I am sick of it

Ksero · Post by **Ksero** » Wed Apr 18, 2007 7:57 pm

spideycw wrote:QUOTE (spideycw @ Apr 18 2007, 09:32 PM) My reasoning for why I hate AB (99.9% of my hate is directed toward small games) is that oft times having a greater number of players is a bigger advantage than being up 15 helo which is in fact not balance at all but a stack to the team that is up 3 players.

[...]

It also sucks for those of high rank because If I could get Aarm or say 8 of CoD or KGJV for that price I'm much more likely to accept the lower rank more numberous players thus denying the higher ranked players a chance to play.

That CoD and KGJV are undervalued are mainly because they haven't played many games. Eventually they will get a more appropriate rank. It's just a matter of time (or recalcing the stats with all the games in the archives). Also, as has been previously said, since they don't play very often, they don't skew the results that often.

That issue aside, would you say that the biggest problem is that novice players are undervalued, or that veterans are overvalued? Because both of those can be adjusted by tinkering with how quickly / slowly people gain rank. If veterans are overvalued, then we should make it harder / slower to get higher ranks. If novice players are undervalued, then we can have them advance faster.
Of course, I would like some hard data before I would support such an adjustment. Say about fifty auto-balanced games, and see how often newbies win, compared to veterans.

Greator_SST · Post by **Greator_SST** » Wed Apr 18, 2007 8:12 pm

...cadillac is exactly right. Unless autobalance is used, current ranks are largely a reflection of a player's ability/desire [edit:added desire] to choose the stronger team. And current player rankings will never change unless AB is implemented. The overvalued vet will continue to remain high, because he/she will always pick the stronger team. The undervalued novice will continue to lag far behind UNLESS he/she embraces the stacking mentality and begins to choose the stronger team.

So you see, what you're witnessing in all these threads is a titanic struggle for survival. For stacking to continue to exist, it must resist at all costs ANY move away from imbalance. Because imbalance is what feeds it and what nurtures its young.

Cadillac · Post by **Cadillac** » Wed Apr 18, 2007 8:34 pm

We must make a choice.

We can either keep our freedom, or we can get rid of stacking.

There will be uproar.

There will be bloodshed.

We must fight for whichever cause we see as the just one.

Fight for freedom, freedom to choose, freedom to stack.

Or fight for equality, fairness and balanced games.

The future of Allegiance is in your hands. Speak now, or forever be silent.

Lykourgos · Post by **Lykourgos** » Wed Apr 18, 2007 8:39 pm

Ksero, two things.

Once, that linear fit doesn't actually fit so well. There's... a lot of variance there.

Two, someone could be winning 1.2 or 1.3 games per loss just because they are that good. Weedman, for instance, really does matter at least as much as 5 average pilots to my chance to win in a standard-sized game.

Certainly I admit that anyone winning 2 and 3 games per loss is stacking.

Edit: Rav, I want a fourth category. Stacking is bad, but inevitable, and I think the best way to fix it is by indirect methods, not by having an algorithm decide which team you are allowed to join. Algorithms are 1. easy to beat and 2. wrong often.

Ksero · Post by **Ksero** » Wed Apr 18, 2007 9:22 pm

Lykourgos wrote:QUOTE (Lykourgos @ Apr 18 2007, 10:39 PM) Once, that linear fit doesn't actually fit so well. There's... a lot of variance there.

Certainly. R2 = 0.59. It just shows a general tendency. A slight correlation between stacking and helo ranks.
QUOTE Two, someone could be winning 1.2 or 1.3 games per loss just because they are that good. Weedman, for instance, really does matter at least as much as 5 average pilots to my chance to win in a standard-sized game.[/quote]
Hmmm... I apologize if I'm sounding like a broken record now, but I really don't agree with this.

Okay... how about this? We clone Weedman, and let his clone play on the opposite team. Would both Weedman and his clone achieve a win/loss ratio of 1.2 - 1.3? No. If the teams were balanced before the Weed's joined, then both Weed's would have approximately a 1:1 win ratio. So in a perfect world, where all games were balanced even Weedman would win about as many games as he lost.

So the reason that Weedman can achieve a win/loss ratio that's higher than 1 is that when he joins a team, the game becomes unbalanced. If you win more than you lose, then the teams you play for are generally better than your opponents. Which means that your team is better. Which means that your team is stacked...

If you don't agree with this, Lyko, can you say what part of this argument that you consider faulty?

And certainly, we can't require everyone to win exactly 50 % of their games. A certain natural, random variation can be expected. But over time, the win/loss ratio should ideally converge on 1:1. Hmmm... I'll have to dig up my statistics book to check how to calculate just how far the win/loss ratio can deviate before we can reliably say that a pilot is stacking. Student's t-test anyone?

Lykourgos · Post by **Lykourgos** » Wed Apr 18, 2007 9:24 pm

let me put it this way, ksero; would you expect a correlation between pilot win/loss and pilot skill? (assuming you are god and know pilot skill).

Reading your post in more detail; I agree with you, I'm just saying that it is very hard for an extremely good pilot to join a team without creating a small stack.

Jormagund · Post by **Jormagund** » Wed Apr 18, 2007 9:42 pm

see and this is where DW's point comes in. At what point do you call stack?

Say, both teams are dead the $#@! even, you want to join one- you suck- you join and then someone mediocre joins the other side but, because you suck, there is now a *small stack* against you. Or vice versa, you rock (or happen to be uber like myself- j/k) you join one of the two dead even teams and now there is a *small stack* in your favor. You are a hopeless stacker.

WTF. WTF. WTF WTF WTF. So what Ksero & co. (otherwise known as the ppl who get pwned cause theyre not good enough to stack) would have us believe is that no matter how good we are we shouldn't be allowed to make a difference because we must make certain that someone equally as good is on the opposing team. I'm sorry, I don't follow. The whole point of playing and getting better is so that at some point you begin to win consistently bc you are capable of beating your opponents. Weedman joining one of two dead even teams does not creat a stack. It creates an edge. A stack is when one team is incapable of standing up to the other. Stacked games by default cannot last longer than 45 minutes and often resolve themselves within 20.

You are all a bunch of whining ninnys. OMFG they are up one player STACK. Seriously, knock it off. This whole "graph madness" as I heretofore will refer to it is just that- madness. You guys are flipping retards. A 1:1 win ratio is a statistical myth. Let's all become cogs within the great machine that is equality(mediocrity) and do our best to achieve the holy of holies- the 1:1 ratio. $#@!tards.

Maybe we can alleviate stacking by flipping a coin to see who wins?

Ksero · Post by **Ksero** » Wed Apr 18, 2007 9:52 pm

Lykourgos wrote:QUOTE (Lykourgos @ Apr 18 2007, 11:24 PM) let me put it this way, ksero; would you expect a correlation between pilot win/loss and pilot skill? (assuming you are god and know pilot skill).

(I see that Lyko has changed his mind... But I've already typed together all this, so I'll post it anyways, in case someone finds it enlightening.)

No, I wouldn't expect that, if (and that's a big if) all games were even (ie. stacks didn't exist).

Your win/loss record isn't related to your piloting skill. It's dependent on how good the teams that you fly with are compared to the opposing teams. If you generally fly with even teams, then you can be the best pilot in Allegiance and still win about 50 % of your games. Conversely, if the best pilot in Allegiance always had to play with a team of (0)'s, then he would most probably lose a lot of his games.

How about this? Suppose that Weedman wins more often than he loses. Then we gradually make his opposing teams better and better. Eventually he only wins 50 % of the games. Wouldn't this be a fair game? When each side has 50 % chance of winning? Which would mean that the games were originally unbalanced.

If Weedman wins 1.2 games for every loss, then his contributions to a team is generally undervalued when the teams are balanced, and he should play with worse teams against better opponents.
QUOTE Reading your post in more detail; I agree with you, I'm just saying that it is very hard for an extremely good pilot to join a team without creating a small stack.[/quote]
Yes... Good pilots might not intend to stack. But they might attract a stack anyways. :\

EDIT: Jormagund's inflammatory post taken to PMs

apochboi · Post by **apochboi** » Wed Apr 18, 2007 9:53 pm

If you believe autobalance seriously is the way to go with "antistacking" then well, perhaps you should limit the rewards for stacking already people get less helo, perhaps lower what they get further or something. Its not about changing the game it's about changing the attitudes, i just dont like the idea of having to play for someone i dislike or dont respect as a decent commander.

Post by **BlackViper** » Wed Apr 18, 2007 9:57 pm

Sneaks in one item to point out. Teams can be formed up. THEN turn on AB and launched. You can then decide what commander to fly for.

Believe it or not, I also agree I do not wish to be forced to fly for certain commanders. I either sit out or wait for the other team to have an opening.