Are you, in fact, a pregnant lady who lives in the apartment next door to Superdeath's parents? - Commodore

Create an account  

 
RB Real-Life RPG #1 [Inc. RB Integrity-Game #1 - Krill+Memphus vs SANCTA RNG.]

scooter Wrote:For those of us who have no idea what you are talking about, can you give us a brief version of what this is about? I tried reading your post and it was so bizarre that I gave up after about 6 or 7 minutes.

Well, I do agree with all of that. I hated making accusations and stuff, and surrounded the details wrt the SANCTA stuff between lots of other stuff. And dressed all in fancy language. All of it is interesting if you're into that kind of stuff [probably not - I've been spending far too much time in theological arguments with evangelical christians, and like silly theoloical poetry, and and much of this is spam from this. but interesting if you're into theology]

Look for the parable of the prodical sons [Krill+Memphus] there are 8 4-line stanzas. Specifically do a search for the term 'Monte-Carlo'.

This is what I think they did:
1. Run map-scrip a number of times, collecting win-loss data round by round. Probably just attacking in any old random order.
2. Reinvert the RNG. Due to the nature of the combat system, they can get good constraints on this. Since each round of battle is not created equal. You need better RNG for the first few rounds, and don't need such good RNG later on
3. With a legitimate battle simulator, try random battle-orders ('monte-carlo'). The important things are :
- You need a cluster of good RNG at the start of the battle for most units
- you need one good RNG (the last in the battle) for flanking units to retreat. One good RNG is eaier that a cluster, hence why it's easiest to abuse flanking. Hence why 'Flanking' is the codeword for RNG manipulation.
By hand, you can practice this. But I guess your computer can do it for you. Important thing is to find units which win in the optimal number of rounds to have good RNG for the subsequent battle
4. The RNG reversal is only statistical. So maybe there is a little by hand tuning

I'll get on and make a summary, might take an hour or two.

Sorry for confusion! I didn't want to cause problems - seems that over-elaborate efforts to avoid being too accusatory backfired.

For the record - they misjudged things, but weren't doing anything bad.

Anyway, I'll start compiling the evidence.
Reply

scooter Wrote:For those of us who have no idea what you are talking about, can you give us a brief version of what this is about? I tried reading your post and it was so bizarre that I gave up after about 6 or 7 minutes.

Short version: In the CFC multi-team demo game, team SANCTA had several runs of exceptional combat luck. Speculation was that Memphus and/or Krill were somehow manipulating the RNG or combat results, possibly by saving the Pitboss game and loading it offline to test different combat sequences. No proof was established. Sullla and the game administrators asked Memphus to step away from being the SANCTA turn player. In response, SANCTA essentially resigned from the game and it was ended shortly thereafter.

Most of the activity starts here in this thread:

http://forums.civfanatics.com/showthread...ost8613624

and the results here: http://forums.civfanatics.com/showthread.php?t=343000

I am a bit confused as to why this topic suddenly needs to be a thread at Realms Beyond.
Reply

T-hawk Wrote:I am a bit confused as to why this topic suddenly needs to be a thread at Realms Beyond.

Probably one relevant factor is that, while I read a lot of boards, I am not really that experienced at writing or posting. So may well have made a mess-up on the etiquette, and I frequently makes maistakes of etiquette IRL [I suspect that I am somewhat Aspie...]. I stuck it on RB since the issues involved were ones central to its philosophy, since i don't often look at civfanatics, and since the MTDG boards are dead.

I delayed a lot also because I was pretty sure that there were no mal-intentions, and further thinking confirmed that - particularly since I very much intuited th Aspie mentality [I think that this has a lot to do with Krill's apparent rudeness sometimes - I simply don't think he realises. And maybe doesn't see the logic when people complain]. But I knew that many others weren't likely to see things in such a way. So I tried to think of a way to do things and explain. Oh, by the way, one thing I had in mind with the abstuse method of bringing it up - to provide an automatic temporary filter so that only those whe were likely to understand all of this stuff would find it - maybe prevent community uproar, flames, unproductive arguments. But it appears that I shot everyone in the foot there too [and am maybe currently just digging the hole deeper... I hope not].

Regarding the Aspie mentality [also some elements regarding the fact that K+M are both highly intelligent, mathematical etc.] and what happenned in the MTDG:
It reminded me so clearly of the time where I in complete innnocence accidentlally cheated on an A level class maths test:
We were told to sit at seperate desks I assumed that it was intended that we should use all of the information available in the textbooke we were provided with. We were not told not to look up the answers in the back of the textbooks. I was used to doing problems/tests where the emphasis is on using the correct methodology, showing your working and prviding elegant solutions, and where some of the numerical answers are included as part of the question. And whilst learning/practicing maths using textbooks, this was my regular approach toward using the answers in the backs of textbooks. So I assumed that a large part of the purpose of our A-level maths test was to provide elegant worked out solutions. Now as to how one can spend 3/4 of anhour making elegant solutions to simple problems in volving a, t, s, u & v? Well, lets just say that I wasted an awful lot of time at school not learning a great deal very efficiently. Since we were told to 'show our working' I would often write out too much working that I didn't really need. Sometimes I crossed out lines of working after re-considering and deciding that they were unnessecary and detracted from the elegance and efficiency of the working. And sometimes I changed my mind again, wrote notes that these lines should be uncrossed, or re-wrote them out... I checked all the answers in the back completely automatically, and never thought to conceal the fact [c.f. the argument in the MTDG - surely someone wouldn't do it in such an obvious manner if they were doing it deliberately?]. Then at the end, we were told to check our answers in the back, and give ourselves one mark for each correct one and to report this. I immediately thought Oh Shit. I mentioned it to a friend, and realised that my so-called explanation wasn't going to go down very well with anyone. So kept very very quiet about it all - I think that the past few weeks are the 1st time in more than 10 years I mentioned it to anyone.

Maybe one good strategy - somebody very important makes some kind of official declaration that all the issues have been examined, there have been studies of Memphus' battle simulator, and that he is in the clear. But due to XYZ, and mathematical and technical details, you can't explain anything further. It seems that Memphus has banished himself from the boards [at least last time I browsed civfanatics 2/3 weeks ago], so this could allow/encourage him to return [at the moment I think that there a lot of lingering suspicions, which people do not voice to avoid arguments].

And why so sudden and now - I gradually realised what was going on [it all started out as a fascination with statistics] over a few weeks, spent a couple of weeks thinking over moral/practical arguments about where to do anything, and finally took swift action.

Considering the IRL mood I was in the time [and a few analogies related to stuff going on in my workplace, but won't detail], it may well have been impulsive and misjudged.

I have to admit that while I still think that it was the right thing to to, it may not have been such a good idea.

Upon reconsideration, I don't think that there should be pressure on K+M do do anything at this stage or maybe ever in fact wrt MTDG - people who are sympathetic & understang the issues can discuss stuff behins the scenes if they decide that this is useful or productive. And decide if it's just going to create more problems than it solves. Maybe they will want to discuss what the community dynamics are at the moment and how to fix, in case fixes are possible.

Definitely one fix would be for the firaxis people to shore up this RNG exploit in their code [instead of spending their time on fancy expansion of game strategy/mechanics etc. - e,g, I noted that a lot of you think that the espionage system was a badly thought out addition to sell more copies of BtS. But stuff like Blakes AI is the right way to go, and appreciated by everyone and very elegant.] Maybe one useful thing out of this accident is that if Firaxis hears about stuff and what is going on in the communities, they might pay more attention to these kinds of matters. I suspect that there is a lot of this kinds of stuff going on in the online communities that they don't hear about, since people can't be bothered to report it - the just have an argumen, go off in a huss, join a different board, play with other players, or get together and swap strategies for how to use the exploits better in the new metagame.

Regarding the teams in the MTDG. I suspect that some but not all the people running all the battles/war-departments will understand the issues and not be so upset [some will of course]. Probably the diplo people will be most annoyed at what their war departments were doing behind their backs.

I'll type up the evidence [discussions on the SANCTA forums in coded language basically] when I've worked on making it more coherent.
Reply

Judging from your description you need the same RNG seed + RNG rolls that is used by the real game to be able to do this. Is there any way to get this info other than using the actual game save? Assuming you're right this makes all sequential turn games without new RNG on reload susceptible, if you can load the save. In PBEM this would be really easy, but can you load and use the pitboss save, if you don't know the admin password?
Reply

Maybe I'm missing something here, but unless there's a rule against it, why would running simulations be considered cheating? It only provides the best theoratical course of action; in practice it may end up very different altogether.
We don't stop playing because we grow old; we grow old because we stop playing. - George Bernard Shaw
Reply

SleepingMoogle Wrote:Maybe I'm missing something here, but unless there's a rule against it, why would running simulations be considered cheating? It only provides the best theoratical course of action; in practice it may end up very different altogether.

This isn't any kind of a problem and probably many of us does this before starting the battle.

However if you do know beforehand the RNG rolls, that definately is an unfair advantage and allows you to adjust/optimize the battle order so that you get more than your fair share of wins. I don't know why Flanking units are considered the most broken. Does flanking refer here to the fact that they are more prone to retreat and this retreat is decided with just 1 RNG roll? If that is the case also Guerrilla3 units probably work similarily.
Reply

plako Wrote:However if you do know beforehand the RNG rolls, that definately is an unfair advantage and allows you to adjust/optimize the battle order so that you get more than your fair share of wins. I don't know why Flanking units are considered the most broken. Does flanking refer here to the fact that they are more prone to retreat and this retreat is decided with just 1 RNG roll? If that is the case also Guerrilla3 units probably work similarily.

Yes, precisely on both the RNG in general [although you could only get upper & lower bound I would think] and also on the flanking. I wouldn't know about Guerilla 3 since I'm not an expert, but I imagine you're right.

I realise the stuff about unfair advantage - I do not believe that was their intention [see lengthy psychology arguments above], although it may well have occurred in practice, so:

It would be useful if people could contribute their thoughts on the dymanics in various MP communities wrt RNG bug [or other exploits]. Anonymously if you prefer [send a PM to me or an admin or a friend or whoever, so that it can be posted?]
I imagine that it would be possible though somewhat difficult to discover for oneself how to use the RNG, but:
I have a strong suspicion that the RNG tequniques are being passed around from person to person. People pass it on to people they're playing against, then happily use it against each other in friendly fashion [e.g. see SANCTA internal game].

Due to game theory - passing on of techniques and social pressure - then the community is probably a system which evolves into one of two states - see the Selfish Gene by Richard Dawkins [genes --> memes]. And it may be possible to get back to the other potential well, at least in some communities [e.g. as has already happened here].
Reply

Frankly, I'd like to know why this is posted here (without having to go through a monster wall of text).
Reply

banshee Wrote:Frankly, I'd like to know why this is posted here (without having to go through a monster wall of text).

- I knew the players involved regularly came here, so it seemed fine. all on the public internet. Probably a mistake in etiquette - I'm an inexperienced poster.
- Laziness, I hardly ever check civfanatics
- The MTDG-1 boards have been dead, so probably no-one would read it. The alternative seemed to be to post in the MTDG-II boards, but this seemed just as bad an idea, since that's also a different, but overlapping set of people.
- I noticed that a lot [but by no means all] of the people contributing the more sensible arguments and comments in the SANCTA thread discussing the RNG were from RB board, so it seemed sensible to solicit their input. And if others not involved in SANCTA contribute, then this gives some independent viewpoints.
- Due to the issues involved, since this is one of the main RB principles, it seemed a good place. I noticed that rule-discussions and sorting out potential violations were always done very amicably here. But I felt that the diplo/demo-game people at the MTDG game don't quite understand the issues [at least in the vie of the warmongers/MP crowd] on integrity wrt the game-mechanics. So arguments and flames. They are perfectly free to look at this board, but it might be best to encourage them only after any amicable resolution.
- I didn't contact the admins at civfanatics, since they seemed to have a more prosecutorial approach - unneccessarily leading to flames, arguments etc.

Maybe I did the wrong thing by posting here - if so, then sorry.
And possibly a mistake to rehash an old debate, but I felt that it could be done amicably. If this was the wrong decision - sorry.

Will try to be more concise in future.
Reply

Just my personal opinion, but can we not have a revival of THAT discussion here, please?
I don't see the point, especially since Memphus is not a regular here (I think) and is unlikely to enter the disussion again here or anywhere else.
It is bad enough that the double move issues seem to wreak havoc in every CIV forum.

mh
"You have been struck down!" - Tales of Dwarf Fortress
---
"moby_harmless seeks thee not. It is thou, thou, that madly seekest him!"
Reply



Forum Jump: