Yes thank you. I already know it can be flawed. But just as I showed above, head to head is not necessarily better, and it fact I believe it's slightly worse. Will there be cases like yours, sure, but yours is a very tiny fraction of cases, and many more times SoS is the better of the two.

By the way, do we actually have any real evidence about the records of the other players from that event? Does anyone really know for sure who the other guy did or did not play?

I probably still have all that stuff filed away in my basement cabinet.

Probably.

Though I don't know how detailed my information is. I mostly had squad sheets. I will look.

I have to admit...when I first started playing in swiss tournaments I was annoyed. One of my first events with Shinja, we were in an FU sealed that was 3 rounds. At the end of the 3 rounds, he and I were both undefeated but he had the stronger SoS. I think my first opponent (who pulled Boba Merc) ended up 0/2 and dropped. I think I was also paired down round 3. At the time, I hadn't played enough to realize the full extent of it.

After regionals (my first 6 round event) my level of respect for SoS is a lot higher. I have learned to respect the system overall but regionals just enhanced it. In WA, we ended up with someone paired down in round 4 and they lost (i think), so by the time we got to round 5 Shinja and I were the only undefeated but we were already in the system for 6 rounds. I lost the game to shinja. Round 6 I played and won and Shinja lost to the player I beat in round 4.

We ended up with 4 players who were 5-1. I played all 3 of the other ones. my other opponents went 3-3, 4-2 and 4-2. I think I ended up with 73% wins on my opponents. It was a tough day. It felt wierd to place above Shinja, who beat me but it was a hard day. He is getting my pass anyway. Honestly Daniel was going to give him the pass aswell if he ended up winning.

Must have been tough for Sithfan who was 4th and his only loss was to me by 1 point (52-51). I just think at the end of 6 rounds and you get 5 wins and 1 loss by only a single point. Thats damn close. Great Player too.

SoS does alot to show how tough it was for that player to get wins. It is helpful to get less drops because thats can really hurt people. I guess my point overall is that in alot of 4 round events I feel like I have one tough game but some events you get 4-6 tough games.

This is excellent and it shows exactly why SoS solved problems that Head to head first would create. Let's look at it the other way, if Head to head were 2 and SoS was 3. You played and beat 2 of the other 5-1s, so you would obviously be ahead of them. Shinja played 2 of the 5-1s and went 1-1 against them. But with H2H as 2, Shinja would place first, and you second. That certainly is no more fair than what TimmerB faced in 2007.

Now, the other issue, that I expect to be brought up here, is the playing of that last round. I know many players think the event should end when one undefeated is made, which can happen (and would have in your event) a round early, so after 5 rounds instead of 6. I have consistently argued against doing so (although in most local events it rarely bothers me because usually we all want to get out of there, particularly in sealed at midnight ). In championship play, you cannot do that. That final round has a lot to say on the final rankings, even in cases where it doesn't end up affecting 1st place. Sometimes you can end up with say 7 4-1s and one 5-0 after 5 rounds, and what if you are placing to 4?

In your case, Shinja's record prior to that last round probably was significantly less strong than yours. Heck, he would have at that point only had one game against a 4-1, and you would have already had 2. Playing that last round is critical in determining the true champion and more accurate final rankings. So I applaud you guys for doing it right. Swiss works best when you use all the rounds given, that is how the algorithm is set up to work, not to cut it short.

I think people often confuse the required number of rounds to get one undefeated (which is how Swiss determines the number of rounds) with the central idea being that you are trying to get one undefeated and that is it's only purpose. That is not the case. Swiss is meant to do several things. It is meant to allow each player to play every round. It is meant to determine a champion. It is meant to give fairly accurate ranks from top to bottom, and there is very strong mathematical proof that with each round, the rankings get better and better. Standardized testing with computer programs (like the GRE) use the same principle of weighting that Swiss formats use.

And I am not just saying it. I was a victim of it in 2007 I remind you all. I was the last 6-0 after 6 rounds. We had to play a 7th, I lost and ended up ranked 2nd in the top 8 that night, rather than 1st. I then had to play a tougher opponent than if I was 1st. There were 5 people at 6-1 that year. I beat number 3 (his only loss) and beat 2 5-2 people (neither of which made the top 8, but one was 9th I think). Number 1 (Dr. Divot) lost to number 3 but beat me. Now in our case, was it totally fair that I got dropped down to 2nd? Probably not because it was likely due to our early opponents, as we both beat equally strong players at the top, although I think mine was stronger overall. And head to head wouldn't have solved it as we had 3 6-1s who all played each other, and were each 1-1 against each other. Not to mention the 4-5 guys who did not play us. So you get my point here I think.

I have a solution to this issue.

It solves the "drop" problem, eliminates SoS concerns, and fixes the system so that everyone's record gets treated fairly.

It's called Double Elimination.

Everyone is paired randomly in round 1. In each subsequent round, people with like records are paired. If you get 2 losses, you are out of the tournament. Being removed this way has zero impact on the people you have already played against.

People in the "double-losers" bracket may continue to play if they wish, but they are no longer part of the official tournament and have no impact on the outcome for the people still playing.

Wow, who would thought the solution so simple?

Seriously though there is as much chance of my idea being implemented as there is of SWM people influencing a change to the DCI Swiss program that WotC uses for all their games. (And for the record, I think my idea has less chance than a snowball surviving in the Sahara desert at noon.)

And now imagine we had a realistic ranking system, that would also allow pairing to be done via seeding, to make it even more legitimate of a championship Oh what dreams may come

The realistic problem of the DCI system is that it does in fact break down in a sense for small tournies (ie less than 16). The sytstem seems to have been designed for large tournies and to create a "top x" playoff after the swiss rounds were over.

I am not sure if ANY system would rank players accurately in a field of 8-10 unless you just did an entire round robin and compared records and even then I am not sure if it would work out.

I find that SOS works fine and has done so over the years (I used to be big into MTG) especially when in a larger field (ie 6-7 or more rounds) and there is a "top x" complete with playoff.

@ Timmer (and not to start an argument) : Over the years I have seen no where that am opponent dropping hurts your record. In most cases when I play/ed a scrub in round 1 I wanted them to drop as I knew that their 0-2 or whatever was much better for my SOS than the likely 1-6 or 1-7 they would have earned if they had played out all rounds. What COULD have happened back in 07 was that the people you beat in the early rounds went on to spectacular 0-7 or 1-6 records thereby almost negating the fact that your losses were to the 6-1 crowd (ie 6-1 +6-1 + 5-2 + 5-2 + 4-3 (guess) + 0-7 (guess) + 1-6 (guess) is 27-22) while the guy who finised ahead of you had his early opponents actually DROP (ie 5-2 + 5-2 + 5-2 + 4-3 + 2-2 drop + 0-3drop) for 21-14 SOS which is slightly better than the 33-22 that you may have earned instead of grinding through loss after loss.

In general if you play good players with good squads you want them to keep playing so they can help your SOS and if you play against bad players/squads you want them to drop ASAP so they don't drag your SOS down with them. This is the one thing that I find that differentiates MTG from SWM. At largert tournies in MTG players usually drop as soon as they are eliminated from top 8 conmtention or if they came with buddies as soon as all parties are eliminated and then do something else (draft, food, go home, hang out, etc.) thereby creating a pool of players at the top who have all played (usually) against players who have dropped and not usually having someone play a deck to an 0-6/7 record. However in SWM this seems to happen much more frequently that people play all the rounds because they are just out to have fun. This leads to some crappy SOS scores for those players that played the guy running his all wookiee theme team (or whatever uncompetitive squad you can think up) at a large tournament.

Again not trying to start an arghument it just seems that some people are confused about how the system works with regards to drops and not drops and how it affects SOS. So unless the DCI reporter does things 100% different for MTG and SWM (or it has changed very recently) I think that Swiss tournies with SOS tiebreakers are about the best we can hope.