Could you prove the opposite, that two games are less effected with chance and randomness if we only keep inherent civ randomness and get rid off random events?
The burden of proof for "that argument is baseless" is not the same as the burden of "this argument isn't baseless". In other words, I'm merely pointing out that people who want them due to "on by default" or asserting "minimal impact" are asserting opinions based on nothing. IMO this isn't fair ground, and while I could say "they have tons of impact", it wouldn't be any less baseless, but it wouldn't be any more so, either.
To get back to previous discussion with Jesusin, more random elements will give a luck percentage with less dispersion. OTOH, absolute values are higher. For example, are 2 games more different in terms of turn difference with 0 vs 1 mine pops or 1 vs 2? This is not so clear at all and no pencil and paper discussion will solve this issue. Unless somebody thinks of a clever test, we won't have an answer to this. I wonder if Firaxis did some serious testing to determine how much impact RE have or don't have on the games?
This is a good point. There are differing beliefs where events lie. For example, if the random-ness is low enough, events have a material impact on the absolute outcome possible without creating enough dispersion. If random-ness is prevalent enough, their relative impact is diminished. I am arguing the former is the case, but that's just based on what I've witnessed. As you've pointed out, that isn't necessarily the case (and we'd have a hard time knowing without testing that's so far not been done). I'd argue the MOST LIKELY thing in this paragraph I'm quoting is that fireaxis didn't do that degree of testing
. HoF nerfs out some events already (indicating experienced players felt stock events imbalanced very quickly after their release), and bugs/glitches like the ones for favorite civic bribes, diplo interface lies, etc suggest fireaxis likely did not put utterly extensive time into testing events...certainly not to the extent required to balance them in MP! I can't think of anybody who put a 1.0 release on anything but the most simple of features that actually had balance, even in games like Warcraft 3 where beta testing was extensive!
My intuition is saying me that relative values have significance, but that's of course not hard proof. Thinking in DoW terms again, it's quite likely that 2 games would be more different in 0 vs 1 DoW since the team not getting DoWed can check WHEOOHRN all the time and keep military low and say adopt Pacifism for GPP, while in 1 vs 2 case both teams have to build an army and research will get hurt. But Civ is to complex to answer this question just like that.
I agree...this could even vary game to game (gandhi won't declare unless you have your head up somewhere it shouldn't be...close spawn shaka on an otherwise isolated continent is all but a lock...other scenarios are more chancy)...definitely not something we'll solve just by discussion...probably never.
If we don't have enough random events or if they are to weak to be statistically significant, then there is no real case for either option. Then it doesn't matter if they are off or on, then it comes down to personal preference. It's the claim that games are effected less with RE off that started the discussion and that's a statistical argument.
And one that remains unsolved either way, unfortunately
. I continue to contend that it's a valid argument based on not only my own but others' game submissions (primarily on strategy and tips), but an unbiased look would be spectacular (and if accomplished could completely 180 my outlook on events...if accomplished).
Play the same map 11 times, alternating with RE on and RE off. Throw out the first three games. Average the results of the RE-on and RE-off games and compare. You only have to play to about 1000AD (after that, the effects of RE are so miniscule its like... who cares -- and interestingly those who care most are probably the folks who win every game before 1000AD anyhow).
You'd probably need more games than 8 to get a true picture. There are tons of holes in this suggested methodology though. Some that come to mind:
1. compare to what? If a RE on or off game finishes 1 turn ahead/behind, does that count? 10 turns? 50? How do you separate out the other RNG effects? Might be hard. AW would remove some of the most powerful events - namely the +3 diplo ones.
2. A lot of winning BOTM submissions (or at least placing) come well past 1000 AD. Diplo ones that allow trade or a diplo win would obviously have an impact beyond miniscule. This is the same ridiculous assumption used throughout this thread. While I don't have evidence to prove a counter claim, you certainly lack evidence to make it in the first place.
3. Player bias WOULD be a major issue. Comical as it sounds, it may actually be better to us an AI "player"
. Horrible quality play is still uniform, and you wouldn't have to throw maps out! Haha! You'd need to run an incredible amount of test games though. Who has a super computer?
