(taken from Seattle Riichi Mahjong Club's blog)
(Final results can be found here)
The 2017 Seattle Riichi Open was played from July 15-16, and could be considered a success for our first major foray into tournaments. Everything went basically on schedule with little downtime (for me anyways) which means that spacing was good.
We had managed to reach 40 registered participants at time of check-in, but late withdrawals brought that number down to 35, which meant that one of our planned subs had to fill in. It also meant that instead of the possible 4 seats on offer, we were down to 3.
The schedule was aggressive in my opinion with the time limits set at G/75 (that's chess term for game in 75 minutes). But by the end of Day 1 most tables finished before or at the time limit so in terms of finishing people were able to get it done. By the end of Day 1, and with only one round left to go in qualifying, the Top 4 were over 25 points clear over 5th place.
Something must have happened overnight though because all 4 wound up on the wrong side of the ledger in the 6th round, effectively crunching the top 7 into a range of just a smidge over 12 points. Factor in the 1/3 division of all scores and it meant that if you finished 1st you were guaranteed to be in the finals. 2nd place would give you a good chance, but not guaranteed.
Another situation that occurred was that 4 of the top 8 already held a WRC seat. Which meant that we actually had 4 people for 3 seats! There was all to play for going into the playoffs.
In the semis, the last hand resulted a tie for 1st in one semifinal meaning that the two players (Zach Francks, Kevin Shi) actually split the uma. At a +10 for each it meant that Charlie McDonell, who had finished 1st outright in his semifinal held a 10.2 point lead going into the finals. With Daniel Moreno rounding out the field it meant that Kevin Shi, the only player out of the quartet who did not have a seat yet, would automatically earn one.
That also meant that for the other 3 - Kira Nebilak, Anthony Hsieh and Sakina Toyota, they were fighting for 2 spots. Kira had almost a 16 point advantage over the other two, while Anthony had just a 0.1 advantage over Sakina! That meant that Kira just needed to avoid trouble and Anthony and Sakina would try to make sure they were ahead of the other.
As the table progressed though, I noticed that with half of the time elapsed the table was still on East. And when I looked closer the reason was apparent. Anthony was dealer on East 3, and was on yonhonba (4th dealer repeat). He also had the majority of points which meant that barring an Atlanta Falcon-style collapse, he all but secured the 2nd WRC seat.
That left the final seat between Sakina and Kira. Kira had a lead of 15.8 over Sakina, so to hold onto the position Sakina could not finish 2nd and Kira 4th, and if Sakina finished 1 place ahead of Kira, she could not be ahead in points more than 5,800. And while I didn't hear the point counts exactly it sounded like Sakina was in 2nd and Kira in 4th which meant there was all to play for.
But when I got the table report, Sakina was in 3rd place and Kira 4th. The point difference? 4,100.
Which meant that by 1.7 points Kira secured the 3rd and final WRC seat!
With the seats determined, there was the matter of figuring out the actual winner of the tournament. Kevin with nothing to lose perhaps went more aggressively than before and paid for it at the table finishing with (5,100) points. That meant that the other 3 players - 2016 PML Open winner Charlie McDonell, 2014 WRC Top 32 finisher Zach Francks, and PML's Daniel Moreno - who has finished in 2nd or 3rd in all tournaments he has played in, were all in contention.
Heading into all last Charlie was in the lead with Zach in 2nd and Daniel just 100 points behind Zach. Going for the title, Zach declared riichi looking to overtake. But when he didn't win the hand he fell from 2nd to 3rd.
Which meant that we now have a multiple tournament winner as Charlie McDonell wins the 2017 Seattle Riichi Open! Congratulations to you Charlie!
Now for a couple of comments as the tournament director side of things...
This tournament featured a pairing format trying to put sense to the pairings that is normally random. It doesn't make sense to me that in one of the final qualifying rounds that say in a cut of the top 8 of a 36 person field like this one, you could have a random pairing of players in 4th-7th-8th-28th. Coming from a tournament director background in chess, I originally thought about implementing a swiss-system format. But with cutoffs for the top players, this made such a system completely unfeasible because you wound up with a situation like the above where someone above the cut is guaranteed to finish 3rd or possibly 4th and drop out right when the cut happens.
That's where I came up with the ordinal pairing system, which just split the groups into quartiles based upon scores and paired top down from each group for the opening rounds, with the final qualifying rounds being paired in groups of a size of the cut*2. Duplicates were going to happen, but that was going to be an inevitability unless we had a large enough field that switching two people in adjacent tables wouldn't still create a duplicate pairing. The system was breaking down about the point I expected it to, and the switch in Round 5 to the cut*2 group size worked well.
But what I learned from this tournament is that having the tables filled from the top to the bottom of each quartile isn't as important as perhaps filling it with one from each quartile. In other words, the importance of having the 1st table filled with 1st-10th-19th-28th in a 36 person field is minimal as just having Top Quartile-2nd Quartile-3rd Quartile-4th Quartile. So I might change the format so that it just has to be the latter, and duplicates are swapped out unless it absolutely cannot be avoided.
Condensing of points at playoffs
In the first tournament I participated in, which was the 2015 NYC Riichi Open, I was introduced to having the points halved when players passed through the qualification rounds. And from each of the tournaments I played in, it seemed like that division was insufficient to collapse the field enough for there to be much drama to who might finish in 1st - especially for those at the bottom of the cut who almost had zero chance to come back. That's where I had thought of the idea of dividing the points by a factor equal to:
- Number of Qualifying Rounds/Number of Playoff Rounds
In this case it would be 6/2 or 3. But when I implemented it in this case, it collapsed the field so much that the difference between 1st and 8th was less than the uma difference between adjacent placings. It almost made it equivalent to Montreal's tournament which was a clearing of points and a straight top 2 advance which was not the intention of this method.
It is my opinion that we need to balance the importance of games in qualifying with the ability for those in those at the bottom part of the playoffs to have a chance (not significant, but greater than say being hit by lightning) to come back to win.
The other thing to think of is that we in the North American region are as a whole are still in its infancy in regards to the riichi mahjong scene when compared to other main organizations in the WRC such as the JPML (Japanese Professional Mahjong League) and the EMA (European Mahjong Association). As such we have players with a wide range of ability which could lead to inflated scores such as what we saw in the 2015 NYC Riichi Open.
But what we've seen in recent tournaments is that perhaps this gap is narrowing as we hold more tournaments and the floor for our players rises. It's possible that there may be no need to have such a division of points after qualifying - especially if I carry my pairings of the final 2 rounds of qualifying grouping people in sizes of cut*2 as those players (in this tournament's case the top 16) will be beating each other up and score ranges may narrow. Perhaps next year we have a schedule of the following
- Rounds 1-4 - Quartile Pairings
- Rounds 5-6 - Cx2 Pairings (where field is grouped into sections of cut x 2, and paired internally
- Rounds 7 - Semifinals (Players grouped into sections = cut size, and paired as follows:
- Round 8 - Players grouped into 1st-4th in each cut section and paired
The location was great for the field size, but if we have a larger tournament, we will need more TV's as players will move outside of the playing area we were at and may not have line of sight to the TV and the timer.
We had extra space and people, and realized that with that public location we could have a place and personnel to teach people if they were interested in learning. Great recruitment tool.
And for some of the personnel I had, the division of labor could have been better with someone on the admin side helping me enter things into the computer and confirm player scores instead of having the runners just verify scores. They can also teach if necessary.
Overall, I think the tournament went well. There were lots of things to take away and work on to make next year better, which should be the case anyways. Hopefully next year will be even better.