GOTM 16 Statistics

[SirPleb]

Originally posted by cavemanf16
Also, I'm wondering if the 'hours of play' can be extracted from the PTW games? That might show some differences in scoring vs. real world time spent playing, thereby giving more of us a true idea of how much real world time needs to be spent on the game to achieve scores of 20000+.

I'm afraid that data won't be useable from my games. When playing the GOTM I leave Civ3 running most of the day on my computer, to keep the reload count down. (Alt-tab back and forth to other stuff.) So my reload count will usually be the number of days I played the GOTM. (Plus a bit for crashes and for times when I want my screen at a different resolution.) Of course this means that the time played for my games will be insanely high. Other people have mentioned this alt-tab approach, so my games won't be the only inflated ones.

BTW, a note for those who use this approach: Civ3 is at most times a huge CPU hog, using all available cycles. This can make one's computer rather sluggish in other work if you alt-tab away from Civ. I've found a useable work-around for this. When inter-turn popup windows are displayed (e.g. a popup indicating that a city has finished producing something, asking what to produce next), at that point the game does not use all available cycles. That appears to be one place in the code where it patiently waits for an event from Windows. So, if you can organize your play time to end a session between turns, at a point where one of those popups is displayed, you can alt-tab away from Civ and have your computer run nicely.

 

[Skyfish]

Great tip.
I play at work so I alt-tab away no end but what I find most annoying is the screen resolution going back to 800*600 when I normally use a 1200 screen when not civing : got any tips for that problem ?
thx !

 

[ControlFreak]

at the following to your .ini file

KeepRes=1

 

[Shillen]

Uhmm what's the point of leaving the game running just to reduce your reload counter? Not accusing you of this, but the only reason I could see to do that is to try to hide reloads. Why does it matter if you play 100 turns per load or 10? Personally I don't think there's any reason to even post data tables on reload numbers. Use it to determine if people are cheating, great, but why does anyone else need to know how long people play civ without quitting? What could anyone gain from that knowledge?

 

[cavemanf16]

Originally posted by Shillen
Uhmm what's the point of leaving the game running just to reduce your reload counter? Not accusing you of this, but the only reason I could see to do that is to try to hide reloads. Why does it matter if you play 100 turns per load or 10? Personally I don't think there's any reason to even post data tables on reload numbers. Use it to determine if people are cheating, great, but why does anyone else need to know how long people play civ without quitting? What could anyone gain from that knowledge?

As some other people have been talking about in another GOTM thread, lots of people submit timelines and several go as far as taking screenshots periodically and writing down their activities as they play. So it's not unheard of, but I'm even more surprised people play AT WORK!! OMG!!

 

[Ribannah]

Hmmm, I do most of my work at home. Occasionally I take a short break from civving to get something done. Sometimes. When I feel like it.

 

[Skyfish]

So it's not unheard of, but I'm even more surprised people play AT WORK!! OMG!!

It's nice to be the boss
Everybody rushing around doing your job...and nobody to tell you to stop civving

 

[EMan]

...........there has been the occasional implication that GOTM games with a high reload count have been cheated!

THAT'S JUST NOT FAIR!

I usually have a high reload count for several reasons:
(Using a Dell Notebook)

1. When travelling, I shut down the computer to transport it. Civ3 when running will not allow my computer to be "hibernated".

2. The game is a CPU hog and sometimes my machine "grinds to a halt" so I have to reboot. (Does anyone know how to reverse/prevent this problem?)

3. GOTM17 had 2 "bug" situations for me ("disband eqWorker" & "scrolling F1 screen") whereby it took me several reloads to figure out what was causing the "crash".

4. Growing up as a tournament chessplayer, I like to study positions and try to make the best moves (viz. micromanagement of workers and cities).........this takes a loooooong time, so for me, GOTM is a verrrrrrry long game and thus I play very few turns per session.

Speaking as a former(?) "Dairy Farmer", I'd love to MOOt this discussion. Here's ONE solution I can think of:

CHAMPIONSHIP GAME MODE
The "Championship" game can only be saved in ONE slot. So if you've loaded the game, you cannot perform another "Load" instruction until you perform a "Save" first. The only exception would be if the game crashes/you have to reboot the computer. The Reload Count would always be ZERO!

Would this be "easy" for Firaxis/cracker to implement? (SirPleb?) Or, am I just being overly sensitive to the "Reload Issue"?

 

[cracker]

Originally posted by EMan
...am I just being overly sensitive to the "Reload Issue"?

You are just be overly sensitive to the issue.

Look at the statistics in the big picture context of the game and respect them for what they say. We have the ability to separate games that have a low turns played per session count for specific late game reasons from those games where players have excessively low turns per session throughout the game.

This screening process is not arbitrary and is based on the longterm observations of player performance and player behavior in literally thousands of games.

Each 3 month period we are able to review approximately 400 game submissions and out of these 400 games we have about 20 games (or 5%) that have low turns per session counts. OF these 20 games we can clearly validate about 12 to 15 games as being OK but the remaining 5 to 8 games either have evidence of gross incompetence or immaculate conception that would make them suspect submissions.

I think you may be being anal retentive on this issue and need to just recognize that we use the statistic as a tool to detect those games that just are not like the other games in the player pool. Once we identify these games as being different, then we manually review and screen the games using expert panel review. We also review games in the main part of the player distribution curve, but the selection process for review there is based on other criteria.

We can also look at the low turns per session count games and say that these games account for approximately 1 day of delay in publishing the GOTM results for every 2 to 3 of these games that are detected. The overhead associated with contacting the players for extra games, reviewing the extra games, communicating with the players in this category, providing assistance to players with recurring problems in this area, and dealing with general issues associated with this tiny 2-3% segment of the player pool really does suck away resources from other tasks. For Gotm16-Rome we could have published the total results package 3 to 5 days earlier if we could have addressed these behavior issues in just a small handful of players.

 

[EMan]

......I'm guessing you can "somehow" tell who is reloading a non-current position in order to replay the game a different way? OR, perhaps people are able to edit the map in such a way as to change the position of units........beyond my knowledge. (Confession: I haven't even used the Civ Editor, yet!)

So much for the "Honor System" words used as part of defining GOTM!

Since, I don't know the specifics of what you know, and I'm assuming you don't want to divulge them here in order to tip-off would-be cheaters, I shall remain mystified and in a state of blissful ignorance!

Anyway, moving on, if you had a Championship Mode as I outlined in my previous post, wouldn't you eliminate all, or most of, this extra "police" work?

BTW, the volume of work, depth of analysis and dedication to the "Fairness of Competition" that you and your team perform on GOTM and in The Forum is mind-boggling and true dedication!

..................UNLESS it's a High-Paying job!!

 

[cracker]

Originally posted by EMan
...So much for the "Honor System" words used as part of defining GOTM!

This sounds a little bit like sour grapes or a cheap wine cooler to me.

In reality the Honor System is alive and well and I can safely say that most of if not all of the GOTM players play an honorable game that sets the standard of play that many other game sites aspire to.

I canoot share with you specifics of the game review process, but it does suffice to say that the review process is not designed to keep the honest players honest. They do this by themselves by providing a reference set that is very impressive.

The review process is designed more to protect the integrity of the community by detecting those players who submit games that indicate they may need more guidance in the form of technical help or assistance. It is rare that this assistance needs to take the form of disciplinary intervention because the wealth of examples is so strong that it creates the desire to be an honorable participant at all levels.

 

[SirPleb]

Originally posted by EMan
Would this be "easy" for Firaxis/cracker to implement? (SirPleb?)

It is an interesting question. I don't think anything unbeatable could realistically be done but I do think that Firaxis could implement, without terribly much work, something which would help Cracker and his team a fair bit.

I don't think a single save slot approach would work, it would have a few problems: it would probably be easy to defeat by copying files; it would probably preclude submitting mid-game saves as checkpoints; it would not block reloading to change battle outcomes or to peek ahead (because crashes must be allowed for there's no way to block that too.)

Best approach I can think of would be to generate a unique id for each map (a thing called a "guid" would work nicely for this), then record in a well hidden place a bit of load history for each guid. (There are some tricks available to improve the hiding part.) Record this stuff only after a player makes a move, to avoid falsely counting loads where the player was just looking. Strongly encrypt this information in save files and the hidden place. The result would be a reload counter which would ignore all normal loads, counting just reloads where the player goes back to a load which has previously been "played forward". Crashes would still count, there's no avoiding that I'd say. Something like this could reduce the workload for Cracker and team substantially. But it wouldn't eliminate it, they'd still have to look at some games and apply their other factors.

BTW, I have the same problem thinking about this as with work I've done in the past to fight software piracy. For every scheme I think of, I can see ways around it. It does not seem possible to stop a sufficiently determined software pirate. Unless that is, you take measures so extreme that they badly impact legitimate users. (And even most of those which have been attempted don't work - remember "dongles"? What a pain!)

Could Firaxis be convinced to add something along these lines to at least help a fair bit? I have no idea...

 

[Ribannah]

What works is if the old savefile records that actions have been made after it was loaded, increasing a counter.
It is certainly possible for Firaxis to make it so.

 

[Ronald]

I was interested how the different victory conditions did in Gotm16 under the old civ3 scoring system and how the would have done with the new Jason scores.

In this file you will find 2 spreadsheets with tables for every victory condition, one with the civ3 score, one with the Jason score:

http://www.civfanatics.net/uploads4/gotm16_scores.zip

This is the table for the top 20 according to the new jason score system:

Ronald

 

[RufRydyr]

@Ronald
Thanks for putting the work into this. It looks great.

But this sux! I went from 6th to 28th place.

My milking days are over and I might actually have to learn how to play!