Matrix Games Forums

This is much to do about nothing.[:)]

ORIGINAL: Itdepends

If the seed is set as Bill says when the scenario is selected then you could use this to exploit/cheat turn one without the Allied players help.

1.Select scenario
2.Save japanes password and then copy the save game file
3. Open original file. Save dummy allied password (as Japanese player)
4. Open file as Japanese and run combat replay.

If you don.t like the result start from step 1. If you do like the result, use the copy of th file you saved at step 2 for all your future opponents.

I.e. This assumes the historical first turn outcome is set once the seed is set when selecting the scenario (which is what Bill implied)

This also can't work without your opponent's cooperation because of the bolded part.

The seed is not set when selecting the scenario. The seed should be set when the Allied player "closes out" the turn. If you want to test this, make a backup save and then close out a turn as the Allies and load it up at Japan. Then open your backup save, make a minor change, close it out, and send to Japan.

In this particular instance, you could test this by:
1) Close out turn as Japan
2) Open turn as Allies, set password, and it closes out
3) Run as Japan. Record results.
4) Re-open original turn from Japan as Allies, set a different password and close it out
5) Run the turn as Japan, record results.

#3 and #5 might differ. Maybe.

ORIGINAL: Lowpe

This is much to do about nothing.[:)]

I really shouldn't be surprised by how much people are holding onto the idea that there's some kind of exploit here and I will continue to rebut it until my computer burns out.

ORIGINAL: tomamars

Way out of this mess is rather simple one. How long does it take for IJ player to create first sav game with those settings and send it to Allies? 1 min? And how long does it take to view results of the 1-st turn? 30 min? How about 10 1-st turns to choose from one of your liking? 5 hours? So solution is rather simple, all allied player needs to do is to observe if file IJ player sent him was ACTUALLY created few minutes ago or few days (or months) ago. Even if an exploit was indeed possible, there is a simple way to counter it by just paying attention.

The Japanese player could run the turn that the Allied player sent to him repeatedly, until the cows come home, and always get the same results.

Always.

Every time.

Just like I'm going to say the same thing every time I have to respond to this. I should really just go back up and find my reply that lays out how it works and quote that every time.

ORIGINAL: PaxMondo

ORIGINAL: Quixote

Aurorus is correct about the potential exploit. Instead of arguing back and forth about what should happen when starting a new game with the same Japanese seed turn, try actually doing it. I did, and the result of the same Japanese seed turn, when sent to two Allied opponents with completely different passwords, was two exactly identical turns. These weren't similar results or close results factoring in fog of war, they were identical. Each Allied player lost exactly the same ships, exactly the same number of aircraft, and suffered exactly the same amount of damage to each and every ship hit. The experiment is easily repeatable if you care to try.

Note that this potential exploit only applies to Historical Dec 7th starts, and still may not mean his opponent was knowingly cheating. It's certainly possible his opponent simply re-used an already prepared first turn without knowing about the exploit (since none of the veteran players here seemed to know about it either), but Aurorus isn't crazy to have at least considered the possibility.

It requires you to set the allied password, which is breaking the process.

I'm sorry, that's not a cheat. that's not following the process and the allied player allowing the IJ player a free turn. since both players are involved, kinda hard to call it a cheat.

Pax, try this yourself. If the Japanese player sets an allied password, then saves the game in a different slot, the original PBEM save does not have the allied PW entered. By then reopening the save with the dummy PW as Japan - not the original PBEM save - the Japanese player can view the results of the turn, while keeping the original save in the state where its waiting for the initial PW setup from the allied player.

However, if the seed is based on the allied save, as others have suggested, that may be moot, but I think that its important not to dismiss the possibility.

ORIGINAL: Lokasenna
Just like I'm going to say the same thing every time I have to respond to this. I should really just go back up and find my reply that lays out how it works and quote that every time.

Except that from what I saw, you're missing a step on how this exploit is said to work.

ORIGINAL: Lokasenna
Well, not really. This is prevented because the Japanese player has to select the scenario, enter a password, and send to the Allied player before doing anything. The Allied player then opens that file, enters a password, and sends it back.

That file would then have the exact same replay generated for it every single time it was reloaded. In order to re-run the first turn until optimal results were achieved, the Japanese player would have to receive the file back from the Allied player each time - starting from scratch, each time, so both players would have to enter passwords.

Let's walk through a potential scenario:
1. Japanese player starts a new PBEM game, sets their password and saves in slot 10.
2. Japanese player re-opens the slot 10 save, sets an allied password, and saves in slot 11.
3. Japanese player re-opens the slot 11 save, runs the turn, and views the combat results.
4a. If the Japanese player likes the results, they then send the save in slot 10, which is still ready for the allied player to set the password.
4b. If the Japanese player doesn't like the results, they then go back to step 1, and start a new PBEM game.

The exploit doesn't come from running the save save over and over, but by instead creating new games over and over and being able to check the results.

I've tested this myself, and it seems to work. The question is if the results would be the same if the allied save was done on a different computer, which I don't know.

ORIGINAL: Lokasenna

ORIGINAL: tomamars

Way out of this mess is rather simple one. How long does it take for IJ player to create first sav game with those settings and send it to Allies? 1 min? And how long does it take to view results of the 1-st turn? 30 min? How about 10 1-st turns to choose from one of your liking? 5 hours? So solution is rather simple, all allied player needs to do is to observe if file IJ player sent him was ACTUALLY created few minutes ago or few days (or months) ago. Even if an exploit was indeed possible, there is a simple way to counter it by just paying attention.

The Japanese player could run the turn that the Allied player sent to him repeatedly, until the cows come home, and always get the same results.

Always.

Every time.

Just like I'm going to say the same thing every time I have to respond to this. I should really just go back up and find my reply that lays out how it works and quote that every time.

Lokasenna, the point they are trying to make is that because the Allied player won't set orders/ won't change the turn in any way, a turn saved with another Allied password will bring the same results. So no cooperation needed by the Allied player, his own password will behave exactly as the dummy one set by the Japanese player

1. Allied and IJ player decide to have a game
2. IJ player creates first save and sends it to Allied player
3. Allied player inspects time save file was created and / or changed and takes it ONLY if it was created in last few minutes - e.g. not being pre-prepaired
4. Allied player confirms the save and sends it back to IJ player
5. IJ player views combat report and starts working on his turn.

But maybe it's best to let Matrix people have their oppinion on this one...

ORIGINAL: paradigmblue
However, if the seed is based on the allied save, as others have suggested, that may be moot, but I think that its important not to dismiss the possibility.

Until there's word from Matrix....according to michaelm:-

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

The quote was from this thread (I assume 2010 info is still applicable today)
tm.asp?m=2420214

ORIGINAL: tomamars

1. Allied and IJ player decide to have a game
2. IJ player creates first save and sends it to Allied player
3. Allied player inspects time save file was created and / or changed and takes it ONLY if it was created in last few minutes - e.g. not being pre-prepaired
4. Allied player confirms the save and sends it back to IJ player
5. IJ player views combat report and starts working on his turn.

But maybe it's best to let Matrix people have their oppinion on this one...

Just wanted you to know, there are free programs available to change these attributes to what ever you want.

Guys, none of that matters.

We know that the seed is set not when the Japanese player ends his turn, but when the Allied player ends his turn (Allied player can alter orders with just 1 change and change the seed this way between 2 otherwise identical turns).

ORIGINAL: paradigmblue

ORIGINAL: Lokasenna
Just like I'm going to say the same thing every time I have to respond to this. I should really just go back up and find my reply that lays out how it works and quote that every time.

Except that from what I saw, you're missing a step on how this exploit is said to work.

ORIGINAL: Lokasenna
Well, not really. This is prevented because the Japanese player has to select the scenario, enter a password, and send to the Allied player before doing anything. The Allied player then opens that file, enters a password, and sends it back.

That file would then have the exact same replay generated for it every single time it was reloaded. In order to re-run the first turn until optimal results were achieved, the Japanese player would have to receive the file back from the Allied player each time - starting from scratch, each time, so both players would have to enter passwords.

Let's walk through a potential scenario:
1. Japanese player starts a new PBEM game, sets their password and saves in slot 10.
2. Japanese player re-opens the slot 10 save, sets an allied password, and saves in slot 11.
3. Japanese player re-opens the slot 11 save, runs the turn, and views the combat results.
4a. If the Japanese player likes the results, they then send the save in slot 10, which is still ready for the allied player to set the password.
4b. If the Japanese player doesn't like the results, they then go back to step 1, and start a new PBEM game.

The exploit doesn't come from running the save save over and over, but by instead creating new games over and over and being able to check the results.

I've tested this myself, and it seems to work. The question is if the results would be the same if the allied save was done on a different computer, which I don't know.

The slot you save it in doesn't matter a whit.

And the key part of this is that you can't do this without the Allied player letting you set their password. That's what you're not understanding. Before the results are generated, the Allied player has set their password.

You can set an Allied dummy password all you like, but when the Allied player reports to you that their password isn't working on your dummy password replay and dummy password turn file, you'd have no choice but to say "Oh, your password is actually (dummy password)."

The only replay and turn you'll be able to actually send back to the Allied player is the one that is based on the turn file they sent back to you, after they set their password - and the replay and turn you send back to them will be the same no matter how many times you run it, or how many different slots you save it in, because the only turn you can run to generate these files is the one the Allied player sent to you with his password.

ORIGINAL: tomamars

1. Allied and IJ player decide to have a game
2. IJ player creates first save and sends it to Allied player
3. Allied player inspects time save file was created and / or changed and takes it ONLY if it was created in last few minutes - e.g. not being pre-prepaired
4. Allied player confirms the save and sends it back to IJ player
5. IJ player views combat report and starts working on his turn.

But maybe it's best to let Matrix people have their oppinion on this one...

Time of save file is unimportant entirely. What's to say I didn't run the turn and send it to you hours or days later? It's happened in my games several times - simply forgetting to send the file.

ORIGINAL: paradigmblue

ORIGINAL: Lokasenna
Just like I'm going to say the same thing every time I have to respond to this. I should really just go back up and find my reply that lays out how it works and quote that every time.

Except that from what I saw, you're missing a step on how this exploit is said to work.

ORIGINAL: Lokasenna
Well, not really. This is prevented because the Japanese player has to select the scenario, enter a password, and send to the Allied player before doing anything. The Allied player then opens that file, enters a password, and sends it back.

That file would then have the exact same replay generated for it every single time it was reloaded. In order to re-run the first turn until optimal results were achieved, the Japanese player would have to receive the file back from the Allied player each time - starting from scratch, each time, so both players would have to enter passwords.

Let's walk through a potential scenario:
1. Japanese player starts a new PBEM game, sets their password and saves in slot 10.
2. Japanese player re-opens the slot 10 save, sets an allied password, and saves in slot 11.
3. Japanese player re-opens the slot 11 save, runs the turn, and views the combat results.
4a. If the Japanese player likes the results, they then send the save in slot 10, which is still ready for the allied player to set the password.
4b. If the Japanese player doesn't like the results, they then go back to step 1, and start a new PBEM game.

The exploit doesn't come from running the save save over and over, but by instead creating new games over and over and being able to check the results.

I've tested this myself, and it seems to work. The question is if the results would be the same if the allied save was done on a different computer, which I don't know.

So basically you're saying here that... the Japanese player could just test his first turn? Big deal. You can do that without an Allied player's help on historical first turn, although I don't know why you would - it's not like you have any control over it, and as I've stressed 180 times in this thread you can't run and re-run and re-run the turn the Allied player sent to you until you get results that you like. It's going to be the same results every time.

This is a CORE COMPONENT of how the PBEM system works. If the results were not 100% repeatable (assuming running in the same version of the program), it would be literally unplayable.

ORIGINAL: Lokasenna

Guys, none of that matters.

We know that the seed is set not when the Japanese player ends his turn, but when the Allied player ends his turn (Allied player can alter orders with just 1 change and change the seed this way between 2 otherwise identical turns).

But in this case, the allied player can't issue any orders because of the historical first turn, and as such can't change the seed.

Stuff about re-running a turn.

We're talking past each other here, I think. This scenario has nothing to do with re-running the same save over and over. It's about testing out a first turn save and being able to see if it's the result you want, and if not, creating a *new* (as in starting at the scenario selection screen) save and testing the new one. If the seed is generated by the Japanese save, this allows the Japanese player to essentially create new seeds and test them one by one.

Again, not by re-opening the same turn over and over, but by using this method to peek at the first turn results, and creating a new game with a new seed if it's not to their liking.

ORIGINAL: Lokasenna
I will continue to rebut it until my computer burns out.

Be careful - and throw it out before your house catches fire too:

ORIGINAL: paradigmblue

ORIGINAL: Lokasenna

Guys, none of that matters.

We know that the seed is set not when the Japanese player ends his turn, but when the Allied player ends his turn (Allied player can alter orders with just 1 change and change the seed this way between 2 otherwise identical turns).

But in this case, the allied player can't issue any orders because of the historical first turn, and as such can't change the seed.

Stuff about re-running a turn.

We're talking past each other here, I think. This scenario has nothing to do with re-running the same save over and over. It's about testing out a first turn save and being able to see if it's the result you want, and if not, creating a *new* (as in starting at the scenario selection screen) save and testing the new one. If the seed is generated by the Japanese save, this allows the Japanese player to essentially create new seeds and test them one by one.

Again, not by re-opening the same turn over and over, but by using this method to peek at the first turn results, and creating a new game with a new seed if it's not to their liking.

OK, I finally understand what you're saying here... Basically, that the IJ player can run their historical first turn any number of times against a dummy password, saving a fresh copy each time prior to the Allied password setting, until they got a result they liked. Then, that file would be sent to the Allied player. This is rendered moot if you are playing with historical first turn set to off, as the Allied player would enter orders after the Japanese player last touched the turn.

This relies entirely on the seed being set when the Japan player loads the scenario for the first time, and not when the Allied player enters their password. You could be right, but I'm not so sure. From observation, the seed seems to be set either when the Allied player closes out their turn or when the Japanese player loads the turn and generates the replay after receiving the turn from the Allied player (this observation comes from lots of experience with sync bugs, and running over 1200 PBEM Japan turns thus far*). Why would the first turn be any different?


*One time, I loaded the turn using the wrong version of the game and it caused a completely different replay AND different turn file. My opponent loaded the replay in the correct version of the game and his replay was what I saw when I loaded in the correct version. This tells me that the seed is probably set when the Allied player closes the turn and the replay is generated based on (1) the seed, and (2) the version of the game (it is probably interpreting the seed differently). Alternatively, the seed could be generated from a combination of a checksum of the turn file and the version of the .exe.

In the midst of all the PC walloping and, umm, belt measuring, I have to point something out.

No matter what happens on Turn 1, in the long run it doesn't matter.

Hey, it's what I do.

ORIGINAL: Lokasenna

OK, I finally understand what you're saying here... Basically, that the IJ player can run their historical first turn any number of times against a dummy password, saving a fresh copy each time prior to the Allied password setting, until they got a result they liked. Then, that file would be sent to the Allied player. This is rendered moot if you are playing with historical first turn set to off, as the Allied player would enter orders after the Japanese player last touched the turn.

This relies entirely on the seed being set when the Japan player loads the scenario for the first time, and not when the Allied player enters their password. You could be right, but I'm not so sure. From observation, the seed seems to be set either when the Allied player closes out their turn or when the Japanese player loads the turn and generates the replay after receiving the turn from the Allied player (this observation comes from lots of experience with sync bugs, and running over 1200 PBEM Japan turns thus far*). Why would the first turn be any different?

Yep, as I've said before, this could be an absolutely moot point, if the allied player is the one that sets the seed. I suppose what I need to do is a test where I run a dummy turn, record the results, then restart the computer and run a dummy turn again - haven't the devs said before that the two ways of changing the seed were a) changing orders b) restarting the computer?

I'll try that after work today.

If you look at post #57 of this thread you will see that Quixote has all ready tested it and the RNG seed is set
when the Japanese player sets up the game.

ORIGINAL: BillBrown

If you look at post #57 of this thread you will see that Quixote has all ready tested it and the RNG seed is set
when the Japanese player sets up the game.

Thanks Bill.

ORIGINAL: BillBrown

If you look at post #57 of this thread you will see that Quixote has all ready tested it and the RNG seed is set
when the Japanese player sets up the game.

This is simply a sad, sad thread.[:(]

Amen!