I'm not suggesting you remove the anti-cheat system, by any means. Just that the savegame stores some more turn-by-turn history, so trigger values don't immediately cause an anomalous score, but they do cause an automated investigation. e.g. storing how a player's empire-wide PQ increases over the course of a game. +5% at the start cause they picked a PQ bonus, +8 on turn X cause they researched tech whatever, and so on. Same for AIs - if it's an incredible AI and it's got a massive PQ penalty, there's a problem. Some unexplained events and technologies could be allowed if they match ones in the approved mod library, but not too much leeway. Especially for techs, since the AI can't use new ones properly.

The same thing with money and trades - someone makes 10,000bc trading worthless stuff with an ai. The save-game stores the AI's difficulty level and the AI's empire state, so it knows if the empire would have made this trade or not, given its current economic status. On beginner, it'll happily send itself into debt all the time. A normal AI might make the trade if it's got lots of cash, but maybe not if it's broke already. A genius AI wouldn't make this trade at all. If an unexplainable trade shows up, flag the game as anomalous.

A possible cheat editor could get around this by adding the bonus, plus adding a bogus explanation to the log, like 100 +1 soldering bonuses in a row picked up by a survey ship, spread out over 100 turns. (If they're all in the same turn, that should be another trigger.) This is the scenario where you'd have to set some threshold of "how lucky does someone have to be before we get suspicious?". With this way, you might get a handful of geniunely lucky people (or save-reload gamers, repeating each time till they get the bonus they want) with anomalous games, but nowhere near as many as now. And it would still keep cheaters out.

The only question is how much stuff you want to score in savegames to do this. Making a savegame log every single detail of a game (e.g. a savegame after 100 turns actually contains a save of turn 1, turn 2, etc., or a list of every single action which took place between turns) could make them too fat, but you could do a lot of decent cheat checks without every bit of info, as long as the critical things, like cash, empire-wide bonuses, and the empire's production capacity are fully logged.




(sigh) I suppose. I guess the thing that worries me the most about it right now is the deadweight effect caused by the current metaverse-scoring (not game-scoring) system. We've seen people make new usernames to get rid of a couple weeks of deadweight scores in the past. As the game gets older, these recalcs are going to have more and more impact. If people feel that they have to make new usernames and throw out their last few MONTHS (not weeks) of submissions in order to remain competitive, there's a good chance they'd just quit in frustration. That'd include some casual gamers, not just crankers, especially if they log-in one day and see there's not a single beginner-level player in the top 25. Maybe a little tweak of the metaverse-scorer, not just the game-scorer, to take this into account somehow?

Some of these posts re the Metaverse are confusing me.

Is it really reasonable to expect a beginner-level player to be in the top 25, when there's more than a thousand players in the Metaverse?

What kind of gameplayer plays a couple of games, learns a couple of beginner strategies, and then expects to top the Metaverse board ahead of a thousand other players who have been diligently working their way up to stronger AI opponents in larger galaxies?

A couple of earlier posts (this thread or the GCCG thread) mentioned that the Metaverse was not supposed to be a measure of skill. Huh? If not skill, then what? If its purpose is to measure dedication, as in number of games played, then "cranking" is the absolute best tactic to use to top the list.

Shouldn't a list order be supposed to represent excellence in something? If not, then just plug in names randomly, like a lottery drawing. Of course, then the list has no meaning whatsoever...

The Metaverse rewards everyone for playing the game to the best of their abilities... what ever the skill level may be...

The thing is to get a balance.

If beginers and normal players (sheesh.. i've only been able to beat normal level once or twice why shouldn't I be able to compete.. ) can't see that they are getting anywhere in the metaverse then they'll simply drop out.

no matter what skill level you play there will always be quirks in them and weak points that can be exploited. If you can find them then good on you.

the problem only becomes such when you start going overboard and start getting anomolous scores be that be using outragous mods, overusing cheeze tactice or exploiting weekness in the game. then when starcock crew tell you why your getting those scores and you don't listen to them.. and expext them to be there at your bec and call to keep editing your scores, after they've told you not to do it.. and ways to stick with your style of play and still get decent scores.(like dumping money before you end the game).

I for one don't use any extra mods because i'm scared of getting anomolous scores. Until something is inplace that will warn me of an anomolous score prior to spending 2-32 days of playing the game i won't play with extra mods.

Its silly playing the same thing when you know its going to give you bad scores..

Stardock goes out of their way to try to make things work and interesting. no where else is there a game with such an involved scoring system.

I can remember games of old where if you got too much gold or too higher score that the counter simply reset itself to 0. or worse went into negatives.. once ppl found this out they simply made sure that they didn't do it again.. I mean its only logical..

Yeah, on a normal ladder, having beginners in the top 25 wouldn't happen, but the devs have said they want this ladder to represent a mix of dedication and skill.

If they wanted the metaverse to measure pure skill, they might as well just take the highest score anyone's ever submitted. One maso game shows more skill than 1000 crippling games, right?

If they wanted pure dedication, they could just get a stopwatch and time the numbers of hours people have spent playing GalCiv.

Right now, they say they want a mix. If that's the case, then there should be some representation of lower-level difficulties in the top 25 (not necessarily the lowest level, but it shouldn't all be crippling and maso).
[Message Edited]

But it's only logical that, if this is the case, then the Metaverse has to first "know" each player's ability in order to assign the proper level of reward.

And that doesn't (realistically, can't) happen.

(And please, readers and repliers, I'd like to leave out all the personal hurt feelings and angst here, and just discuss the problem from a gameplay point of view.) Basically, it seems to me that playing rapid games at low difficulty, when the player could possibly play the game at higher difficulty levels, is being likened to cheating at Metaverse. Sorry, but I gotta disagree with that point of view.

This "from each according to his abilities, to each according to his needs" is a nice concept, but no one has ever figured out how to make it work in real life.

You may WANT the Metaverse to reward each according to his best ability, but the simple fact is that it's a point score. If a player on beginner huge gets the same number of points as one on maso huge, then there is no legitimate reason that the maso player should not play a bunch of rapid beginner games if he wants to get the greatest Metaverse reward.

It's just ordinary, simple, stimulus/response behavior. If the skilled player wants the reward the Metaverse gives him (high placement on the list order, bigger empire, bragging rights, whatever), then he should crank out as many low-level games as he can, as fast as he can. Because, the way the thing is set up now, that's how he can collect the most Metaverse points and advance farther, faster, in the ratings.

I note that this is apparently changing, starting with v. 1.092. I also note that a number of players aren't motivated all that much by the Metaverse--otherwise the entire top 250 would be beginner players.

If someone could come up with a system that actually did reward players the most points for playing at the top of their ability, that would be very cool. But that's not the way it's set up now.

The current Metaverse point system has the same problem this philosophical approach has always had: it encourages those with highly developed skills to perform below their peak abilities--and rewards them most for underperforming. The GCCG is just a symptom of the flaw in this type of system in that it optimizes the point reward.

Now, if someone could figure out a Metaverse system that really DID reward according to ability, then you'd REALLY have something--not just for playing the game, but for all of human society, too.

An interesting conundrum.

First 20 games at any given level scores 100% of the metaverse amount. the 21st-30th get 95%. The 31st-40th get 90%, ect.. Tweak the range or percentage for balance. Games past the 90 day (or whatever it is) degradation cycle are not considered in the calculations. Could even reward new players by giving the very first 10 games 110% of the score (would not reset due to score degradation time limits)

Two ways to approach the condrum.
1. rewards the person for playing at max ability.
2. penalize the person for not doing so.
or, both.

No?, ok, need more coffee then. LOL

An intelligent approach, couslee.

But, probably only 5% reduction isn't much disincentive. So, for the sake of argument, let's consider a 90% reduction (games 21-30 get only 10% credit, etc.).

Now, instead of encouraging sandbagging for an infinite number of games, it's only encouraged for 20 games--not a total solution, but a definite improvement. And, a sandbagger will only play 20 games at low level in 90 days. Even though he can go back and play another 20 games 90 days later, someone cranking out 30 min games would definitely have to move up in difficulty--so, that's also very good.

But consider the player who really is playing at his max ability. Try as he might, he can't get past beginner or normal (or crippling, for that matter). After 20 games, the Metaverse basically tells him to get lost for the rest of the 90 days--and that's bad, because Stardock wants to encourage fan(atic)s. Heck, I've played 10 games in about 20 days.

Also, a maso player that finished 20 games would actually have to move DOWN in difficulty to keep his ladder scores high. (Actually, that applies to anyone playing at his highest ability: after 20 games you'd be forced to drop down in difficulty to keep your ladder spot.)

Of course, instead of 90% score reduction, we could go with the 5% you suggested, but then the disincentive effectively disappears, and a cranker can still max his Metaverse score with low level games.

Another tweak would be to include the score reduction, for the number of games at the same difficulty, along with an increased reward for higher difficulties. Currently, in v. 1.05, the score for a maso and a beginner game are about the same (for the same size map). But then you're back to the highest scores on the ladder almost certainly going solely to those who perform at the highest difficulties.

Hmm, maybe a 30% score reduction after 20 games in 90 days coupled with a reward 50-100% higher for playing the next higher level of difficulty? Maybe also slightly increase the bonus for map size? I would think that even the most dedicated would have a hard time finishing more than 20 maso gigantic games in 90 days.

Of course, the ladder would still sort out according to difficulty much more stringently than it does now. Those who can't stay on the learning curve after 20 games in 90 days will get slammed when they fail to move up in difficulty. But then, is it reasonable to expect someone to continue to pound the keys if he can see no improvement in his game after 20 tries? Maybe that's not a real-life objection.

Comments, please.

I know the initial 5% reduction is a low number. But I chose that so that people that are stuck on a lower than maso level will be encouraged to keep trying harder, without being penalized too much. And as you said, if someone can't stay on the curve after that many games, learn to duck.

Good point about ppl that submit maso scores. I would suggest that anyone who can play that many maso games deserves to be at the top. I would exclude maso games from the reduction.

Alternately, the multiple submission degradation could also be graduated. 5% reduction after X# of games at crippling, 90% reduction for cakewalk games after the X# submitted.

"StarFleet Battles I" & "StarFleet Battles II" by Prof. Trevor Sorenson; had a very good system on scoring and promotion through the ranks..

Every game you play and win gets harder.. you don't choose the difficulty level.. mearly the size and duration on the game..

you start of and its extreamly easy to beat.. and it lets you get to know the game. and as you win so it slowly increases the dificulty level.

If you loose then you stay at the same level.

This way it doesn't take long till your playing at the highest level that you can .. without cheating or getting dispondent

It would be interesting to see GalCiv using a similar system for the metaverse

but then again GalCiv is a different game

I do wish to make two points. One, this is not all about judging people for flooding the system with tiny games at low difficulty levels. In the end, it comes down to the devs trying to make the newbies feel good by giving them high scores at the outset that do not seem fair to most compared to more difficult games. If the small, simple games were scored low compared to harder games, then the crankers could submit 200 games and no one should complain if the crankers are in the top 10 because the Meta does and should reward commitment. And I am not even mentioning the fatal flaw which is being able to take extreme advantage of a low level AI through tech trading, another fault that is in the dev's hands for not closing that loophole fast enough once they saw the results.

Point two: if what the devs say is true about EBZ knowing ahead of time his strategy was tripping flags and that Stardock people were not wanting to be "on call" to fix anomolous scores for crankers, then EBZ should have seriously reconsidered going public with it and the devs should have said upfront the moment he did go public that it would cause problems. I am not taking sides there on who is telling the truth or not, I am just saying that if there was foreknowledge of anomolous scores due to this method then there was a responsibility on one or both sides of the fence to not encourage this strategy from the outset.

Gerakken,

SD mentioned that the reason why this was not closed is because the AI was designed to be easily taken advantage of in terms of tech trading at that level. AIs at lower levels were designed to be dumb to ease beginners into the game.

~SDC~