I like the discussion that this topic has generated!  Interesting options you presented M57.  Ultimately I think there should be a point where we agree that a tie really does make sense, which I think is when, in the case of RR, the leaders each beat the other.

So in other words, I agree with Hugh's preference, which is to say that I agree with my own proposal (I know, it's self-serving!).

Ties?  TIES?  This isn't some pansy chess game.  This is war! You know ..with pillaging and all that good fun war stuff.  Do you think Khan, Hitler, or Gadhafi would  settle for a tie?  What's wrong with having a playoff game?  All the real games out there like baseball, football, tennis, basketball and soccer have things like extra innings to solve the problem.  Even Golf!  OMG!  ..Golf!

Oh well, I suppose if we MUST have ties, it's OK.  I guess I'll just have to be ready to share my trophy.  It's a good thing these aren't Oscars.  I'd have to kill the person I'm suppose to share it with so I could keep it on my mantle 24/7. 

BTW, Thanks for the compliment.  I know I can't hold a candle to some of the mathematicians/gaming theoraticians on this site.  I'm just a lowly musician.  At least I'm able to come up with out of the box ideas (some are way too far out, but then you all are very good at putting me in my place)

Hugh wrote:

tom wrote:

So is there a consensus here? If so what is it exactly?

As far as I can tell not much has changed. We like Conan's system:

Swiss: Wins, then Sum of Opponents Scores (SOS), then Sum of Opponents' SOS (SOSOS).

Round Robin: Wins, then head-to-head, then Sum of Defeated Opponents Scores (SODOS).

In the event that these fail to break the tie, it is called a tie.

Though it took some posts to understand each other, it was brought up that there are round robin tournaments that SODOS doesn't break that a further iteration (using the same logic) would break. There is no consensus on usage yet (it is a recent proposal after all).

Tom, is your plan for FeatureID #92 what is listed above, or something else?

Two questions:

1) Does the proposed plan correctly calculate a winner even if rounds are played out of order?  i.e. if some players choose to start /finish all of their games at once vs. a player who only plays one game at a time

2) How would the system handle a tournament like this: http://www.wargear.net/tournaments/view/321? Ender and I both won 1st round, then I beat him H2H, then he beat me H2H.

AttilaTheHun wrote:

Two questions:

1) Does the proposed plan correctly calculate a winner even if rounds are played out of order?  i.e. if some players choose to start /finish all of their games at once vs. a player who only plays one game at a time

2) How would the system handle a tournament like this: http://www.wargear.net/tournaments/view/321? Ender and I both won 1st round, then I beat him H2H, then he beat me H2H.

1) Order of play does NOT matter.  Tiebreaker scores would have to be updated after each round (to see who is currently "winning") but in the end it won't matter when people play, but rather who people played against. Also, the system does not always produce a clear winner.  But ties, I believe, would be uncommon.  See thread for discussion on possible ties (which many of us think is okay).

2) You would win.  You and Ender tied the first tiebreaker, Sum of Opponents' Score (SOS), knocking off Caleb... and falker1976 (who each had lower SOS).  In the second tiebreaker, Sum of Opponents' SOS (SOSOS), you had a slight edge over Ender. 

Conan wrote:

Hugh wrote:

tom wrote:

So is there a consensus here? If so what is it exactly?

As far as I can tell not much has changed. We like Conan's system:

Swiss: Wins, then Sum of Opponents Scores (SOS), then Sum of Opponents' SOS (SOSOS).

Round Robin: Wins, then head-to-head, then Sum of Defeated Opponents Scores (SODOS).

In the event that these fail to break the tie, it is called a tie.

Though it took some posts to understand each other, it was brought up that there are round robin tournaments that SODOS doesn't break that a further iteration (using the same logic) would break. There is no consensus on usage yet (it is a recent proposal after all).

Tom, is your plan for FeatureID #92 what is listed above, or something else?

Yes - I'm going to re-prioritise this one as I think it's important to get this bug fixed asap.

I have written the tie-breaking code for Round Robin tournaments. You can see this in action for all prior tournaments here (non-teamplay for now).

http://www.wargear.net/tournaments/ties/Round%20Robin

Interestingly there is only one tie so far - a pure 3 way tie (http://www.wargear.net/tournaments/view/366)

Let me know if you spot any mistakes. I'll work on the Swiss System tiebreaking next.

tom wrote:

I have written the tie-breaking code for Round Robin tournaments. You can see this in action for all prior tournaments here (non-teamplay for now).

http://www.wargear.net/tournaments/ties/Round%20Robin

Interestingly there is only one tie so far - a pure 3 way tie (http://www.wargear.net/tournaments/view/366)

Let me know if you spot any mistakes. I'll work on the Swiss System tiebreaking next.

So is this tie-breaking and different winners going to be retroactive?

AttilaTheHun wrote:

tom wrote:

I have written the tie-breaking code for Round Robin tournaments. You can see this in action for all prior tournaments here (non-teamplay for now).

http://www.wargear.net/tournaments/ties/Round%20Robin

Interestingly there is only one tie so far - a pure 3 way tie (http://www.wargear.net/tournaments/view/366)

Let me know if you spot any mistakes. I'll work on the Swiss System tiebreaking next.

So is this tie-breaking and different winners going to be retroactive?

I am fine if it will be changed that way for the tournaments to be started in the future, as I believe it will give a more fair result (I never liked the present score calculation). IMO looking for a fair winner in a RR 1 vs 1 tournament should before all take into account the number of times you started to play first (giving the win to the one with less starting first positions). It is indeed a pivotal advantage in these 1 vs 1 games. In case of a tie, then the head-to-head result could be used as a tiebreaker of the tiebreaker.

But nothing should be changed for the finished tournaments or already started ones. It was indeed part of the strategy to win the tournament to take more or less risk in the coming games knowing your own results and the results of the other odds-on favourites against yourself or other players and estimating what the final score will be, especially in the 3 player games RR tournaments.

And it's before all an inviolable principle not to change the rules after the fact.

I partly agree with Toto - I think it would be unfair to strip winners of past tournaments of their titles, even if the system was not fair before.

However it will be difficult to apply a different set of rules for ongoing tournaments vs newly started ones so the new rules will apply to all ongoing / future tournaments at the same time.

tom wrote:

I partly agree with Toto - I think it would be unfair to strip winners of past tournaments of their titles, even if the system was not fair before.

However it will be difficult to apply a different set of rules for ongoing tournaments vs newly started ones so the new rules will apply to all ongoing / future tournaments at the same time.

Fine for me.

I hope nobody thought I was only defending my trophies ;)

tom wrote:

I partly agree with Toto - I think it would be unfair to strip winners of past tournaments of their titles, even if the system was not fair before.

However it will be difficult to apply a different set of rules for ongoing tournaments vs newly started ones so the new rules will apply to all ongoing / future tournaments at the same time.

This makes sense.  On-going tournaments don't have a winner yet so the rule change should be invisible.

You can see how Swiss System tournament ties are resolved now - http://www.wargear.net/tournaments/ties/Swiss System

There are no ties created using a combination of SOS and SOSOS which is great news.

tom wrote:

You can see how Swiss System tournament ties are resolved now - http://www.wargear.net/tournaments/ties/Swiss System

There are no ties created using a combination of SOS and SOSOS which is great news.

After looking at this chart, I really think this is a good system. 

The tournament that I was mad about losing, I would have won:(http://www.wargear.net/tournaments/view/13)

and the one that I backed into the win on, I would have lost:
(http://www.wargear.net/tournaments/view/324)

So this system seems to get it right.

So nearly all of the Swiss System style tournaments have different winners when looking through the new system lens?

In the Tournament "Captain Planet" I am credited with them win under the old system. In this system Red1 is credited with the win.  I lost to Red1 in game 1, then won the next two games.  In particular in game 3 I beat Red1 this time around.  From reading this chart it seems to me that no matter what I did in game 3 I could not win the tournament?  Does "strength of schedule" come into play here then?  Seems a little weird -

Amidon37 wrote:

In the Tournament "Captain Planet" I am credited with them win under the old system. In this system Red1 is credited with the win.  I lost to Red1 in game 1, then won the next two games.  In particular in game 3 I beat Red1 this time around.  From reading this chart it seems to me that no matter what I did in game 3 I could not win the tournament?  Does "strength of schedule" come into play here then?  Seems a little weird -

Yes, the idea is to give the win to the person with the most difficult strength of schedule. Strength of schedule is measured here by the sum of your opponents' scores. (In the even that is tied, you sum your opponents' sum of scores.

so assuming someone (Djembe) doesn't go and win 11 games who would win this tourney (http://www.wargear.net/tournaments/view/281)?

weathertop wrote:

so assuming someone (Djembe) doesn't go and win 11 games who would win this tourney (http://www.wargear.net/tournaments/view/281)?

Good question - One thing that would be good is to have an updated SOS, SODOS, etc on the tournament page. With that information, you could get a clearer picture, but it might depend on games yet finished.

Another question: Are we treating 3-player round robins like Swiss or like 2-player round robins?

this is all assuming that the new changes are taking effect to tourneys in progress and not just new ones (which i thought i read somewhere).