I have been a big EiA fan since 1986, when I first purchased the game and in the past I have spent untold evening playing EiA. I have been reading the forums and the constructive criticism of the game. I am not a programmer, just an accountant, so I know next to nothing about designing a computer game. There are bugs and the Matrix team is addressing those bugs. As with anything we all think that we can make something a little better etc.
Currently, I own the Matrix EiA game and I am in several games that are at different stages. I did not buy EiA to play against the AI, I bought the game to play against other human opponents. As with any game there is the chance of hacking, exploiting, etc. But I am involved in a PBEM game were one player accused another player of cheating and posted this argument on the AAR report section. A gentlemen that helped to design Matrix’s EiA posted on the AAR that it is possible for an individual to use a copy and paste function to replay a turn before you send that turn to others. I found this to be very troubling because it can be easily done I guess.
This is a game that many of us will spend our valuable time playing and plotting, but now to find out that the game results can be manipulated in such an easy manner has greatly decreased my desire to play. People will post on here that you need to have more trust and that people sometimes find ways to exploit in any game. From an accounting perspective, internal controls are put into place to decrease risk and to, eventually catch foul play.
Again, not being a game designer, I do not understand why a set of randomly generated numbers cannot be generated and thus fixed for that phase so that the same results would still happen no matter how many times the copy and paste function is used and in what order the action takes place. Assign pre-generated die rolls to each country or each counter, etc. (this is not a debate about whether or not a set of fixed random numbers, etc. would work) Perhaps this is not feasible, but I find it hard to believe that something can’t be done to stop this potential exploit.
EiA is a game that I am passionate about and I am strongly debating on whether or not to continue playing myself, because this is a time investment for me, an enjoyment of the game, but to know that an individual has the ability to so easily manipulate the results, I am questioning my desire to continue playing.
I really hope that this issue can be solved in a manner that will not detract from the game and in no way should this be taken as a way to decrease sales or hurt EiA activity. I would love nothing more than to be able to find opponents and play this game for years to come! It might not even be Matrix’s duty to put internal controls into place to stop manipulation. Instead it might be my burden to find seven players that I trust, so that I can play this game at a level in which I do not need to worry if there is manipulation happening or not wihout me using a stats program to see if somebody's result are falling outside the bell curve.
Thomas Whitfield
The possibilty of manipulation
Moderator: MOD_EIA
RE: The possibilty of manipulation
Chit selection and variable battle order.
That is, yes -- it would be technically possible for EiANW to incorporate a custom pseudorandom number generator (if they decided not to rely on a stock one) and to preserve the state in the save file. A generator + state results in a replicable stream of numbers, provided that there are no arithmetic inconsistencies (like cross-platform issues when approximating floating-point computations).
However, that would still leave
(a) the ability to reload a file and choose different chits, and
(b) the ability to use the same numbers in different battles, whenever one has multiple battles per turn, by going through them in a different order
(c) the ability to introduce different die-roll requirements by doing/not doing things like guard commitment and reinforcement attempts
(a) might be most important. If an attacker can reload a file and choose different chits, then even with a frozen die roll stream he can repeat until he reaches an optimal case (for that set of rolls) -- minimizing losses, at worst. If one were willing to increase the number of e-mail exchanges, then one could minimize this problem by requiring a file transfer between decision / resolution. For instance, attacker would not immediately resolve the battle after choosing a chit, but would send to defender. If the defender then chose a chit (instead of having it selected prior to attack, or AI doing it for him), he'd have to send it to the attacker before seeing the attacker's selection. The same holds for guard commitment and reinforcement attempts.
Even taking chit selection out of the hands of players AND having a single frozen PRNG stream still leaves it to be exploitable, if he can run and re-run battles in a different order.
(b) That might suggest basing the PRNG state on the particular battle -- perhaps computing a PRNG seed based on a global seed specific to the game, somehow blended with such things as location, date, troop counts, et al -- so that one cannot exploit the sequence of rolls by resolving battles in a different order. Provided that the PRNG is not particularly bad, and the factors going into the battles are well-selected so as to be difficult to replicate. ex. picking bits out of a cryptographic hash / digital signature of a string including various factors for a seed.
...
Considered an approach in which chit files et al are sent to a third party for execution who then computes results but does not see them, but that still leaves room for exploits; he can send the results to just one side and re-execute if they weren't what was desired. So the PRNG state would still need to be preserved, and if the "third party" conspired with the second player to select chits, that selection could be changed.
That is, yes -- it would be technically possible for EiANW to incorporate a custom pseudorandom number generator (if they decided not to rely on a stock one) and to preserve the state in the save file. A generator + state results in a replicable stream of numbers, provided that there are no arithmetic inconsistencies (like cross-platform issues when approximating floating-point computations).
However, that would still leave
(a) the ability to reload a file and choose different chits, and
(b) the ability to use the same numbers in different battles, whenever one has multiple battles per turn, by going through them in a different order
(c) the ability to introduce different die-roll requirements by doing/not doing things like guard commitment and reinforcement attempts
(a) might be most important. If an attacker can reload a file and choose different chits, then even with a frozen die roll stream he can repeat until he reaches an optimal case (for that set of rolls) -- minimizing losses, at worst. If one were willing to increase the number of e-mail exchanges, then one could minimize this problem by requiring a file transfer between decision / resolution. For instance, attacker would not immediately resolve the battle after choosing a chit, but would send to defender. If the defender then chose a chit (instead of having it selected prior to attack, or AI doing it for him), he'd have to send it to the attacker before seeing the attacker's selection. The same holds for guard commitment and reinforcement attempts.
Even taking chit selection out of the hands of players AND having a single frozen PRNG stream still leaves it to be exploitable, if he can run and re-run battles in a different order.
(b) That might suggest basing the PRNG state on the particular battle -- perhaps computing a PRNG seed based on a global seed specific to the game, somehow blended with such things as location, date, troop counts, et al -- so that one cannot exploit the sequence of rolls by resolving battles in a different order. Provided that the PRNG is not particularly bad, and the factors going into the battles are well-selected so as to be difficult to replicate. ex. picking bits out of a cryptographic hash / digital signature of a string including various factors for a seed.
...
Considered an approach in which chit files et al are sent to a third party for execution who then computes results but does not see them, but that still leaves room for exploits; he can send the results to just one side and re-execute if they weren't what was desired. So the PRNG state would still need to be preserved, and if the "third party" conspired with the second player to select chits, that selection could be changed.
--
Not a grognard.
Not an optimizer. It's a game to me, not a job.
Not a grognard.
Not an optimizer. It's a game to me, not a job.
RE: The possibilty of manipulation
Huh? [&:] Too techy 4 me.[:'(]
-
Soapy Frog
- Posts: 282
- Joined: Sat Jul 16, 2005 12:33 am
RE: The possibilty of manipulation
Using encryption, chit choices for a given land combat phase could be held, encrypted, by the host until all chit choices are received and then executed. This would require a seperate phase each land combat segment for chit selection.
Simply put; the current system is utterly broken from a "security" standpoint. There should be NO WAY that I can reload the game until I pick the right chit or get the right sequence of die rolls.
It could be done with a PBEM format but you do need to have a trusted host which cannot be tampered with. I don't think much thought was given to this in the current design.
Simply put; the current system is utterly broken from a "security" standpoint. There should be NO WAY that I can reload the game until I pick the right chit or get the right sequence of die rolls.
It could be done with a PBEM format but you do need to have a trusted host which cannot be tampered with. I don't think much thought was given to this in the current design.
RE: The possibilty of manipulation
Hm, 'k.
PRNG = pseudorandom number generator. Combines a state -- a number or set of numbers, chosen however -- and a method for using that state to generate supposedly random numbers, plus the next state. If the method is designed well, the numbers are difficult to predict (without knowing the algorithm and state).
So if you keep a single PRNG and save the state, then you can request a series of supposedly random numbers and get the same ones in the same order.
However, if you just use a single stream amongst all battles, then the player can reload the battles in a different order. If, for instance, he knows he'll get a certain sequence of rolls, and he knows that as a consequence he'll roll badly in the first battle he resolves and well in the second, he could choose a minor battle for the first and resolve the more important battle second. Or he might resolve a siege attempt at a different time.
Likewise, if the system let you decide whether or not to forage, then start resolving battles before you committed to your decisions by sending files to the next player -- then you could check your die rolls. High die rolls might cause you to prefer depot supply and use the rolls in battle.
Hence, to cut down such exploits -- you'd want to rig the PRNG in such a way that a player can't get different results just by reloading a file over and over again (so repeatability is good) but also so he can't get different results just by making a slightly different decision that alters how the same numbers are used and repeating until he's happy. So either the sequence of random numbers should be specific to each situation, so that no matter what he does he can't change that; or you prevent him from doing 'try this and see what happens' repeatedly by making him commit the first time (by sending the files to a hostile player with no reason to let him 'do over', before he can see the results of his choices).
The same holds for such things as foraging rolls; if you're going to forage in many places, and you know you're going to roll badly sometimes, you might have an incentive to forage in a different order if doing so lets you distribute losses in a more preferable manner (or have fewer losses, due to varying forage levels and modifiers).
PRNG = pseudorandom number generator. Combines a state -- a number or set of numbers, chosen however -- and a method for using that state to generate supposedly random numbers, plus the next state. If the method is designed well, the numbers are difficult to predict (without knowing the algorithm and state).
So if you keep a single PRNG and save the state, then you can request a series of supposedly random numbers and get the same ones in the same order.
However, if you just use a single stream amongst all battles, then the player can reload the battles in a different order. If, for instance, he knows he'll get a certain sequence of rolls, and he knows that as a consequence he'll roll badly in the first battle he resolves and well in the second, he could choose a minor battle for the first and resolve the more important battle second. Or he might resolve a siege attempt at a different time.
Likewise, if the system let you decide whether or not to forage, then start resolving battles before you committed to your decisions by sending files to the next player -- then you could check your die rolls. High die rolls might cause you to prefer depot supply and use the rolls in battle.
Hence, to cut down such exploits -- you'd want to rig the PRNG in such a way that a player can't get different results just by reloading a file over and over again (so repeatability is good) but also so he can't get different results just by making a slightly different decision that alters how the same numbers are used and repeating until he's happy. So either the sequence of random numbers should be specific to each situation, so that no matter what he does he can't change that; or you prevent him from doing 'try this and see what happens' repeatedly by making him commit the first time (by sending the files to a hostile player with no reason to let him 'do over', before he can see the results of his choices).
The same holds for such things as foraging rolls; if you're going to forage in many places, and you know you're going to roll badly sometimes, you might have an incentive to forage in a different order if doing so lets you distribute losses in a more preferable manner (or have fewer losses, due to varying forage levels and modifiers).
--
Not a grognard.
Not an optimizer. It's a game to me, not a job.
Not a grognard.
Not an optimizer. It's a game to me, not a job.
RE: The possibilty of manipulation
I would agree. However, before I spend any time on the corrections available for the problem, I would like someone to verify that it actually exists.
At LAST! The greatest campaign board game of all time is finally available for the PC. Can my old heart stand the strain?
RE: The possibilty of manipulation
I recently posted a suggested solution to this in a thread on the tech-support subforum:
http://www.matrixgames.com/forums/tm.asp?m=1676424&mpage=1�
Please tell me what you think about that? (see no reason to duplicate)
http://www.matrixgames.com/forums/tm.asp?m=1676424&mpage=1�
Please tell me what you think about that? (see no reason to duplicate)
Guy
RE: The possibilty of manipulation
ORIGINAL: trw2264
I have been a big EiA fan since 1986, when I first purchased the game and in the past I have spent untold evening playing EiA. I have been reading the forums and the constructive criticism of the game. I am not a programmer, just an accountant, so I know next to nothing about designing a computer game. There are bugs and the Matrix team is addressing those bugs. As with anything we all think that we can make something a little better etc.
Currently, I own the Matrix EiA game and I am in several games that are at different stages. I did not buy EiA to play against the AI, I bought the game to play against other human opponents. As with any game there is the chance of hacking, exploiting, etc. But I am involved in a PBEM game were one player accused another player of cheating and posted this argument on the AAR report section. A gentlemen that helped to design Matrix’s EiA posted on the AAR that it is possible for an individual to use a copy and paste function to replay a turn before you send that turn to others. I found this to be very troubling because it can be easily done I guess.
This is a game that many of us will spend our valuable time playing and plotting, but now to find out that the game results can be manipulated in such an easy manner has greatly decreased my desire to play. People will post on here that you need to have more trust and that people sometimes find ways to exploit in any game. From an accounting perspective, internal controls are put into place to decrease risk and to, eventually catch foul play.
Again, not being a game designer, I do not understand why a set of randomly generated numbers cannot be generated and thus fixed for that phase so that the same results would still happen no matter how many times the copy and paste function is used and in what order the action takes place. Assign pre-generated die rolls to each country or each counter, etc. (this is not a debate about whether or not a set of fixed random numbers, etc. would work) Perhaps this is not feasible, but I find it hard to believe that something can’t be done to stop this potential exploit.
EiA is a game that I am passionate about and I am strongly debating on whether or not to continue playing myself, because this is a time investment for me, an enjoyment of the game, but to know that an individual has the ability to so easily manipulate the results, I am questioning my desire to continue playing.
I really hope that this issue can be solved in a manner that will not detract from the game and in no way should this be taken as a way to decrease sales or hurt EiA activity. I would love nothing more than to be able to find opponents and play this game for years to come! It might not even be Matrix’s duty to put internal controls into place to stop manipulation. Instead it might be my burden to find seven players that I trust, so that I can play this game at a level in which I do not need to worry if there is manipulation happening or not wihout me using a stats program to see if somebody's result are falling outside the bell curve.
Thomas Whitfield
Thomas,
I hear you, but there really is no way to prevent the unscrupulous enterprising character from cheating in PBEM games (Grognot pretty well covers it). The power of the copy paste function eradicates all in-house efforts (pre-chosen random number events, read writes to third party files, etc.) Obstacles can only curtail the action not prevent it.
Anyway, same offer that I already made in another post . . .
For PBEM games without Quick Combat option enabled, if the playgroup wants a neutral (non-playing) arbitrator to settle all or some of their battles for them then the arbitrator will need to have the following files and player instructions in order to perform the action.
1. The attacker’s saved game files (with password) immediately after picking his chit
2. The battle file (from the attacker)
3. The defender’s saved game files (with password) and chit pick instruction
4. Instructions for casualty preferences (including pursuit) from both players
5. Reinforcement and guard commitment instructions from both players (if applicable).
I’d be happy to arbitrate battles if needed. Just PM me.
Richard
RE: The possibilty of manipulation
Or, you have the host resolve it, and that cuts it down to only the host being able to cheat, and really, if you have a host who cheats, go find someone else...
-
yammahoper
- Posts: 231
- Joined: Fri Apr 23, 2004 7:14 pm
RE: The possibilty of manipulation
Wow. And I thought cheating in EiA was slipping extra garrison counters onto the map or trying to "drop roll" a specific number on the die.
Times do change.
yamma
Times do change.
yamma
...nothing is more chaotic than a battle won...
RE: The possibilty of manipulation
Thanks Richard and I know that it is impossible to balance internal controls with playability. I am actually typing up something now that I would like to try and I will be posting it to Opponents Wanted section. I am going to try and get a principles-based game or two going. I have received a lot of PM's and private emails about this post in the past couple of hours, so I get the impression that there is a desire for fairness or at least some comfort level of fairness for a game that people are passionate about and realize the time commitment that is involved.
After I completed the post, I went back through the forums and saw that others have had concerns as well, but to be honest I never even thought about manipulation or ways to manipulate this game until it was questioned in a current game that I am in. I can see in your replies that you have been upfront in your posts on this matter and that is appriciated, especially your offer of arbitration.
After I completed the post, I went back through the forums and saw that others have had concerns as well, but to be honest I never even thought about manipulation or ways to manipulate this game until it was questioned in a current game that I am in. I can see in your replies that you have been upfront in your posts on this matter and that is appriciated, especially your offer of arbitration.
RE: The possibilty of manipulation
You could have the die roll emailed to all participants of the PBEM. If all players enter their email address at the start of the game, then everytime the dice is rolled, an email is sent to all players indicating the reason for the die roll, and the result of the die roll. This way, it does not matter if you reload or copy/paste.
just my 2 cents.
just my 2 cents.
RE: The possibilty of manipulation
If the game were mass-market (*), there might also be people trying to decode the save files to see whether they can bypass the fog of war, or even to edit them. *shrug* I don't know whether the files contain readily-accessible information that players shouldn't have, or whether they're protected against alteration with checksums and so forth. It's an old, old problem in MP game design.
Yup, trusted third-party arbitrators are helpful.
(*) Or even not. I do recall that there was somebody who figured out how to edit his resources in "Dominions II" PBEM games (before eventual discovery and fix), and that's a fairly obscure game.
Yup, trusted third-party arbitrators are helpful.
(*) Or even not. I do recall that there was somebody who figured out how to edit his resources in "Dominions II" PBEM games (before eventual discovery and fix), and that's a fairly obscure game.
--
Not a grognard.
Not an optimizer. It's a game to me, not a job.
Not a grognard.
Not an optimizer. It's a game to me, not a job.
RE: The possibilty of manipulation
The problem with this is that my email program is under MY control. If I am trying to
cheat; I simply shut OFF my internet connection & all outbound email sits in the "out basket". If I reload the files; I also delete the outgoing email. Sorry.
cheat; I simply shut OFF my internet connection & all outbound email sits in the "out basket". If I reload the files; I also delete the outgoing email. Sorry.
Guy
RE: The possibilty of manipulation
Perhaps the phasing player can send an email to the Host or a non-playing neutral party with his chit selection before he sends the file to the defender. This email can then always be checked if need be. This assumes the attacker has to pick his chit first. I have not go that far in a PBEM so I could be way off.
-
Soapy Frog
- Posts: 282
- Joined: Sat Jul 16, 2005 12:33 am
RE: The possibilty of manipulation
The host should resolve everything, after chit choices are recieved independently form participants. Communications should be encrypted. The host should have no control over the "order of operations" i.e. the order in which combats are resled. This means the host could do reloads to try and get better die rolls, but if this is combined with PRNG of sufficient sophistication then even that would be impossible.
In short it could be made WAY HARDER to cheat.
Did I ever mention that a TCP/IP CLient/Server solution would be far superior to PBEM?
In short it could be made WAY HARDER to cheat.
Did I ever mention that a TCP/IP CLient/Server solution would be far superior to PBEM?
RE: The possibilty of manipulation
ORIGINAL: gwheelock
The problem with this is that my email program is under MY control. If I am trying to
cheat; I simply shut OFF my internet connection & all outbound email sits in the "out basket". If I reload the files; I also delete the outgoing email. Sorry.
True, but in this case the game is its own email software. The player would have no control over the email, and so should be prohibited to delete them. I do not believe that the email has to go through your normal email browser. The program can save the copy itself and send it once it detects an internet connection.
If the player were to reload the other players would be getting many different die roll for the same event and would become suspicious. I am not a programer but this does not seem to be too difficult to implement.
RE: The possibilty of manipulation
ORIGINAL: lavisj
ORIGINAL: gwheelock
The problem with this is that my email program is under MY control. If I am trying to
cheat; I simply shut OFF my internet connection & all outbound email sits in the "out basket". If I reload the files; I also delete the outgoing email. Sorry.
True, but in this case the game is its own email software. The player would have no control over the email, and so should be prohibited to delete them. I do not believe that the email has to go through your normal email browser. The program can save the copy itself and send it once it detects an internet connection.
If the player were to reload the other players would be getting many different die roll for the same event and would become suspicious. I am not a programer but this does not seem to be too difficult to implement.
Actually; the game does NOT have its own email & DOES go thru your normal email browser.
(You can see this because there is no provision to setup the SMTP/POP3 email connection in the game itself)
It would be possible to ADD an email section (they are fairly "out-of-the-box" plugins); but then if I still
wanted to cheat; I would simply set up my own Exchange Server; route the program's connection to THAT
(so that the email appears "UP" to to program); but not let the Exchange forward it & kill the extra stuff there
(Microsoft Exchange & other freeware equivalents are also "out-of-the-box" apps [:(])
(And I AM a programmer[8D])
Guy
-
Grapeshot Bob
- Posts: 222
- Joined: Sun Dec 16, 2007 8:35 pm
- Location: Canada
RE: The possibilty of manipulation
Hey, I was just thinking ...
Would it really be so bad to have everyone "optimize" their turns?
Maybe you could have "optimize" PBEM games where everyone agrees to play "optimized".
GSB
Would it really be so bad to have everyone "optimize" their turns?
Maybe you could have "optimize" PBEM games where everyone agrees to play "optimized".
GSB
RE: The possibilty of manipulation
Hi, new to this. Im bit confused.
Question, combat phase if quick combat disbabled ,
is combat not resolved by the host ? Sending the combetants resoults after each round back and forth ?
Or is it just the defender and the curent-phase player exchanging files ? Things should run thorugh the host to avoid messy stuff.
This is cryptic.
Now i dont remember if zip still can have password, but if, then one option players could include is, when at combat phase.
The phasing players involved in combats, send a zipped(with password 2 letters 6 digits, offcourse i can still be cracked, not sure about unzip pgms to...) text file with chit selection, to the host. Or to the defender, before combat is resolved. Then come combat they "actually" exchange the combat files. And the password is released. So that they can compare the text file chit selection to the textfile send. Atleast this way, phasing player cant chose chit acording to the defenders chit.
Regards
Bresh
Regards
Bresh
Question, combat phase if quick combat disbabled ,
is combat not resolved by the host ? Sending the combetants resoults after each round back and forth ?
Or is it just the defender and the curent-phase player exchanging files ? Things should run thorugh the host to avoid messy stuff.
This is cryptic.
Now i dont remember if zip still can have password, but if, then one option players could include is, when at combat phase.
The phasing players involved in combats, send a zipped(with password 2 letters 6 digits, offcourse i can still be cracked, not sure about unzip pgms to...) text file with chit selection, to the host. Or to the defender, before combat is resolved. Then come combat they "actually" exchange the combat files. And the password is released. So that they can compare the text file chit selection to the textfile send. Atleast this way, phasing player cant chose chit acording to the defenders chit.
Regards
Bresh
Regards
Bresh