Detecting people replaying GOTM

[Capt Buttkick]

And part of that is making "finding out how not to be detected" too much effort.

But there's absolutely no effort to "finding out how not to be detected."
You just described one way yourself.

 

[Thrallia]

That's their purpose and reason for being in the first place. I was thinking more along the lines: why are some successful and others not so? For instance, wikipedia is based on people contributing to the wiki. But much more than that, it's based on the belief (and maybe now a confirmed fact) that for every person destroying a part of wikipedia, there will be hundreds or thousands others standing by to correct mistakes, blunders and sabotages.

I'm sorry if I don't get my meaning across at all times. I can't really blame it on English being my second language either, cause I've spoken/written it for about 25 years now. Come to think of it, that may be pretty close to your experience too

lol I actually hate wikipedia for anything that is not pop culture or computer knowledge...I find it extremely unlikely that most of it is edited by people who actually know what they are writing(especially in the history sections)

 

[Capt Buttkick]

I agree you should be more critical to info on wikipedia the more obscure the subject is. To me, it's still an invaluable first sourceof reference if I want to know more on a subject. I start with wikipedia then go on to other sources if I want to know more, or if I don't rely on the info on Wikipedia.

Anywho, you can't argue with it's successfulness, though, and that's due to open participation

 

[DynamicSpirit]

Anywho, you can't argue with it's successfulness, though, and that's due to open participation

Is Wikipedia completely open? Do the organizers for example publicize the best ways to hack into wikipedia, supply passwords for anyone to log into their servers with admin privileges, make it known which pages may be susceptible to SQL injection attacks, etc. etc. in order that the community might work together to close those kinds of things? Or is the openness of wikipedia merely confined to allowing people to freely contribute information within specified user interfaces that have been set up by wikipedia?

 

[humbe]

Decompiling code isn't really that difficult. How easy it is differs as to how it is compiled. Java code, for instance, is ridiculously easy to decompile, and a decompiler comes with the SDK if I remember correctly. A java decompiler can almost restore the original source files. C++ is harder as less debug info is present in binary, but in no way incomprehensible.

I can't think of any reason why people would go through so much work to be able to cheat on an online ranking list of a game though. If you're cheating, you can't tell people how you play, because if you explain in details it will be obvious that you're not doing well because you have a better strategy. Where's the fun then?

After having scanned through a lot of posts in the forums, I see far worse things in the forums that would unbalance the game than reloading though. I dunno what exploits aren't fixed yet, but it sounded like it was a lot of ways to exploit how the AI works. As I don't find all of them in allowed/disallowed list of GOTM, I assume some of them are fixed though. For instance the 'gift GPT to AI, to have him pay more for resources and then cancel GPT gift' thing sounded really overpowered.

HOF crashes my game when I try to load games after I've already started one, but as I don't think I'll be using it for non-GOTM games and it is automatically loaded when I start GOTM game it isn't really any problem.

 

[Niklas]

I have a C++ decompiler on my computer. Like Capt Buttkick says, no code is safe from reconstruction. But like the previous speakers, I can't see who would go through the trouble. Reading decompiled C++ code is not my favorite pastime, it takes forever to glean the meaning of even a short program. There are obviously much much simpler ways to cheat for those who really feel like it, which have been pointed out in this thread already.

 

[DynamicSpirit]

If you cheated, you either stop cheating and continue playing, you keep on cheating the realy silly way and are thrown out of the competition. OR, you learn how to cheat the "smart" way...
Does the staff really think it's so hard to find out ways one can cheat undetected now? Cause that would be naïve...

I think the point is that, by keeping our 'cheat-detection techniques' under wraps, we force potential cheaters to keep guessing what we might or might not be able to detect. Chances are in the process of guessing, and trying out different things, the dedicated cheater will get so many games rejected (because he thought we wouldn't detect something that actually we do detect or even specifically look out for [1]) that it'll become not worth it, and besides, the staff will have been very well alerted to that person in the process, probably to the point of having permanently banned them anyway.

[1] Indeed there's a good example of that in this thread. Someone mentioned a possible way of cheating and held it up as an example of something that would be completely impossible to detect. Yet I can think of several ways that that particular activity might in principle be detected. Bottom line is - the GOTM staff have had years to think of possible ways a person can cheat, and to formulate and in many cases implement ways of detecting those cheats - and some of those ways might be not at all obvious.

 

[Capt Buttkick]

HOF crashes my game when I try to load games after I've already started one, but as I don't think I'll be using it for non-GOTM games and it is automatically loaded when I start GOTM game it isn't really any problem.

Does that with my game too. I still use it, but it means I don't reload even lax, private games anymore

 

[Capt Buttkick]

I have a C++ decompiler on my computer. Like Capt Buttkick says, no code is safe from reconstruction. But like the previous speakers, I can't see who would go through the trouble.

I think Thrallia might disagree with you. He's pretty suspicious to human nature

Like I said at the start, I knew I was
I'd like to thank the staff for taking the discussion, though. If nothing else, it demonstrates to the community that you've thought these matters through, although I disagree with your conclusion

Oh, and Gotm is still

 

[Thrallia]

Capt Buttkick: I wouldn't be so suspicious if I hadn't had so many examples and reasons in my life to feel that way...every time I try not to be cynical, I'm given more reason to be.

 

[Bleys]

HOF crashes my game when I try to load games after I've already started one, but as I don't think I'll be using it for non-GOTM games and it is automatically loaded when I start GOTM game it isn't really any problem.

I noticed something about this in my regular games. I loaded the BTS HoF mod a while back, and use it for regular games, and never had a problem saving atnd reloading. Now, however, that I installed the Vanilla HoF mod, even my BTS games crash if I try to reload from a previous save. Kind of annoying, because I play a lot of games with 2 or 3 "optional crossroads" in an effort to learn how the AI and game itself respond to slight differences in a given situation.

As for the topic at hand, /shrug. If people want to cheat these games, given the amount of effort it would take, then I truly feel sorry for them and their puny lives. Cheating to be the biggest fish in a tiny little pond like Civ Fanatics HoFers is pathetic. My goal is to be a better, less sloppy player, and these types of games are perfect for it, regardless of "where I rank". Having played Everquest for many years from the day it went live, I found it constantly annoying how many people felt the need to "prove" something, as if it actually matters.

And for the record, thank you MODs, for being willing to put up with all the crap that I know for a fact goes into running these kinds of games. /salute, my hat is off to you.

 

[whb]

Indeed there's a good example of that in this thread. Someone mentioned a possible way of cheating and held it up as an example of something that would be completely impossible to detect. Yet I can think of several ways that that particular activity might in principle be detected.

Though that might in part be because the original poster did not actually want to publish a viable method of cheating on the forums for nefarious persons to pick up on!

In any case, it also appears true that we aren't tracking some things that we should be - for instance, as I've said many times before, anonymous browsing of the spoiler threads. (One thing I am naturally suspicious of is late submissions that suddenly blitz an award that few people in the spoiler thread have gone for.)

Yes, I realise that some of this can be picked up with behaviour analysis methods (eg, looking for convenient settles near uncovered resources and assessing whether the settlement made sense without the foreknowledge but only the techs known at the time), but this gets harder to pull off even as the information becomes only slightly more subtle. I'm getting a touch suspicious at how rarely I hear about a failed rush in the spoiler forums - does this mean it is only picked as a strategy when people know it would work for this map, and if so was that knowledge just regular good judgement or something worse...

 

[MarkM]

I'm getting a touch suspicious at how rarely I hear about a failed rush in the spoiler forums - does this mean it is only picked as a strategy when people know it would work for this map, and if so was that knowledge just regular good judgement or something worse...

I recall reading about a number of failed rushes in latest BOTM. I definitely had at least one failed rush in some old WOTM or GOTM. But even if what you said about rushes were true, I think what it more likely says is that anyone who tried it and failed is understandably less proud of their game, and so less likely to boast ... err, post ... to the spoiler threads. And they may even have just quit the game entirely & not even submitted. That seems like a much more reasonable explanation. What you're basically saying is that anyone who regularly rushes and has never posted that they failed in it looks suspicious, which I think is going too far. I'm not going to say that I'm not surprised myself how many people seem to pull off a successful early rush building many fewer attack units than me (I'm cautious I'll concede, but I never seem to have that safe of a margin of success), but I just figure what I just said is the explanation -- the ones who failed tend to just stay silent.

 

[Niklas]

I think what it more likely says is that anyone who tried it and failed is understandably less proud of their game, and so less likely to boast ... err, post ... to the spoiler threads. And they may even have just quit the game entirely & not even submitted. That seems like a much more reasonable explanation.

Been there, done that, didn't bother to buy the t-shirt though.

 

[whb]

I recall reading about a number of failed rushes in latest BOTM. I definitely had at least one failed rush in some old WOTM or GOTM. But even if what you said about rushes were true, I think what it more likely says is that anyone who tried it and failed is understandably less proud of their game, and so less likely to boast ... err, post ... to the spoiler threads. And they may even have just quit the game entirely & not even submitted. That seems like a much more reasonable explanation. What you're basically saying is that anyone who regularly rushes and has never posted that they failed in it looks suspicious, which I think is going too far.

No, what I said is that I'm getting a touch suspicious and nothing more*. I've very carefully not said that cheating is occurring -- in fact the very point that I made in the post is that it can be difficult to discern foreknowledge from good judgement from behaviour/gameplay alone. A point which your reply would seem to back up (there might be plenty of alternate explanations)

* To be more precise, I was quietly making a suggestion on an analysis technique that I don't know whether the mods are using but might be a good idea (and I thought DynamicSpirit might have been referring to), and then pointing out its potential limitations.

However, I think "foreknowledge of whether a rush would work" might be a useful concrete hypothetical example: tangible and realistic enough to think about how we as a community might detect it, but rushes are widespread enough that it can't possibly be construed as accusing any particular player of cheating.

 

[dr_s]

Been there, done that, didn't bother to buy the t-shirt though.

Ditto. In fact 2 of the last 5 GOTM's I've abandoned because of failed rushes. One was definitely a lost game. The other I could still have won (it was the recent Mansa Musa game) but I was so disgusted by my own idiocy I threw in the towel.

 

[whb]

Hmm, I'm going to stop replying to this thread (probably should have stopped a couple of posts ago). While I think "how to detect cheating" is an interesting problem, if we hypothetically deconstruct the mod to work out what techniques can and can't work, then that's probably as bad as posting the mod code.

On the other hand, the argument that most cheats will already have revealed themselves by being incompetent at their first attempts doesn't convince me -- Dwaine Chambers was only caught after Balco was revealed by an insider; he didn't have six failed attempts at using detectable drugs first...

From my perspective it's all about maintaining a believable level of confidence in the competition, and then hoping that the smart honest players outnumber the lazy villains. (Which is why I don't want the mod opened up -- in case the mod is weaker than I hope and destroys my confidence!).

 

[Capt Buttkick]

Hmm, I'm going to stop replying to this thread (probably should have stopped a couple of posts ago). While I think "how to detect cheating" is an interesting problem, if we hypothetically deconstruct the mod to work out what techniques can and can't work, then that's probably as bad as posting the mod code.

Or as good as...

On the other hand, the argument that most cheats will already have revealed themselves by being incompetent at their first attempts doesn't convince me -- Dwaine Chambers was only caught after Balco was revealed by an insider; he didn't have six failed attempts at using detectable drugs first...

Yes. There are just too many smart people here Some of them are also bound to be cheating and they'll certainly find ways to get away with it.

From my perspective it's all about maintaining a believable level of confidence in the competition, and then hoping that the smart honest players outnumber the lazy villains. (Which is why I don't want the mod opened up -- in case the mod is weaker than I hope and destroys my confidence!).

I'm certain the mod wouldn't disappoint you. The staff has earlier stated that the mod have helped a lot with cleaning up the competition.
I'd like the mod opened for much the same reasons as you'd like to keep it closed

BTW: I was active in the pre-game discussion for two ongoing gotms. I made a mess of both and might not post in spoilers for the obvious reason that I don't believe anyone will find anything interesting in my game (maybe I'll post on one of the games, haven't decided yet).

 

[DynamicSpirit]

No, what I said is that I'm getting a touch suspicious and nothing more*. I've very carefully not said that cheating is occurring -- in fact the very point that I made in the post is that it can be difficult to discern foreknowledge from good judgement from behaviour/gameplay alone. A point which your reply would seem to back up (there might be plenty of alternate explanations)

FWIW in the past the reports I've tended to be most suspicious of are the ones where someone is reporting an outstanding game, but then no GOTM submission from them turns up - in most cases there's probably good reasons - loads of things can crop up to stop you submitting, but I'm sure it's crossed a few people's minds to make up stories of incredible achievements I'm pretty sure in most cases any preponderance of great games being reported in the spoilers is down to people who play good games being more likely to post and submit. (Which is a bit unfortunate for the newer players as it can give the impression that the community has higher average game skills than it actually does)

* To be more precise, I was quietly making a suggestion on an analysis technique that I don't know whether the mods are using but might be a good idea (and I thought DynamicSpirit might have been referring to), and then pointing out its potential limitations.

You were roughly correct. I was hoping to get away without specifying what I knew about how to detect that , though in this case it may be worth while - just to emphasize the point that we have thought about all the possibilities. You can get more elaborate too (deliberately placed honeypots in multiple games etc.). Of course a 'cheater' could guard against that by not taking obvious advantage of their foreknowledge of the map, but that in itself reduces the advantage of cheating in that way anyway, so making it less worthwhile.

I think your point about maintaining a believable level of confidence in the game is a good one, hopefully we're managing to maintain that

 

[LowtherCastle]

Given the security in place, there are still two ways that I can think of to detect cheaters:

1) GOTM not announce any new security implementations so they can catch the work-around cheaters off guard.

2) Monitor game openings--cheaters are the ones who will invariably have perfect starts, for example, place their settlers in an unlikely, yet ideal, place to get a resource, or explore directly toward the nearest AI or some other early strategic target. We all get lucky on occasion, but it eventually becomes obvious when certain player moves that defy all common sense lead to 'good luck.'