ERDoll wrote:
Some posts to the forum thread call Swiss pairing a lottery.
No not the Swiss pairing - the Swiss system as a whole is a lottery. The pratical examples of weaker performing players finishing above stronger performing players are numerous. It is called leapfroging and is a consequence of the swiss system being totally inappropriate for our game. Controversies like this will continue to pop up as long as we don't address the "worm at the core". There are numerous tournament forms were placement is guranteed to depend only on the individual players own performance not on luck of the draw and where colluding is impossible.
ERDoll wrote:
The Swiss system is a compromise attempt to produce a fair tournament when all players cannot play each other. It is widely used in checker tournament play.
If the Swiss system is used because people thinks it is fair something is REALLY wrong. I don't think that is the case though - I think it is used because other factors have higher priority than fairness e.g. ease of running the tournament, ensuring everybody plays maximum number of games etc.
ERDoll wrote:
It is a terrible fear of a tournament director that a mistaken pairing will create controversy.
Wouldn't it be nice if tournaments where run with a system where pairings don't matter so much and where what mattered were the individual players own performance ?
ERDoll wrote:
Some have suggested that we require recording and submittal of all games. Since we do not publish our tournament games, recording games has never been a requirement in Illinois. While doing so might have provided additional evidence in this situation, I just see it as an unnecessary burden on players and don’t recommend that recording of games be required in the future.
When a group of players meet and play checkers and don't use a clock or don't record the games it should not be called a tournament but a practise session or a club meeting. No titles should be given out on the basis of practise games and no ratings should be changed.
Since you asked The " A " Class at this past week-end's tournament requires ACF Membership to play in the Masters. Dr. Beckwith and Kim Willis and myself have received several emails and the three of us as ACF Executive Officers are engaged in discussion . All I can offer for now..............................................
Sincerely:
Alan Millhone, President
American Checker Federation
Greetings Alan,
Hope all has been well.....
One thing which I hope you and the rest of the EC strongly consider during your discussions......
If the allaged conduct at Illinois is indeed true and accurate, and there are no consequences,......
this goes the whole nine yards towards perpetuating the myth (if it is a myth ??!!??) that the ACF, and organized Checkers in America (in general) are nothing more than a "good old boys" club.
How could we ever hope to attract the youth of our country to this game with such a stigma attached to it...? Corruption from within has destroyed many an empire............
Regards,
Jay "Loyal to the game" H
Jay,
your comments are good,
"If the allaged conduct at Illinois is indeed true and accurate, and there are no consequences,......"
Now answer
"If the allaged conduct at Illinois is indeed untrue and in-accurate, and there are no consequences,......" what then? It has already been shown by Nick that his comments have not been completely truthful\accurate, what then, do we just continue to write this off as immature youth and is acceptable?
Ken
Regards
After being Beyotch slapped by Bugs I am afraid to post....I feel like a yard dog who snuck into the house and got caught sniffing around the kitchen stove.
My tail is firmly tucked between my legs, and I am now back in the yard where I belong.
As I posted before,
I am,
Jay "sticking to the game's history from this point forward" H
PS....Regarding this situation, I have sent my thoughts to Alan Millhone who is an ACF EC member.
In general, I believe the system itself is flawed when there is a State Championship Tournament, held to determine who is the State Champion, and in that tournament there are only 4 players from that State, and they do not all face each other ...... It seems to me that in a tournament where there are only a few players from that State, they should all play against each other during the normal course of the tourney, and only their results against each other should count in determining who wins the Title of State Champion.
Well said Jay. A little bit of logic and common sense go a long way.
Incidentally, in the Swiss System, the Maximum Compression rule is an inherent feature of the Pairing Methodology,
and not an optional add on. It follows inevitably from the basic pairing rules:
1. No repairing (before the final round or rounds).
2. After the first round players should be paired with opponents on equal points,if possible, subject to Rule 1.
3. In pairing, no dropping down in the points table should occur, except where forced under Rule 1.
So, taking a typical example:
Player A on 10 points
Player B on 9 points
Players C and D on 8 points
Assuming A and B have already played each other, then they should be paired if possible against C and D.
Assuming A has not played C or D and that B has played C, then A must play C and B must play D.
To arbitrarily pair A with D thus forcing B to drop down to an opponent with less than 8 points is totally wrong and a violation of Rule 3.
Last edited by liam stephens on Mon Apr 25, 2011 2:40 pm, edited 1 time in total.
liam stephens wrote:Well said Jay. A little bit of logic and common sense go a long way.
Incidentally, in the Swiss System, the Maximum Compression rule is an inherent feature of the Pairing Methodology,
and not an optional add on. It follows inevitably from the basic pairing rules:
1. No repairing (before the final round or rounds).
2. After the first round players should be paired with opponents on equal points,if possible, subject to Rule 1.
3. In pairing, no dropping down in the points table should occur, except where forced under Rule 1.
So, taking a typical example:
Player A on 10 points
Player B on 9 points
Players C and D on 8 points
Assuming A and B have already played each other, then they should be paired if possible against C and D.
Assuming A has not played C or D and that B has played C, then A must play C and B must play D.
To arbitrarily pair A with C thus forcing B to drop down to an opponent with less than 8 points is totally wrong and a violation of Rule 3.
If this is the correct standard then how do people constantly get paired twice during tournament play??? It is definitely time for the checkers world to have some formal training on how to pair correctly. It also seems that in many tournaments the Swiss system is being used when round robin is possible. We dont have THAT many players at tournaments besides nationals to be doing Swiss all the time
As I understand it, in the USA repairing in the last round is generally the rule.
This rule also applies in Ireland. In Britain there is no repairing.
Allowing repairing in the final round has several advantages.
1.Better to end the tournament with a bang, rather than with a whimper. (How often have we seen tournaments decided by the lottery of who draws the weakest opponent in the last round)
2.With repairing, the top two players are forced to meet again and the player in second place has the opportunity to attempt to beat the leader.
3.The leapfrogging effect is suppressed.
If there are 10 players or less in a section then a Round Robin, certainly, seems the best option.
I agree that with 10 players or less a round robin is always the best choice.
Swiss system should be used when there is not enough time for a round robin event.
It is a lesser evil, but it is designed to detemine the first three places.
Let us look at the final Ilinois result:
1. Moiseyev 22 points,
2. Michael Holmes 19,
3. Larry Keen 16.
So it did the job.
Someone who was overtaken in the last round claimed it was unfair.
It was indeed unfair, but only to him ;-
In a rather subjective way, but it was not unfair in the way that it was arranged , ie intersubjective unfair.
It would have been unfair if he, after scoring only two points in the last round, could not be overtaken by a player with 1 point less before that round.
And that is what happend.
And it does not matter if that player played his brother, or if you did not see the games.
As long as the brothers play fair, it is okay.
And if you did not see the games as you played some skittle games, then where is the evidence?
You simply refused to see the games, but it seems they had been played, Nick.
And as someone that has two brothers that he loves, I find it irritating that you claim brothers would cheat a third person.
Nick, if you don't like to be overtaken, just play better.
Nothing more to say.
----------------------------------------------------------------------------------------------------------------------------------------------
As for leapfrogging may I question if this is anything of relevance to the discussion, ie if that "leapfrogging" really exists at all?
No matter if u have round robin or swiss system u can always get between 0 and 4 points in the last round, and it is always possible that anyone in striking distance can overtake if u score less.
So what?
That is only fair, as it is a sport.
It has to do with competition, fairness and sportmanship!
Do you really want to minimize chances to overtake the leader in the last round?
And why not in the earlier rounds?
As all rounds have to treated equal, or not?
Let us just declare the best rated player the winner and do not bother him with the task to prove his supremacy (Irony!)
In fact, it is only fair that one can overtake someone that scores less in a given round.
To demand to minimize that is simply not approriate.
For any round.
May it be the last or the 1st.
As pointed above, leapfrogging is also possible in round robin.
As long as you are in striking distance, and you score better, you can overtake from behind.
But what is wrong with that?
Vettel leapfrogged some drivers in the last F1 race.
Yet you can't complain.
His opponent had the chance to score better, but they did not manage to do so.
Richard Beckwith overtook some players that were ahead of him at the QT in 2009, yet he did it with fair means, and there was nothing wrong.
Michele Borghetti overtook Ron King in the last round in Dublin.
But he was only two points behind, and Ron King scored only 2 points, while he scored 4 points.
Everything was fair.
--------------------------------------------------------------------------------------------------------------------------------------------------
And to argue with probability is simply leading nowhere.
As it is possible to have a pairing that has a 0,1095 percent chance.
In fact, given the huge amount of possible pairings in this event, all other possible pairings are also very "unlikely".
What u said Nick, is simply that it is not posssible to win in a lottery for an individual, as the chances are against you each time.
Yet: someone wins each time, so what?
Did all lottery winners that has a probability to win less then 1 or 50 per cent cheat?
Your conclusion is wrong as it makes no sense!
Your probability to win was as low as that of the others, that is why it was fair.
And the probability of the pairings of your opponets were the same.
So what?
You proved it was possible, as you calculated the probability of that paring as above O (impossible).
And yet you complain it is impossible.
You are a witness against yourself.
----------------------------------------------------------------------------------------------------------------------
As for repairing:
1. All rounds should be treated equally.
To allow repairing only in one round or in some would simply violate that rule and manipulate the result accordingly.
It hurts the rights of of any third player.
2. What about the player in place 3?
Suppose two players are on 22 points, one on 21 , and the next on 16.
The player in 3rd place has not yet played one of the two leaders, who had palyed against each other in an earlier round.
With repairing he now gets the fourth ranked player.
If he scores a 4-0 and the other two play 2-2, he wins.
It is not unfair that he "leapfroggs" them, as he was in striking distance.
It is only unfair that he had the chance to have an inferior opponent in the last round, compared to a fair pairing.
Repairing is to blame, not any form of overtaking.
Fair means you can play each player once, and only once.
Like in a round robin.
Repairing is unfair.
As you dont say in round robin that in the last round the two with most points have to be "repaired".
Obviously that is unfair!
Nothing else.
Did you know that a round robin of 10 players and 9 rounds swiss system with 10 players should provide the same result in mathematical models?
In fact that was how Swiss system was invented:
Try to imagine a round roin with all players, and find a model that allows to have the same result at the top ranks, with less rounds then the round robin.
If you expand that number of rounds to the number of the round robin, it should provide the same result.
If you allow repairing that does not work anymore!
If you would allow repairing in a round robin, you would thus allow some players to play others twice, and others did not play some opponents at all, thus paving the way for manipulation.
If you do the same in swiss system, you excatly make the same mistake.
-------------------------------------------------------------------------------------------------------------------------------------------------
-------------------------------------------------------------------------------------------------------------------------------------------------
Conclusions:
1. The discussion about leapfrogging simply indicates it would be against any ethics to overtake a player from behind despite big odds.
But the opposite is in fact ethical, so this discussion should be stopped for once and for all.
2. Swiss System makes only sense without repairing, as does round robin, which is in fact only a special ("ideal") case of a swiss system.
To allow repairing is unfair and manipulative.
In round robin or swiss system.
It should not be done.
3. Low probalility does not mean it can not happen or that something was unfair.
As long as all had equal chances, it was fair and it can happen, it is even likely to happen, as 0,1095 percent means it is likely, not that it is unlikely.
The glass is 0,1095 full, but not empty at all!
Greetinx from sunny Europe,
Ingo Zachos
You can rent this space for advertising, if you like!
Someone does not win the lottery every time. Eventually, (thats why the jackpots grow larger and larger) someone hits it. This could be days, weeks, months years. After so much time and purchases eventually enough numbers are played that someone wins, but that individual's odds, as they may have only bought one ticket or only played once, are extremely low. The odds of any random someone winning the lottery of all players at somepoint in time is much higher than 0.1095%.
#2.
I have pointed out what I believe was incorrect pairings, but I do not care nor claim the pairings are the reason that I didn't win the title. I only went through the statistics to point out what I believed to be incorrectly and unlikely events, in effort to support my case. Assuming my understanding of swiss pairing is correct, some of the tournament pairings were not possible so 0.00000 probabilities. If true, this would strengthen my case by:
a) Tournament had incorrect pairings, so this begs the question if the statistically low occuring pairings happened by chance or were fixed
b) If the question is asked abou the integrity of the pairings, one might understand the gateway to giving 2 losses in the final round to secure a title. Bad actions beget bad actions
So even if the tournament was run incorrectly, that's not my main case, I don't care. I just think given the manner of how and where the Ellisons played their last round, and the quickness of it. It was all deceptively quick and purposely unnoticed. This leads me to believe that one brother simply gave up, and gave his brother 2 wins. Even if the first game was a loss from the weak side, the evidence stated even shows that he most likely just gave up the second game because he knew he couldnt win the title anymore, but his brother could. If Gene played his all out best, and tried for the maximum points he could earn, and received 0, then I would legitimately have lost the title. I dont believe this is the case.
The Ellisons always play at the front table, and could have began their match quite quickly. Yet, they played this match off in the corner when no one was watching, and made sure to finish before anyone realized they had started.
Maybe the Ellison's weren't trying to cheat me out of a title personally, but seeing as a 3 point score would make another 14 point score. The Ellisons would have to reconfigure all the standings and prize money given out to all of the other players (many) who scored 14 points in that tournament. So not only was ensuring an 11 and 15 point score (the only 11 and 15) attractive monetarily, and title wise, but it was also attractive with the sense of ease as I explained in my decision to take 2 draws. It simplifed the calculation of the results. So I believe 11-15 was planned, as it is the strongest move in the game, especially in this scenario.
Metaphorically, stop focusing on a random sentence in a novel, and study the underlying themes....
True is true ... we just proved and tested true with various percentage many times in this topic ... isn't ? Statistic says that people who say 50% true live longer
There is one more thing i would like to add here and clarify things.
It's about pairing in Round 5. I said in my earlier post that this pairing could be done better, however at the time when Gary/Gene did it, they just followed exactly the rules in IL Association. So - rules can be improved in my opinion (an I hope they do it next year), but Referee(s) didn 't do anything wrong - intentionally or even un-intentionally.
Ironically, when Dr. Beckwith developed his Guide, he used Torunament Rules of IL Association as model
Alex - What is your "state title recovery"rule and what is maximum compress pairing from the top? Since i have not come across these terms in the past 50 years i am skeptical of their validity.
The " A " Class at this past week-end's tournament requires ACF Membership to play in the Masters. Dr. Beckwith and Kim Willis and myself have received several emails and the three of us as ACF Executive Officers are engaged in discussion . All I can offer for now..............................................
