Hot Topic w/ New Messages Hot Topic w/o New Messages
Locked w/ New Messages Locked w/o New Messages
 Post New Thread
 Reply to Message
 Post New Poll
 Submit Vote
 Delete My Own Post
 Delete My Own Thread
 Rate Posts


Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI

0.137