The possibilty of manipulation

[Ashtar]

I am sorry to insist, but THERE ARE ways to avoid much of the cheating with pre-generated random numbers and to beat much of the "reloading and doing things in a different order" stuff. Here it is my proposal:

1a. Combat phase - Non trivial combat. Simple way of doing it:
a) The attacker choses his chits and sends to the defender.
b) The defender choses his cheats and the program generates the needed die rolls. None of them are shown (so no reload is possible), and
the file is sent back to the attacker.
c) Results of the first turn are shown to the attacker, he decides about guard commitment and sends back to the defender.
d) The defender receives, check for first turn results and decides about guard commitment. Needed die rolls are secretly generated
and stuff his sent back to attacker...
Go on until the end of the combat. No way of cheating here, large battles are often the crucial and more important ones in a game
and are now secured. Moreover, no more file exchanges then now are required.

1b. Combat phase - Trivial combat and assaults. Random numbers (different from the ones used in foraging duing the land phase) are
pre-generated during the previous player land turn (or naval phase if this is the first turn of land phase).
True, wicked players can alter the order of assaults and trivial battles to look for the most convenient die distribution, but still
these are generally non-decisive gains. Moreover, if you do not trust your opponent, do not give pre-assigned order to single corps.
Otherwise, the order in which assaults and trivial combat are taken can be fixed (by ascending area number, for instance) to avoid
cheating. It is a small price to pay (order of larger combats need not to be fixed) to avoid cheating.

2. Naval phase. This is the most delicate one. Intercept rolls and naval battle rolls here. The player playing the previous
turn pre-generates two different sets of random number, one for interception and one for combat. This way there is often
not so much room for reloading and rearranging order. Still, if you still do not trust the French from trying every turn
to beat GB blockade and invade, only to sistematically cancel his move when his ships are sunk, introduce a Naval combat phase.
It will not take too much time, just a single extra file exchange is needed per battle (and naval battle are much rarer then land one).
Procedure: attacker send out his naval battle file, defender generates the needed die number without seeing them, the file is
sent back to attacker and combat is resolved. Easy, fast and bullet proof.

3. Land phase. Only rolls here are foraging ones. Lets have your foraging rolls be pre-generated
by the player playing the previous turn. Of course, if you do not like the outcome of foraging
you can change their order or decide to pay for supply. This is (almost) unavoidable, but
advantages from this kind of cheating are at best mild.

4. Diplomacy and Economic phase: rolls here are not of extreme importance (influencing minors,
spanish gold, piracy...), but they can be pre-generated for the entire phase when the last
player playing the previous phase ends his turn. This will avoid reloading issues.

It is long to explain in detail, but as you can see, there are really no programming issues here. I guess it is pretty easy to implement.

[Soapy Frog]

I would tend to agree with Ashtar, his solution is a simple and elegant one and will eliminate 95% of cheating.

[AresMars]

Good proposal Ashtar!  Well done!
 
 

[zaquex]

Its a good proposal, im however surprised how few have picked up cunctators suggestion to provide for the use of a dice server. It should be relativly easy to implement and would definatly provide a very reliable control feature that is very hard to bypass.
 
The use of a dice server seems to me like the obvious choice to easily and fast resolve these issues.

[Soapy Frog]

I'd prefer something built into the game that does not require extra emailing.

[zaquex]

As the response from the dice server is automatic it should not create any noticable delay in game progress.

[Grognot]

Yes, if you're willing to do completely interactive combat -- the frequent e-mail exchanges -- that will largely work in a full (seven-human) game.  The main difficulty with land combats is if you're doing PBEM quick combat or the like in order to minimize the back/forth e-mails, in which case a trusted third party becomes useful... well, have to look into the "non-trivial combat against AI" case for non-full games.

I'd also note that a system which generates 'secret' die rolls and saves them in files available to the person who is not supposed to know them, should take some care to ensure that the rolls cannot be readily extracted or tampered with.

[zaquex]

I would argue also that the use of an external game server probably is easier/faster to implement than Ashtars solution wich means (if im right) that its possible to solve the issue faster, this might be the strongest argument for this solution. I also think that the use of a dice server potentially is harder to tamper with than any ingame solution running on your machine. Its also the only solution ive seen that completly deals with issues concerning trivial combat intercepts etc that dont require an exchange of emails. 

[Ashtar]

zaquex:
I would argue also that the use of an external game server probably is easier/faster to implement than Ashtars solution wich means (if im right) that its possible to solve the issue faster, this might be the strongest argument for this solution. I also think that the use of a dice server potentially is harder to tamper with than any ingame solution running on your machine. Its also the only solution ive seen that completly deals with issues concerning trivial combat intercepts etc that dont require an exchange of emails.

It could be that you are right, especially noting that an external dice server should fix completely the problem while my solution still leaves spaces for minor cheats. But:

a) Actually I find my proposal neat and easier to implement then setting up an external dice server which should receive dice requests and send to all player correct logs for each die roll. EIANW it is not tied to any mail client (it tries to use outlook, but you can easily bypass it and send out files with your preferred method). Actually I sometimes do my move offline to connect later on to send it out to my fellow players. This would not be possible if you need a die server.

b) Keping pregenerated rolls secret is not a big issue. Either encrypt them (if someone goes so far to decript them - he should better invest his time attempting credit card frauds then cheating at EIANW) or simply write them in binary format in the middle of the turn files. They should contain hundreds of numbers anyhow, so finding them it is not a practical task.

c) What I do not really like is the idea of receiving hundreds of e-mails with all the die rolls performed by my opponents. This would make my mail box a mess and checking for cheating quite unpractical. Suppose the French tries to avoid British interceptions, he fails and then he decides to reload the turn and keep his fleets in port. As a British, should I check that he is not cheating looking between all the e-mails for intercept rolls not corresponding to actual naval movement? Or, worse, should I check that all my opponents are not reloading land turn switching from failed forage to supply?

Anyhow I would be happy to read that Marshall and his people are taking this issue seriously, whatever solution they chose...

[AresMars]

I think programming realities and costs will determine which possible solution gets looked at.
 
Does Matrix want to host a game server/dice server?  Will they?
 
Both arguements have weight - however, economics still rule the day.....
 
 

[zaquex]

I accept that Ashtars arguments have merits, and agree that it would be interesting to get Marshalls input on the plausability of implementing any of these two in my oppinion very competent solutions. 

[Cunctator]

This discussion, by far, exceedes my personal knowledges, so I'm afraid to say something stupid or useless.
 
Anyway, I thought about a dice server because in 99% of boardgames that are played by email dice servers are employed.
 
They are completely free, you need only to register.
 
Every player receives an e-mail containing a die roll and in this e-mail the player who requested the roll specifies in the "subject" field of the e-mail the event that is related to the roll (forage in moscow, second round of battle of Vienna, pursuit roll for battle of paris, interception in the Area 321, etc.etc.).
 
With a simple glance to the summary of e-mails received in the last turn, every player can check if the opponent has changed his mind, reloaded, and "left his fleets in port", in other words, if he has CHEATED.
 
Surely this method can be limited to major land (or naval) battles only avoiding the "one hundred e-mails" effect, importing the results generated by the die rolls enabling the "third party combat system" option, already existent in the game. 
 
It is just an idea.
 
Anyway, I'd pray Marshall and Matrix to solve this matter that is vital for the vast majority of players.
 
Personally I'm happy to play this game also if cheating is possible, but I'm happier if it is not.
 
Homo homini lupus (Thomas Hobbes).
 
 
 
 

[zenmaster]

Considering the game does not currently have sufficient AI to be a strong single player game and is really only usable via PBEM, I think fixing easy cheating is critical.
Furthermore, most games will be played with unknown persons.
Hence, there will not be any inherent level of trust.
 
Nothing against anyone on this forum, But I have no reason to trust anyone here. 
I have not yet tried the PBEM while I was busy learning interface and reaquanting myself to the game after many years away.
 
However, realizing that cheating is so easy, I really have no interest in playing PBEM.
 

[isandlwana]

The solutions all noted above seem pretty cumbersome.  I agree with some of the posters we need a built in dice server or other seamless solution that allows only a single e-mail in a verifiable form to respond to chit selection etc prior to chance events/encounters.  This seems to me to be an issue which should be addressed by programmers for the 1.02 patch...No?  [&o]