Prepare for the AA tourney (in FGF07)

[libervisco]

Well looks like some may be ignoring the last thread, perhaps because of losing faith in this tourney ever happening. Who knows, I'm just speculating. It doesn't really matter.

What matters is that the Armagetron Advanced tournament as part of Freedomware Gamefest 2007 is actually proceeding. I believe I found a good solution to the most controversial issue so far: rubber.

It's not any genius of course. It is summed up like this:

"In qualification matches, two times will be played on low rubber (5) and once on high rubber (15 or more) and in the final stage once low and once high. Alternatively, those involved in a match can vote whether to play high or low rubber, even to the exact number."

So the default is some qualifier matches low and some high. Most are low because that's what most players seem to prefer. But if you want to be even more granular you can just vote. If your group happens to be all 4 low-rubber people then that's what you'll have. Same if it's 3 low-rubber players and 1 high rubber.

This way we don't impose one way or another.

You can check out the updated set of rules here: http://www.libervis.com/wiki/index.php? ... Tournament

And I still welcome some final feedback for adjustments if necessary, at least for this weekend. Then I'll be preparing servers to play on, form groups of those who signed up and proceed to schedule first matches with those groups.

So, please don't write this off. It might be a quick little tourney, but it may be fun and the winner gets to win a prize. In fact the winner also becomes eligible for the big prize. You can increase your chances for that by participating in one of the other tournaments in this gamefest.

As for why it took so long to get something done with AA.. well we have a multitude of other tournaments. OpenArena and Nexuiz 1vs1 and CTF and now Tremulous. Almost all have already begun or are about to begin matches. We're well on our way. No reason to leave Armagetron Advanced out of the party when we have *at least* 12 players signed up for it.

Cheers!

And not to forget the last but not least, invisible maid-bots are reported being detected as already sweeping the floors of Armagetron spaces where advanced tron riders will engage each other in a light speed race of reflexes. -- Reporter

[libervisco]

I'm trying to rechecking the rules linked above, just to make sure everything is clear.

I'm not yet 100% clear on the thing about the one winning the match being the one reaching 10 points first because that's the idea I picked up from someone in the AA community (QBrain actually). I'd assume it is one who wins 10 rounds first, so in a match we play rounds until any of the players win 10 of them.

Alternative way could be to set a fixed number of rounds and then just count who won most of them when we're done.

[Lucifer]

The problem is that winning a round in armagetron means surviving to be the last one. So if you count the number of round wins to determine match winner, you encourage camping.

QBrain's idea is right, and standard fare. Count core dumps, the match winner is the one who reaches 10 first. Or use timed matches, but don't give any points for a round win, only for a core dump.

There are some specialized arenas where the last person surviving is probably the one who did the best, like in a sumo game, but regular ol' last man standing is best when you give points only for core dumps.

[libervisco]

I see, so points refer to core dumps. I think that makes sense. I'll edit the wiki to reflect that.

You're indeed right, that's the best way to prevent camping and yet is fairly clear.

Btw, core dumps are counted in game? I think I saw that the score in-game is actually a number of core dumps which gets reset on every round. This leads me to ask, is it possible to set AA to make an automatic screenshot on the end of every round?

Btw, so to recap. The match winner in this AA tourney would be one who reaches 10 core dumps across all rounds (not just 1 round).

However, since the rule is not fixed just yet (I just sent an email allowing everyone to reply back with comments) alternative way would be to have the one reaching 10 core dumps first be the winner of a round and then after 10 rounds the one who won most of them is the winner of the whole match. This would make a match longer.

Of course, another simple way to make a match longer is to just increase the maximum number of core dumps.

Thanks

[ed]

Btw, core dumps are counted in game?

you could use settings something like these:

Code: Select all

LIMIT_ROUNDS 100000   # You would have (effectively) infinite rounds.
SCORE_KILL 1   # 1 point for each kill.
LIMIT_SCORE 10  # First to 10 points wins.
LIMIT_TIME 30   # match will last no more that 30 mins.
SCORE_SUICIDE -1  # If you're in a rut, suiciding (in order to prevent giving points to another) is really not in your favour.
SCORE_WIN 0  # no points for winning a round

[libervisco]

That looks good ed. I'll definitely use it as a reference. Thanks.

We didn't get all of people respond to my email though and the deadline for that is the end of this week (pretty much only one day left). I could perhaps assume that anyone who confirmed before will participate despite not replying to the last email, but then again I specified in the email that replying is a final confirmation of participation while not doing so may mean that they're out. :S

I hope more people respond. Everyone should respond at least with their location which helps with forming groups.

[ender]

I could perhaps assume that anyone who confirmed before will participate despite not replying to the last email...

Yeah, I wouldn't assume anything with the a.a. community or whatever... You should get confirmation for sure. Players always make promises that they end up not being able to uphold due to real life and whatnot.

[Lucifer]

Speaking of not replying, I've had my head buried in code again, sorry.

AA tracks score between rounds just fine. Score gets reset every match. The settings ed provided are what might be considered the standard, if there is such a thing, for last man standing. The only thing is I wouldn't penalize suicides. It's hard to detect exactly who killed whom, and generally you want it to err on the side of crediting a suicide rather than giving someone a free point they didn't earn, even if occasionally you don't give someone a point they did earn, and penalizing a suicide is basically penalizing the player for the game's error. In a perfect world, I whole-heartedly agree with penalizing suicides and teamkills (which I didn't talk about). This is not a perfect world, however.

[libervisco]

I'll consider your opinion Lucifer. Thanks.

RIght now I'm still concerning myself mainly with doing best groups as possible with who replied and confirmed so far, which are:

goodkoen - EU
epsy - EU
free-zombie - EU
zion - EU
Lackadaisical - EU
Durka - US
mkzelda - US
ender - US (edited)
drågøn -

MaZuffeR - EU (only after January 2nd)

I've written more on this here: http://www.nuxified.org/topic/organizin ... ment-11727

Basically the issue is that having groups which are half EU and half US is an issue due to ping so it seems we'd have to do some balancing the odds thing with the ping like playing same match in two parts on two servers (one US and one EU). Also the number of people who can currently participate, since MaZuffeR is only after January 2nd, is 9 which can only be split into 3 groups of 3 rather than ideal groups of 4.

Also, please please ender and drågøn, since you so nicely forgot to mention where are you from (even though I asked this in my email), tell me your location! It is really important as you may see!

EDIT: OK, ender is from Texas, US, as seen on his profile here. I emailed drågøn about his location again, but something tells me he's in EU (look at the weird letter.. let me guess.. dutch! ). This means we have exactly 3 US people.. maybe 3-player groups might not be such a bad idea hum?

Thank you!

[libervisco]

OK I'm seriously considering 3 groups of 3 players. As explained here it wouldn't change much and it would work out.

So one group would be completely US (durka, mkzelda and ender) and the other two would be EU whose members can be randomly selected.

So it could look something like this:

Group 1:

goodkoen
zion
free-zombie

Group 2:

epsy
Lackadaisical
drågøn

Group 3:

Durka
mkzelda
ender

Behold the selection of the light riders! What do ya think?

[libervisco]

Looks like Drågøn is from US after all. So we have 4 people from US and 5 from EU which isn't good for groups of 3 nor groups of 4.

If we wait for MaZuffeR (January 2nd) we'd have 6 from EU and 4 from US which still isn't good because one group would end up having one EU player among all US.

It would be great if 2 more players could sign up for this then before January 2nd! Either that or someone from US join now so we have 5/5 ratio between US and EU.

Thanks

[libervisco]

Groups defined: http://www.freedomware-gamefest.com/con ... tournament

Rules: http://www.libervis.com/wiki/index.php? ... Tournament

Scheduling topic: http://www.nuxified.org/topic/armagetro ... scheduling

But I will be watching this topic as well.

So yes, we're finally getting this started. Please come to the scheduling topic to specify your availability and when best to do a match, as well as your rubber vote.

I'll make sure to set up servers.

Cheers

[Luke-Jr]

I'm not listed...

[libervisco]

I listed you now and then changed groups to be of 3 players (four groups) since you make it 12.

Everyone please let me know if you object the way groups are set now.

Thanks

[Concord]

group 2


wow