|
Post by Dubber on Feb 27, 2008 14:15:16 GMT -5
but if they will have 5 points each, why Vermont is the winner then? did you calculated tiebreakers already? Vermont has 2 kills, Citanest 1 kill -- there's no tie
|
|
|
Post by BioLogIn on Feb 27, 2008 14:24:39 GMT -5
Vermont has 2 kills, Citanest 1 kill -- there's no tie In past we didn't took that into account. We measured points, then tiebreakers. How were these points acquired (via kills or vie surrenders) didn't matter. Proof: a tie between ExDeath and Spacca in December (http://slarty.proboards56.com/index.cgi?board=league&action=display&thread=1196943157&page=4) was solved by Slarty using tiebreakers. Note that they both had 7 points, and one had 4 kills, and the other had only two.
|
|
|
Post by Dubber on Feb 27, 2008 16:13:56 GMT -5
That's all a little esoteric, in my view.
Spacca won in December due to 5 wins (2 of them kills) & 0 losses vs. ExDeath's 4 wins (all 4 kills) & 1 loss.
The simplest way to impartially calculate (which jives with how the decisions were decided) is PointCount (descending) then SurrendersPlusDeaths (ascending) then KillCount (descending).
If we ever end up with tied Points and tied [Surrenders+Deaths] and tied Kills then the breakouts would follow like this:
higher win count wins if tied lower death count wins if tied *then* we would have to go into point differentials and on the obscene chance we're tied out to there *then* we'd have to get into the swiss insanity of which Slarty is so fond.
It's all pretty mechanical until we get to where I think the swiss insanity should start, the swiss insanity will almost always support the higher win & lower death count measures I outline above.
|
|
|
Post by Citanest on Feb 27, 2008 17:14:25 GMT -5
But Dubber, wouldn't that mean ExDeath shouldve lost December by having died?
I really thought if points were tied, whoevers opponents did best wins (although I would guess that either way Vermont wins, as my opponents all finished quite low).
|
|
|
Post by Dubber on Feb 27, 2008 20:10:29 GMT -5
Spacca won December & made our January Rule... ExDeath won January & made our February Rule... Vermont has, I think, mathematically won February & will make a March Rule... Someone please check that no remaining warlock can unseat Vermont (assuming VT does not outright lose his final battle -and- both Citanest & VT do not complete their battles) Obviously, if Citanest wins his last battle & VT doesn't complete his, Citanest would win
|
|
|
Post by vermont on Feb 27, 2008 20:19:34 GMT -5
Technically, emrys could win as well, with three matches left to go.
|
|
|
Post by Dubber on Feb 27, 2008 21:24:42 GMT -5
emrys always enjoys playing the spoiler
|
|
reds
Ronin Warlock
Posts: 31
|
Post by reds on Feb 28, 2008 8:30:55 GMT -5
Actually My battle with emrys is over, see result page, so that problem is solved.
Reds
|
|
|
Post by Dubber on Feb 28, 2008 10:03:53 GMT -5
So, Vermont & Citanest... (for Vermont to win) either both not finish or Vermont to win outright (for Citanest to win) Citanest will need to win or Vermont to lose or not finish
One day to go...
|
|
|
Post by spez on Feb 28, 2008 14:43:00 GMT -5
Even though it clearly doesn't matter since i've lost all my matches this round (ouch), I suppose my game against Rucchys is going to end in a "no count" since even though we started with 10 days left in the month, we only managed to play 13 turns...
|
|
|
Post by Citanest on Feb 28, 2008 15:42:48 GMT -5
Sorry to be a fusspot, but how exactly are we working out the tie? The Slarty add-up-opponent-scores way, or Dubbers way?
Cos if you add opponent scores Awall wins by a mile...
|
|
|
Post by Dubber on Feb 28, 2008 15:55:35 GMT -5
Well, we should really use Slarty's system since the League Battles are technically under his auspices -- but I have no idea the calculations he uses... I'm a liberal-arts major not some fancy-pants coder/mathematician
|
|
|
Post by Citanest on Feb 28, 2008 15:59:10 GMT -5
I thought it was just add up opponent scores. If that's the case:
Vermont = -12 Citanest = -8 Freesoul = -4 Awall = +8
I dunno. I lose any which way you work it, so I leave it to you...
|
|
|
Post by awall on Feb 28, 2008 16:47:06 GMT -5
I think opponent scores makes more sense... going by most kills seems to overvalue kills even more; a kill is already two wins even without contributing to your tiebreakers at all. Going by opponent scores compensates people for having to deal with spaxdeatucker, going by kills rewards them for picking on newbies who don't know when to surrender.
But I sort of have an ulterior motive here, so take my opinions with a grain of salt. :-p
|
|
|
Post by vermont on Feb 28, 2008 16:54:41 GMT -5
Use whatever system you like as long as I win!
|
|