As you are all aware the game has been paused for some time now. The reason was talk of "impossible" RNG results. The game admin, the moderator-team and our contacts with the creators of the game at Firaxis performed an investigation. Here are the results:


While the results are unusual, there is no evidence that can prove cheating.
Restrictions on the accused player or team are not required.
To remove any future doubt on RNG results we're adding a few new rules:

1 - The bulk of the turn must be played in one sitting, or at least in a way that moves more than one unit at a time.
Getting out to discuss what to do next when battle does not go according to plan, is OK, in and out for each unit is not.

2- Battles must be separated from other actions. Meaning: When you've started a battle, you must continue with doing battle and cannot do other things that may influence the RNG like diplomacy, hut popping, spying etc.
Basically your turn becomes a three phased sequence -> pre combat phase - combat phase - post combat phase. Move your units into position and decide a battle order is the "pre combat phase". If you want to do scouting, city management, spy operations, diplomacy, hut popping etc. before battle, this is the time. The combat phase is the phase where all actions allowed are selecting a unit and attack an enemy unit. Get a result and select the next unit to attack the next enemy unit. Rinse and repeat till you do not want to do combat any more that turn. When ending this phase, no more battles are allowed. The post combat phase is where you do after battle movement, diplomacy etc.

There is an exception; if you defeat an enemy unit (outside or inside a city) and want to attack another unit (outside or inside a city) that became available because you've won a battle, unit movement(s) for a combat-goal are allowed. Diplomacy, city management etc. are not allowed.
.
We've installed a program that detects RNG tinkering. This is done not to accuse a player or a team, but to remove any grounds for future accusations. It will prevent a theoretical possibility to influence RNG-outcome to a certain degree.


Please continue the game.
Any discussions may take place in this thread, but no (more) accusations.

This seems reasonable. Clarification question:

Can you a unit with multiple moves be left far away from a target? two scenarios come to mind:

1) A stack is attacking another stack. There is a horseman 2 tiles away - can he move and attack?
2) A team is defending against an invasion. After combat has started, can a rifle move 6 tiles on a rail and attack?

I think the answer to these questions is yes, but I wanted to be sure.

Basically your turn becomes a three phased sequence -> pre combat phase - combat phase - post combat phase. Move your units into position and decide a battle order is the "pre combat phase". If you want to do scouting, city management, spy operations, diplomacy, hut popping etc. before battle, this is the time. The combat phase is the phase where all actions allowed are selecting a unit and attack an enemy unit. Get a result and select the next unit to attack the next enemy unit. Rinse and repeat till you do not want to do combat any more that turn. When ending this phase, no more battles are allowed. The post combat phase is where you do after battle movement, diplomacy etc

Comment on bolded part: sometimes you need to use workers, settlers and occasionally great people during what you would call the combat phase, for tactical reasons in the battle (ie roading and settling mainly), which could be planned prior to the battle, but the usage of the none combat units is dependent entirely on the outcome of a battle.

Is the rule as described now stating that such a move is illegal?

This seems reasonable. Clarification question:

Can you a unit with multiple moves be left far away from a target? two scenarios come to mind:

1) A stack is attacking another stack. There is a horseman 2 tiles away - can he move and attack?
2) A team is defending against an invasion. After combat has started, can a rifle move 6 tiles on a rail and attack?

I think the answer to these questions is yes, but I wanted to be sure.
1) Sure, that is an "unexpected combat phase" for the defense.
2) Sure, that is also an "unexpected combat phase" for the defense.

The spirit of that rule is create a distinct "combat phase" instead of having many combat activities mixed up with RNG-altering activities. The spirit should be clear and firmly applied. In the reality of in-game happenings there should be no reason to refrain from incidentally moving units or even incidentally do a combat action as a response to a direct threat. This rule is not to limit your options in the game, but to minimize RNG-influences.

Of course, in your planning (and playing) of the turn you should seperate pre-combat phase, (planned) combat phase, and post combat phase and stick to it.

Comment on bolded part: sometimes you need to use workers, settlers and occasionally great people during what you would call the combat phase, for tactical reasons in the battle (ie roading and settling mainly), which could be planned prior to the battle, but the usage of the none combat units is dependent entirely on the outcome of a battle.

Is the rule as described now stating that such a move is illegal?No, those worker actions or settler actions are combat-oriented and are not illegal. Even spy missions may be combat-oriented and not illegal during the combat-phase. As long as you've planned them before playing the turn to perform those actions in the combat phase. If planned as combat-activities they will not (illegally) influence the RNG, which is the goal of the rule(s).

Just refrain from doing non-combat oriented worker, settler or spy, (or whichever unit I've forgotten) actions during your combat phase.

Thanks for the clarifications there.

2- Battles must be separated from other actions. Meaning: When you've started a battle, you must continue with doing battle and cannot do other things that may influence the RNG like diplomacy, hut popping, spying etc.

Was this sort of thing happening?

Was this sort of thing happening?We have found no evidence that this happened.

In our investigation we learned about theoretical ways someone can influence the RNG outcome. Most of the ways to do that require a very deep level of understanding both the cIV game and computer technology. It is unlikely that this has happened, but it is theoretically possible.
The new rules and the RNG-tinkering-detection software that is installed on the server now are designed to eliminate the possibilities of RNG-tinkering and to detect it if it happens.

Since new random seed on reload is not clicked for this game, do you have any way of knowing if someone loaded a save (by cracking admin password) played the battles and depending on results took it to real game?

Not saying that Sancta used this method BTW.

You don’t need the admin PW, you can load every save.

To reiterate IS, can it be detected that the game was saved?

If rng differences can be detected, then I don't think it's really an issue in any case, and you'd think that they were taking care of the easier solutions before tackling the hard ones.

I'm content with their solution, it looks like they really tried their best to address every team's concerns, and put a significent amount of work into getting a much more foolproof method of detecting cheating.

Of course... just asking out of curiosity for other multiplayer games.

There's no mention of this. Only the loosely enforceable recommendation to perform the turn in one sitting (as there are countless reasons to break this rule). On the day in question, there was a log in and exit with the completion of the turn some 21 hours later. Also some diplomacy occured mid battle, something now stated as against the rules. It seems to me like all their effort was put towards detecting manipulation of the RNG, when, it seems as though one can produce desired results without such manipulation.

Of course, this would be a clear violation of rule 1.5, although it appears that rule went largely neglected by some teams before it was put on the books.

Is it possible is all I ask. I know everything's really hard to prove but is it possible?

On the day in question, there was a log in and exit with the completion of the turn some 21 hours later. Also some diplomacy occured mid battle, something now stated as against the rules. It seems to me like all their effort was put towards detecting manipulation of the RNG,when, it seems as though one can produce desired results without such manipulation.

Of course, this would be a clear violation of rule 1.5, although it appears that rule went largely neglected by some teams before it was put on the books.

Is it possible is all I ask. I know everything's really hard to prove but is it possible?

:salute:

I will now walk away from this game.

I sat through and patiently wait while the mods did thier investigation, on the legitimacy of SANCTA's (but me being the only turn player, my word).

The Mods came to their conclusion, but it doesn't seem to be over. There is no point playing a game, when it starts to feel like a chore, and the other players continue to accuse you of being a cheater. As a professional I found the original accusion insulting , but undertood the reason for the due process to take place.

I will log in to pause the game on SANCTA's turn while they find a new turn player.

Congradulations Cav you just won the game. :clap:

What is your point? Where were you when all investigation was happening? Why did you choose to go through Sulla, when you know that you are going accuse again, if you don't like the verdict?


Is it possible is all I ask. I know everything's really hard to prove but is it possible?

Is it possible is all I ask.
Several of us know the answer to this, but I for one don't like the idea of what some people might do with the answer. However it does not appear that any worse damage can be done at this point... :(

Yes, it is possible to obtain an optimum result in that way, with the game settings we're currently playing with. In particular with preserve random seed and sequential turns.

Both of the alternatives have their own problems. If simultaneous turns were used, then there could be abuse of turn order and it would be impossible to replay exactly if a crash recovery were necessary. New random seed allows someone to crash the pitboss on purpose if they get a result they don't like, and on the reload from autosave a new random number would guarantee a different result. There are enough network glitches to make either one a significant risk and difficult to detect.

In the end, I think we should either trust and respect the people we play with, or leave it to those who prefer an honorable game. Personally I would need a reasonable expectation that proof was possible before even making an accusation, and if told there is no proof I would drop the matter, not keep picking at it until the accused gets annoyed and leaves.

My question was the earlier one, was is possible to see if the game had been saved. To me, it was unclear if this avenue had been pursued. We got to sit back and wait three weeks and all that we see are the words at the top of this thread. This is the time for discussion. I want the details.

I did a test on a different pitboss game I'm playing in. I saved the game to my hard drive, tried to load the game without being connected, and got a screen asking for the "ADMIN PASSWORD", which I don't have. The game would not load without it, and defaulted to LAN set up screen I believe.

I personally am not a "tech guy", I just play the games, but it seems to me that it would require a great level of sophistication, and LOTS OF TIME, to get any useful data out of a save that can not be opened without a password. I for one, am not that sophisticated, and have nowhere near that kind of time to waste. I personally believe that Memphus, like me, is way to busy to have spent that amount of time trying cheat at a flippin game. :(

My question was the earlier one, was is possible to see if the game had been saved. To me, it was unclear if this avenue had been pursued. We got to sit back and wait three weeks and all that we see are the words at the top of this thread. This is the time for discussion. I want the details.

You're accusing Memphus of cheating a couple of days after the game has been unpaused because there was no evidence of cheating.

You're just being petty, and destroying the game.

So then he's lucky? It was all some good-natured luck?

Unless you can provide evidence suggesting otherwise, yes, it was luck.

All I got for evidence are the "impossible" results and maybe the plako declare (still don't get that one), but nothing's holding up in court

Unlikely results aren't proof of cheating, nor is an accidental war declaration.

Nope, proof won't exist in this leaky game, and so all that's left is speculation. I would like to think that how unlikely the results are would start to head in the direction of proof, but let's not leave that up to me.

Nope, proof won't exist in this leaky game, and so all that's left is speculation. I would like to think that how unlikely the results are would start to head in the direction of proof, but let's not leave that up to me.

It doesn't matter how unlikely the results are. It's not proof of any tinkering...

If this were Casino then you might get fired for letting three machines hit the jackpot in twenty minutes. But that's in the billions, we're just talking about 10 to the 6th

I just went ingame and saved a file to my computer. I was able to open it without an admin code.

My question was the earlier one, was is possible to see if the game had been saved.
Every time you enter a pitboss game, a save is created on your computer. You just have to know where to look. Besides that, there is the save from menu or ctrl-s, and no way to track that.

I did a test on a different pitboss game I'm playing in. I saved the game to my hard drive, tried to load the game without being connected, and got a screen asking for the "ADMIN PASSWORD", which I don't have. The game would not load without it, and defaulted to LAN set up screen I believe.

There is a technique for loading a saved pitboss game without the admin pw. I believe it has to be sequential turns and during your own turn in order for it to work. There was a discussion about this during the setup phase because teams wanted to be able to go back to previous turns and grab screenshots for comparisons.

we're just talking about 10 to the 6th
Let's apply a little math to this. Suppose there are 10 battles and you need to find the correct order to get optimum results. There are 3,628,800 possible orders of those 10 battles. To successfully find the correct sequence, you'd need to load and try 151,200 battle orders per hour, or 2,520 battle orders per minute. On my system it takes a minute to get out of a save and get back in. Clearly it is impossible to find the perfect move order within the time limit, and that's for a relatively short battle sequence.

Brute force obviously won't work -- so what about trial and error? Suppose out of 10 battles on the 1st iteration, there are 3 losses in positions 3, 6, and 9. To avoid a loss at step 3, one has to try up to 648 different combinations of moves 1-3. Even then, if a winning 1-3 is found it might have changed the results of #4 and #5, which then take up to 30 attempts -- if you don't have to backtrack looking for another order of 1-3 which allow a win on #4/5. Repeat for repairs to the sequence 1-8 to make #9 work, etc. Suffice it to say that even being lucky on finding alternatives the number of combinations of permutations will run into the 1000's quickly. On my system, being able to manage an average of 30-60 attempts per hour, 1000 attempts would take between 16.5 and 33 hours.

Sounds like a good Mythbusters spot. Myth: you can get perfect battle results in a turn-based strategy game if you're willing to try enough move orders. Answer: busted, for battles having more than (let's guess 6) battles in the sequence in games with a time limit of 24 hours, because for sequences longer than that you won't have any time left to eat drink or sleep.

Let's slow down here. It's not random luck to figure out the correct sequence. A bit of human thinking can whittle it down some. Clearly stronger troops would be used earlier in the battle - Despite the factorial availibily of options, no right-minded human would attempt all those options. Use a riflemen here and then these options are available to you. For the turn in question the odds of that result were more like 1 in midway thru 10 to the 5th. Taking in consideration all our defensive battles (three extra) was my 10 to the 6th quote. Most of the battle were not that unlikely to win, it was just the cumulative affect of winning all those battles that was so unlikely, thus there were many, many options available to them in their factorial number of outcomes.

Regardless, it seems we need to find a way to prevent this from happening in the future (loading the goddamm save and exploring such options) and again,
IS THERE A WAY OF KNOWING IF THIS HAS HAPPENED IN THE PAST?

Regardless, it seems we need to find a way to prevent this from happening in the future (loading the goddamm save and exploring such options) and again,
IS THERE A WAY OF KNOWING IF THIS HAS HAPPENED IN THE PAST?

I think the answer is: no!
Nobody can say, what I (or someone) do at my comp.
In case of cheating, there is allways asuspicion in my mind. But in normal SP i 've so many bad or good luck results, and the forum is full of posts "the AI cheats", so I can never be sure about it in MP. If the battle sequence is reasonable then I swallow my annoyance, if not I lose the fun in the game. Additional there is the mistake in the official/ingame odds.

Can it be detected that a game was saved from a pitboss application to another computer? If this can be answered yes (and proven this did not occur) then my point is shattered

We've been challenging whether it was reasonable. Our position is no. Statitically, outcomes have deviated too far from expected bell curve results.

I can't continue to debate this in a reasonable manner, so I'm going to spend my time on more productive things.

First let me say I am sorry as I feel responsible for reviving this...it was not my intension to accuse anyone and I don't think I did, all I wanted to know is whether it is posible to detect if someone took a save from the pitboss game (for future personnal use in my pitboss games)....quite naive of me as I forgot about the autosave.

I totally agree that this is a matter of faith on opponent's sportsmanship and you take it or leave it.

Sancta had some extraordinarry luck, not only against Cav but against Kazakstan as well earlier in the game, winning consecutive very low odds...however this is just luck and it is within the game. They were taking the piss out of their luck however with ironic posts and that escalated feelings abit.

Personnaly I don't think they cheated and I know they played a dam good game so you can say luck favors the good.

The rest is just pointless and ugly.

If this were Casino then you might get fired for letting three machines hit the jackpot in twenty minutes. But that's in the billions, we're just talking about 10 to the 6th


10^6=1 million

Can it be detected that a game was saved from a pitboss application to another computer?
No, current technology does not have this capability. Plus, the game itself creates a save when you log in, and it's impossible to do anything about that autosave.

I believe it is possible to create a mod for a hypothetical next game that would close most if not all of the holes that have been discussed. If asked, I can explain what needs to be done to someone who has the right setup to actually mod it -- perhaps the BUG/BUFFY team would take it on.

But first let's see if this game lives on.

The best is like in ISDG, with new rng after reload.

Then you can take every save for experiments and can look for the best sequence of battles with the odds. The only problem is, if you must reload for game-reason you get another chance. But there is never the reproach of cheating if you win with low percentage.

So then he's lucky? It was all some good-natured luck?

All I got for evidence are the "impossible" results and maybe the plako declare (still don't get that one), but nothing's holding up in court

Nope, proof won't exist in this leaky game, and so all that's left is speculation. I would like to think that how unlikely the results are would start to head in the direction of proof, but let's not leave that up to me. I have heard about this controversy and am suprised that it still has such legs. Maybe I can help stick a fork in it ...

First of all, the results were not "impossible", since they occured. One could say they were improbable, if we were in fact dealing with independent random events (which we are not, see below). But for the moment, supose we were dealing with an independent random system.

Hopefully, someone has pointed out what independence of events means: the probability of a 10th coin flip being heads after 9 heads in a row is still 50% even though the probability when you start of getting 10 heads in a row is 1 in 2^10.

Also, it is important to understand what we can and cannot conclude using frequentist statistical inference. In reserach, when we say a result (usually expressed as difference between an observed result and an expected result or control result) is "statistically significant", we mean that the probability of observing that difference in our study, if the truth in nature is that there is no difference, is below some ARBITRARY threshold (usually 5%) ... and based on that, we are willing to ASSUME that the difference is not due to random error in our sampling (or in our radomization for randomized trials). We don't PROVE it is not due to chance (random factors). Sometimes our assumption that results are not due to chance are wrong, despite "statistical significance".

But of course, (as I understand RNG's) we are not dealing with random, independent events. We are dealing with a hopefully unpredictable (by players) deterministic system. So a lot of this statistical talk really does not apply here. My sense is that given a particular seed, the algorithim may generate sequences which will potentially contain segments that highly favor the attacker at times, and highly favor the defender at times. So a long string of success at low or even odds is not the "luck" of individual "rolls", but perhaps more the luck of attacking in a RNG segment favorable to the attacker.

In which case, maybe the particular combat order would only make marginal differences to the outcomes? So instead of winning 100% after siege, one wins 80% with various other combat orders? It might well be that similarly "improbable" results would occur in this scenario regardless of attack sequence? That seems to make basing accusations on the alleged improbability of outcomes pretty dangerous: any innocent attack in that situation would generate an improbable result subject to accusation by that criterion. Is that were we want to be?

The best is like in ISDG, with new rng after reload.

Then you can take every save for experiments and can look for the best sequence of battles with the odds. The only problem is, if you must reload for game-reason you get another chance. But there is never the reproach of cheating if you win with low percentage. This comment I don't get. Are you saying you want to be able to pre-play the position before you play the real game turn ??? :confused: :crazyeye: :confused:

dV

In many Pbem-DGs we (at GWT) make (with WB) a similar environment game (not only for battles), set the same units and test it (10 or 20 times). So you can look for the best sequence, because there is in my opinon in most times a favor for the attacker independent of the offical (ingame) chance. With new random seed you can break the battle at any time and discuss the the next moves.
In my opinion: one problem is the reproach and that we are unknown together. If you play cards and you have the idea the other cheats I (you) have possibilities to clear the sitruation, here we have only the hope that the other is fair and this is only a game for fun. But you get the idea of cheats more often, when there is a fix chance to win some turns without combat and then one decided to attack with a good reason (perhaps because the supply is near) and win. Then for the defender there is automatically the question: why now?

:shake: how did it come to this?

sad, sad, sad.

:shake: how did it come to this?

It was just luck, I guess.

:shake: how did it come to this?

sad, sad, sad.

I suppose it all began with some rather unfortunate game settings that left the game open to this sort of loophole.

da Vinci I just don't buy your argument, sorry. There's no fingerprint, video footage or dna sample but the highly improbably results combined with the highly suspicious logging and out repeatedly is like playing in a pub quiz where you're not allowed to use mobile devices and one team gets an improbably high score but throughout there was one guy in the team who had to keep going to the toilet every five minutes and took his iPhone with him. Gimme a break, I wasn't born yesterday.

I suppose it all began with some rather unfortunate game settings that left the game open to this sort of loophole.

da Vinci I just don't buy your argument, sorry. There's no fingerprint, video footage or dna sample but the highly improbably results combined with the highly suspicious logging and out repeatedly is like playing in a pub quiz where you're not allowed to use mobile devices and one team gets an improbably high score but throughout there was one guy in the team who had to keep going to the toilet every five minutes and took his iPhone with him. Gimme a break, I wasn't born yesterday.

This type of post is the problem in a nutshell! It IS PERFECTLY LEGAL to log in and out as often as we want. It IS PERFECTLY LEGAL to TEST battle options as much as we want. After 1 battle goes differently than you expect, you have to reevaluate ALL of your options. This incessant whining that someone is cheating with ABSOLUTELY NO EVIDENCE is exactly why we ended up here. :(