I think it's funny that the "strategy" is to win fast and play slow the good players you are more likely to lose against, when it's also been said that it's better to lose early and win late. Not to mention that just because you have the multiple rounds turned on doesn't mean you are going to play multiple rounds, or many multiple rounds in which you really get to choose your opponents. Hence I really do think it is going to be tough to "predict" in which way to play your games prior to knowing the outcome of them.

I'd say it's fairly obvious that which order to win/lose, if not then player's don't understand basic Ranked games either.

Maybe tie should just go to the person with the highest Win%. (Joking!)

I don't think it matters as much whether or not there is a "strategy" to playing the current system, more that is capricious in that your score is determined somewhat by luck.

I will put a +1 in for deciding the tie by the head-to-head and, if that (very rarely) does not decide it then playing tie-breakers as needed.

And if you are looking for ties in the record there was one in this tourney;

http://www.wargear.net/tournaments/view/71

Even though I won the head-to-head against Blackdog it was the longest back and forth I have played on that board. I would have liked to declare the tourney a tie.

So I quit reading at some point and want to add two things:

First my comment before was intended to suggest that whatever tie-breaker you come up with will be unfair by some measure (so why not just roll a hundred sided die for all the tied players and declare the victor by highest first?).

Second, it seems that everyone likes the scoring in swiss where it is done round by round, but not in RR as fast players may be punished as all there opponents could have score of 10000 ... .
Seems like the easiest thing to do is wait for each successive round to finish prior to taking into account games of future rounds. So the tournament score will change when each round is finished instead of when each game is finished (this means you may have to wait until the end of the tournament for the score to populate).

Alpha wrote:
Seems like the easiest thing to do is wait for each successive round to finish prior to taking into account games of future rounds. So the tournament score will change when each round is finished instead of when each game is finished (this means you may have to wait until the end of the tournament for the score to populate).

This makes sense to me.. There's no advantage to slow playing either.

Alpha wrote:  but not in RR as fast players may be punished as all there opponents could have score of 10000 ... .

Punished or Rewarded as opponents would have a score of at least 10,000 instead of less than 10,000.

Alpha wrote: Seems like the easiest thing to do is wait for each successive round to finish prior to taking into account games of future rounds. So the tournament score will change when each round is finished instead of when each game is finished (this means you may have to wait until the end of the tournament for the score to populate).

This will take away the ability to strategically play your way through the points, but does not really help the points determine who played better and deserves the win in a tie.  Might as well flip a coin.

Alpha wrote: First my comment before was intended to suggest that whatever tie-breaker you come up with will be unfair by some measure (so why not just roll a hundred sided die for all the tied players and declare the victor by highest first?).

That's why we should all agree on some fixed criteria for establishing a score used for tie breaking. So far we come up with a modified version of the Buchholz score, which should work fine, except in one case (see above). Then play-off round should do the rest.

Alpha wrote: Second, it seems that everyone likes the scoring in swiss where it is done round by round, but not in RR as fast players may be punished as all there opponents could have score of 10000 ... .
Seems like the easiest thing to do is wait for each successive round to finish prior to taking into account games of future rounds. So the tournament score will change when each round is finished instead of when each game is finished (this means you may have to wait until the end of the tournament for the score to populate).

And that's why the actual score system is unfair for RR tournaments. What should matter is the final result, not living to play wargear tournaments. I'd like to have to worry only about playing to win, not planning my schedule to take turns at specific hours. Despite being an active player, this is not a full time activity, and I don't want it to became one (and this is the reason why I can't stand complains about the vacation, as seen in other threads or in some game messages. But this is not the topic here).

(=

Yertle wrote:

Alpha wrote:  but not in RR as fast players may be punished as all there opponents could have score of 10000 ... .

Punished or Rewarded as opponents would have a score of at least 10,000 instead of less than 10,000.

agreed, that is why I do not believe anything is necessary, but at least making so that scores are calculated in the order of the rounds and not the order of the games takes away any strategy or worry, just luck of the schedule.

Tesctassa II wrote:

Alpha wrote: First my comment before was intended to suggest that whatever tie-breaker you come up with will be unfair by some measure (so why not just roll a hundred sided die for all the tied players and declare the victor by highest first?).

That's why we should all agree on some fixed criteria for establishing a score used for tie breaking. So far we come up with a modified version of the Buchholz score, which should work fine, except in one case (see above). Then play-off round should do the rest.

I don't see the need for a different scoring system, but I am also indifferent to one being implemented and did read the above situation for when Buchholz breaks down.  The current system with my above suggestion may not take all things into account, but no system will.   It seems that introducing a new system when we already have an easy to understand and used elsewhere system around does not make sense to me.  Again, whatever is the consensus, I don't really care what is used as long as it is decided upon.

Alpha wrote: I don't see the need for a different scoring system, but I am also indifferent to one being implemented and did read the above situation for when Buchholz breaks down.  The current system with my above suggestion may not take all things into account, but no system will.   It seems that introducing a new system when we already have an easy to understand and used elsewhere system around does not make sense to me.  Again, whatever is the consensus, I don't really care what is used as long as it is decided upon.

The need is that RR tournaments actually don't use that kind of score anywhere (=

And the reason is quite simple. In swiss tournament you need alternative scores because you don't play against all other players, hence you need a way to measure the strength of two tied players, since tied players couldn't have played against the same opponents, hence one of them might have encountered weaker opponents.

In RR instead, the concept is that since you play against all the others, if you end up tied with someone else, then it means you have equal strength. Other scores are used only if it's not possible to have tied winners, but those scores must be as neutral as possible, that is they shouldn't depend on the round order or things like this, since this wouldn't respect the idea behind the RR system. In other words, instead of the current system, in case of tied players, we could randomly pick one of them and it'd be the same.

That's why for RR tournament the actual alternative score should be changed.

(=

There is still the issue in RR tournaments that three people have the same record at the top (say A,B, and C have a 7-2 record). In their individual games A beat B, B beat C and C beat A. How will this be decided?

I called the proposed tie-breaker goes to whoever won the one-on-one a one-off solution because it doesn't always apply and doesn't always decide. So something else is needed.

If something else is needed, then to avoid complicated decision rules, the something else should just be the decider. I do not agree that the current scoring system is irrelevant or equivalent to a coin flip. It does take into account the strength of your oppenents at the time you played them.

The fact you're missing is if I play my games faster than you and all my opponents that are going to lose, haven't yet, they still have more points than they will when you play them. Therefore you get less points for defeating them and I get more. My current "strength" would be more and even though our record may be the same I just played faster and you took your time.

While we may not have the answer to a 3-way or more tie situation at this point anything is better than the random "coin flip" system that's in place now.

Alpha wrote: There is still the issue in RR tournaments that three people have the same record at the top (say A,B, and C have a 7-2 record). In their individual games A beat B, B beat C and C beat A. How will this be decided?

I called the proposed tie-breaker goes to whoever won the one-on-one a one-off solution because it doesn't always apply and doesn't always decide. So something else is needed.

The modified version of the Buchholz which count only the opponents a player has won against should work fine.

Alpha wrote: If something else is needed, then to avoid complicated decision rules, the something else should just be the decider.

That's also true, but I believe that when you have to decide criterias for such things, you need to evaluate also the probability that a scenarios where one of the criteria fails has to happen. This because, as you said (and I guess we all agree on that), there isn't a method that works in every situation and/or that satisfy everyone.

That said, I think that if criteria A fails 5% of the time, then it's good, and you can decide to apply another criteria, say B, which also fails 5% of the time. So the probability that both fail is extremly low. And if they do, then it's acceptable to flip a coin or repeat the game or games (such a playoff as proposed above).

Otherwise you can find another criteria, much more complicated which fails only 1% of the time, or even more complicated that never fails (but I think it's reaaally hard).

Alpha wrote: I do not agree that the current scoring system is irrelevant or equivalent to a coin flip. It does take into account the strength of your oppenents at the time you played them.

The problem is that ability of a player isn't something that changes within a tournament (in my opinion). And furthermore, the score actually doesn't measure the strength, but how much you win and, indirectly, how much you're opponents did when you played them.

That's why this system it is criticised and I said it is the same as flipping a coin. Because in the end it matters which order you played the others.

(=

Viper wrote: While we may not have the answer to a 3-way or more tie situation at this point anything is better than the random "coin flip" system that's in place now.

I think we do have a solution for every tie situation.

(=

Viper wrote: The fact you're missing is if I play my games faster than you and all my opponents that are going to lose, haven't yet, they still have more points than they will when you play them. Therefore you get less points for defeating them and I get more. My current "strength" would be more and even though our record may be the same I just played faster and you took your time.

While we may not have the answer to a 3-way or more tie situation at this point anything is better than the random "coin flip" system that's in place now.

They will not get it until it happens to them.

Viper wrote: The fact you're missing is if I play my games faster than you and all my opponents that are going to lose, haven't yet, they still have more points than they will when you play them. Therefore you get less points for defeating them and I get more. My current "strength" would be more and even though our record may be the same I just played faster and you took your time.

Lol again, or you play fast, win fast, then lose later and end up losing more points than had you played slow, lost earlier, won later ended with more points...

Alpha wrote: but at least making so that scores are calculated in the order of the rounds and not the order of the games takes away any strategy or worry, just luck of the schedule.

I could get on board with that too I think, that way you do kind of alleviate the strategy as to what order you play your multiple rounds at once games.

Yertle wrote:

Viper wrote: The fact you're missing is if I play my games faster than you and all my opponents that are going to lose, haven't yet, they still have more points than they will when you play them. Therefore you get less points for defeating them and I get more. My current "strength" would be more and even though our record may be the same I just played faster and you took your time.

Lol again, or you play fast, win fast, then lose later and end up losing more points than had you played slow, lost earlier, won later ended with more points...

I'm trying to figure this out. Viper's argument makes sense to me.

If you play fast and win, then yes, you will play against players with lower scores in later rounds but isn't this preferable?  When you win or lose in games with more than 2 players, you always win more than you lose, right? Using this logic, I would rather win more and lose more than win less and lose less.

Obviously, in two player games, this argument doesn't hold water.

Rules should include first to the top wins, so even if two or more people have the same score in wins, the first one there is the tournament winner.

CiscoKid wrote:

Rules should include first to the top wins, so even if two or more people have the same score in wins, the first one there is the tournament winner.

So you want to reward quick play?  Or are you specifically suggesting this for Swiss System tournaments where at the end of the tournament; if the top two players have a 2-1 record, the player that was 2-0 should win and the player that was 1-1 should lose.

Currently, with four players per game, there should be a 2-0 player 13000 points and 1-1 player with 11000 points that are fighting for the tournament win.  The 1-1 player would actually take the lead by beating the 2-0 player.  This reinforces the need for a better decision method for ties in the Swiss System, but does not work all of the time and still leaves the need for a tie breaker decision system.