A little better quality control on this month's GOTM

[Aeson]

Originally posted by EEKthedog

There has also a game been found with 567 loads. Don't know what type of game it was, but as of yet it's the record. I doubt anyone would have 19 playing sessions every day of the month.

I probably have exceeded that in my HOF submission (took me closer to a month and a half though)! I'm not exacly sure as to the overall number of play sessions, but I would often save and go do something else 2 or 3 times a turn. At one point I had close to 1600 overall workers (700 captured) and for a while a couple hundred tanks in action each turn. Moving workers around was especially boring, and at the peak would take close to 45 minutes a turn. Add in the setting all the specialists by hand each turn, moving military units around and attacking (keeping in mind this was back when you couldn't use 'j' because it would empty your armies out), and setting build orders... and playing through a single turn becomes rather daunting. Then the game would crash half the time I went to diplomacy (so I had to save every turn before that stage, and reload of course if it crashed), and occassionally between turns. Could be 1000+ load count on that game very easily.

If a person were to spend 7 hours a day on their GOTM milking, I don't think that number of loads would be out of the question. That would be an average playsession of around 20-25 minutes, which while milking a game can be too long to deal with. Probably have to be a large/huge map to be likely though.

---------------------

It does sound like this reload counter should only be used to support other more evident claims to cheating. The first game it was quite apparent that something wasn't right, this one I'm not so sure of. The reload count seems high, but maybe this person only had a few minutes to spare at a time? Playing it at work and hiding from the boss... or playing it at home and hiding it from the wife... I was thinking about going through the map and seeing how early it could be beaten (default rules) with reloading. It seems I have been excluded from trying though.

In any case, I think the benefit of the doubt should always go to the player. Better to let some cheaters slip by than to wrongly accuse someone who's innocent.

 

[Serg]

I find the place of "reload counter". My own is so big...

 

[Cartouche Bee]

Bamspeedy,

To make this sort of strategy work you need to discover horsemen and writing. Moving the capitol to the sea facilitated making a discovery and moving closer to the prey. The 'slow' start is rewarded with a settler from the goody hut to the north. Items like temples and barracks were built while waiting to produce the first horsemen, a handful of warriors provide early exploration and the popping of goody huts. Temples aid in pop rushing when and if required. A number of cities were won through peace negotiation, that explains the size one cities that are acquired the turn they are founded. Hitting all the capitols first guts each target civ by taking what is likely the most developed city. The under lying skill and knowledge is excellent and I doubt that there are 3-4 people that could play the game to this conclusion even with reloading and having forgone knowledge of the final result.

Aeson,

My first milked game I had some turns that took two hours each. I was ferrying workers from one continent to the other, plus all the other things that one has to do. My initial reaction to the "loading" count was the same as most of the comments made by everyone because I know how long some of those milked games took. It is embarrassing to admit to the time it took to manage thousands of workers, military units, city inspections and micromanagement. I personally don't think it is fair to judge games based on how long each individual can play each session of civ.

 

[Cartouche Bee]

Originally posted by Serg
I find the place of "reload counter". My own is so big...

I believe that we will be able to determine if the field has been edited.

That of course would be illegal tampering.

 

[MSGT John Drew]

Far different from anything that most of you have said about playing Civ3, I actually am happy with the autosave feature and have grown to rely on it even more than manual saving. that is shown by the fact that I only have 17 manual saves from GOTM12. In my engrossment in the game, I even forgot to make that 10AD save and the nearest I have is 130BC. I actually did not save until 230 AD!

All in all that's 17 saves in 4 days and without writing a gamelog while playing it. Removes the boredom factor -- *snaps finger* -- just like that!

Another thing that surprised me was that my first manual save for GOTM12 was not until 630 BC. The cons, though, are that I will not be able to compare my game at, say 1000BC, to same-date screenshots posted at the spoiler thread.

Still, I may not be saying the same if I ever did 'milk' the next GOTMs although I could not think of a reason for doing so.

ps: I imagine that if you need to find that 'reload' key, you need to look for it in autosaves(uncompressed) and not the compressed 'manual-saves', right? After what I saw 'inside' the autosaves you might even find out the build order of each civ (maybe). One thing's for certain though, the first few pages of 'text' in the savefile for each savefile is EXACTLY the same!

ps: i'll be submitting my game as soon as I can decide if I write a gamelog for it or not. looks like too much work for such a good game.

 

[ainwood]

Hee Hee! This has really put the proverbial cat amongst the pidgeons, hasn't it!

I would make a few observations on the previous comments.

1.) We don't want to let people know exactly how the reload counter works, in case they try to edit it. However, I will say that from my testing it appears to be completely reliable. And its not as easy to "edit" as people may think

2.) The counter DOES NOT count the number of saves. Only reloads. Save all you like; if it crashes, then reload. Crashes are infrequent, and it will make little difference to reloads.

3.) The reload counter is only one bit of information that can be used to identify potential cheaters. There are other ones, which I am reluctant to go into for now.



I would suggest that if people find time limitations mean that they are having to play over many sessions, they should submit more saves. The turns at the start of the game are very short, so submit the save at the end of your first session (2000 or 1000 BC for example). If you are taking a long time over many sessions due to milking, then submit a save prior to the milking phase. Eg - if at 800 AD, you have 78% of the land and 10 reloads, and at 2050 AD you are at the domination limit and have 300 reloads, then submitting the 800 AD save will add a lot of credibility to your score, with little effort.


One other issue: When I first load up Civ from the desktop, then choose the save file and load it, the whole process takes around 3 minutes late-game on a reasonably fast PC (1 GHz, 256 MB SDRAM). 4 minutes if I have to boot the PC as well. If someone is making 100 "legitimate" save and reloads (ie starting a new session), then I estimate they are spending around 300 - 400 minutes (thats over 5 hours!) a month just waiting for civ to load and then to load the save game. Perhaps a rationalisation of the playing sessions would be beneficial?

 

[Phillip_martin]

Simple question if I:

save in 1000BC
go to bed
reload the 1000BC save
Save in 10AD
go to bed
reload the 10AD save
crash
reload the 10AD save
play until completion

Is that counted as 2 or 3 reloads?

 

[ainwood]

Phillip,

For my own (twisted ) reasons, I think that answering this will do more harm than good (I do know the answer FWIW).

Needless to say, crashes are now fairly irregular, and I would expect that a few crashes would not add anything like as many reloads to the count as malicious reloading does. If this has happened to you, then rest-easy.

What I would say though, is how much have you played after 10AD and before the crash? I think that trying to reload from a save-game would be more in the spirit of the game than reloading a save that is potentially 50 or more turns old.

 

[Phillip_martin]

My question was (without getting the source code) to see if it distinguished between a set of 'string' reloads (saved to continue later) and a 're-start' reload (same save used more than once). This would be a fair question that players would need to know prior to accepting the GOTM in which the checker is used.

I agree with the concern some players have that grabbing an hour of playing here and there may be seen as a high number of reloads.

I also agree that players should keep more saves in case questions are asked.

 

[Aeson]

ainwood,

I've had several games that crashed about every other turn (mostly in 1.17f). GOTM7 was the worst for me, crashing several times a turn and eventually I just gave up and didn't finish. Just disregarding game crashes as insignificant when considering the reload counter isn't very sound. Crashes may be infrequent for you, but that doesn't mean they are for everyone. 1.29f is certainly more stable for me, I couldn't say if it's more stable for everyone though, just my configuration.

Also I wouldn't think it uncommon for people to be switching inbetween games played, especially when milking. I know I do, it's just too boring otherwise. That could easily account for hundreds of loads over the course of a game. Then you have to account for people's situation, they might not be able to sit down for hours at a time and play a game. If these things do affect the counter (as it sounds like), it doesn't seem very reliable to me. It's just as indicative of length of gaming session, and game stability as it is of reloading.

Firaxis put this counter in, it might mean something that they didn't tout it as an 'anti-reload'.

---------------------------------------

I played through GOTM10 trying to beat alain's conquest date tonight. I'm a bit rusty at early conquest, had so-so luck in combat, but still was able to finish off the last AI at 410BC. I did do some reloading, probably about 30, though I'm not exactly sure yet how the counter works and didn't keep count (partly cause I didn't know which would count ). Most of my reloads were to get a good start, playing with city placement, tech path, and huts. I also tried to explore as though it was my first time on the map, sending warriors west as well as south and east. It certainly could be done without any reloading, although it would require a lot of luck. I'm not sure I believe that alain's game was played by the rules, but it is possible as far as I can tell.

If there is other evidence you guys are holding back, I understand. I personally don't think it's right to keep back that information, makes it impossible to explain or defend for anyone who really is innocent. Also, it really cuts back on the amount of testing that those methods.

Here's the final save.

 

[col]

It seems to me that what we're heading for is that the number of reloads is something to look at for games that have caused suspicion for other reasons and that seems reasonable to me.

Its just another piece of evidence that MAY be used to identify potential cheaters.

Similarly asking for more saves MAY be useful in certain cases.

I'd hate to see a rule about how many saves you're supposed to have, or how you're supposed to save and reload.

Most sports events test the winning competitors in any event. I dont anything wrong with checking out the leading games in each category a little more closely. It seems the early conquest games are the ones to look at most closely.

Not something that is likely to affect me!

 

[Phillip_martin]

Originally posted by col
Not something that is likely to affect me!

You and me both

Still I can't help but argue for a system which is clear and understood. If you then break the rules then you get no sympathy.

If rules are vauge and unclear (say, to protect the checking system) then I hope there is a good dose of "due process" before someone is publicly labeled a cheat. I still have not heard a clear ruling on Alain's game!

 

[Aeson]

Originally posted by ainwood
One other issue: When I first load up Civ from the desktop, then choose the save file and load it, the whole process takes around 3 minutes late-game on a reasonably fast PC (1 GHz, 256 MB SDRAM). 4 minutes if I have to boot the PC as well. If someone is making 100 "legitimate" save and reloads (ie starting a new session), then I estimate they are spending around 300 - 400 minutes (thats over 5 hours!) a month just waiting for civ to load and then to load the save game. Perhaps a rationalisation of the playing sessions would be beneficial?

Takes 20-25 seconds for me (desktop to actually playing). 1.33GHz 512MB. Switching between games only takes a couple seconds...

 

[col]

And me - I like to switch between GOTM / Tournament / RBepic games - and some just for fun.

Takes about 10s to save and laod a different game once civ is running.

 

[ainwood]

Phillip,

I may seek consensus of the GOTM police on releasing the details, although my feeling is that for a genuine player, it makes little difference whether its the same file reloaded again and again, or a continuous game reloaded. However, a requirement to submit an "early" save will help clear this up.

Aeson;

I was thinking of V1.29f with my comments. If people want to protect themselves from suspicion, they could simply keep a save from before the frequent crashes and submit that as supporting evidence.

In terms of the "other" evidence, I don't think you need to worry. Without giving details, the things I am looking for are direct signs of cheating - normal play simply won't result in these circumstances. Rest easy that no-one needs to be 'careful' about the way you play to prevent their game looking suspicious.






Aeson,

What I can say is that switching to other games will not confuse the reload counter directly. The reloads counted are game-file specific. The only thing is that when you reload your GOTM game file, the counter will be incremented by one.

 

[ainwood]

Originally posted by Aeson


Takes 20-25 seconds for me (desktop to actually playing). 1.33GHz 512MB. Switching between games only takes a couple seconds...

Looks like I might have to invest in some more ram!

 

[Cruise]

4 minutes for a save? I have 384Mb RAM with a 700mhz Duron and for me it also takes 20-30 seconds.

 

[ainwood]

Originally posted by EEKthedog
4 minutes for a save? I have 384Mb RAM with a 700mhz Duron and for me it also takes 20-30 seconds.

Maybe it just seems like 3-4 minutes..... I really shold time it. But my PC does tend to spend an age actually loading up the OS, and then loading civ.

I am jealous of you fast booters though!

 

[thefrenchzulu]

Personally, I'm not happy with someone being accused in public when there might be doubt. Just not cricket.

I appreciate the work the GOTM police are doing, especially as far as modified .bic files are concerned. Thanks.

The loading and reloading issue worries me. Like Aeson, I tend to switch between games. HOF, GOTM and Tournament. I do that to break some of the boring stages of the middle and end game.

I also enjoy writing my spoilers before I'm finished with the game. That sometimes involves going back to key stages in the game. I try to save these stages specifically for spoilers.

I also tend to perform detail city investigations. Unfortunately this involves screen capturing a city, going out, saving it in Microsoft Photo Editor and going for the next city. Alt tab results in an immediate crash. This will result in reloads for every city investigated. (Not saving and reloading will be cheating in reverse) I perform this detailed spying before engaging in any war and do it a few times during a game.

I will not even mention pressing the wrong buttons during the game. (Like when you are busy bombarding and suddenly you build a stupid colony.)

Nor the fact that my notebook's mouse button that gets stuck. Immediate save exit, swear words, opening of keyboard, further swearing and the reloading.

All these factors results in many reloads during a game, but all in good faith.


A few points on cheating, if I wanted to cheat:
1. I could play the game on two or three PC's. (I do have a few)
2. I could get the world map and start from scratch
3. I could use all kinds of illegal utilities!
4. I could read the spoiler before starting the game


Taking all the factors into account, these competitions cannot take place without a certain amount of trust and ethics. If the current competitions rules are going to become impossible to adhere to without the possibility of being labelled or accused as a cheat, I will no longer participate, but rather wait for PTW, where cheats will not last long…

These games are suppose to be fun…

 

[Cruise]

Originally posted by thefrenchzulu
Taking all the factors into account, these competitions cannot take place without a certain amount of trust and ethics.

There are also rules set in this competition, and now that there are methods to verify if someone possibly might have broken the rules, we should definately use them.

We won't just put a label on someone as easily as you fear. But when someone has 70+ reloads on a (370) BC victory; already looking odd when loading it up yourself, it does not take a rocket scientist to conclude something is wrong. This example more or less came out because somehow it was not in the data the GOTM police reviewed. When the results were posted we first had a look at it. Someone else also saw it and posted about it and that way it became public. I doubt we'd even post our finds in the main forum. We'd simply keep it in private.