RESULTS: 2011 Amethyst 7

[coxcomb777]

Yuki won Yippeeeee. ~There can be miracle, when you believe~

[Zorya]

before I read through all your reply(it's a bit too long), i think that you did not understand what i said at all.

all seven matches of TierA vs TierB in one period was LINKED as a unit.

for example: try swapping seven Tier1 vs Tier 2 matches in Aquamarine with those in Topaz.
that will create a new schedule and it still works. because all the 14 matches are happened between those 14 characters, it will never affect other matches.

[maglor]

before I read through all your reply(it's a bit too long), i think that you did not understand what i said at all.

all seven matches of TierA vs TierB in one period was LINKED as a unit.

for example: try swapping seven Tier1 vs Tier 2 matches in Aquamarine with those in Topaz.
that will create a new schedule and it still works. because all the 14 matches are happened between those 14 characters, it will never affect other matches.

Please do read. They were linked on purpose of MAXIMIZING the variance of SDO. They needed to be linked to GUARANTEE uncertainty needed to deterr large groups from coordinating massive tactical voting. Swapping is possible way as randomizer, but then it becomes impossible to stop charge of schedule tempering with blind scheduling ( that is neede for uncertainty generation ) and the fact 99% of the voters won't read any posts in this forum.

Simplified questionaire + extras

1. Please define what you mean by 'better'.
2. Please tell how how these random swapping will solve the biggest problem, the misassignment of character's tier
3. Please suggest improved way of determining seeds for all 50 characters.
4. Please show us a example of your full schedule. Let's indeed see how much it can be better, in the definition of your choice.
5. You also made charge of necklace group match being of no consequence. Please elaborate.

[KamiyaCheap]

Fate 2 consecutive win! My happiest day ever! XD

[Zorya]

1. that's always a problem. but randomized schedule has a better chance to weaken the effect than current one.

2. it's impossible

3. current schedule makes LOOP all into one group, that's even unfair for those 7 characters, especially the wildcard. randomly created LOOP is not a major problem.

4. in my opinion, the better SDO variations in Tier1&2 should within certain range. not the same, not like current status. additional necklace arena + SDO variation of 2010 is good enough. to maximize the SDO variations make necklace arena seem stupid.

5. we've talked about these things last year.

6. i think it will work. just put a download link on the Rules page is enough. people who don't care to download it won't complain about it. if they do not trust staff to such a degree as you said, the voting process and multiple votes deciding method are more suspicious.

7. in fact, i did not blame on the seeding result. it's only enlarged the problem of current schedule. although we can really make it better. three rounds of 25 1v1 games with some calculations (such as ELO system) will give a better seeding result. since multiple characters voting varies from 1v1 a bit and is greatly affected by the composition of the constants. it's not a big problem that some characters are shifted a bit unless it goes too far.
and finally, my suggestion for wildcard character is the 10th to 15th seed since there's an additional necklace round.

[maglor]

4. in my opinion, the better SDO variations in Tier1&2 should within certain range. not the same, not like current status. additional necklace arena + SDO variation of 2010 is good enough. to maximize the SDO variations make necklace arena seems stupid.

I think we are now getting to the core of the debate, the SDO range. I am disturbed that you still haven't given me single concrete number to look at, however. Please do tell us what was the SDO variation in 2010, and how it compares to 2011 so far. Unequal sample sizes, unequal variance T-test should tell us some details as to how different they are. ( http://en.wikipedia.org/wiki/T_test ) If you are unwilling to do it, please ask someone who isn't a staff to do this for you. ( Team Rocket Elite and lihuazou are not involved with schedule creation nor vote checking , thus they might be willing if you ask them ) . Also the thing we should consider is

1) why is maximizing the SDO variation bad? some case studies will be nice as well.
2) what is the SDO variation range we should shoot for.
3) How can we attain the desired SDO variation from above 90% of the time? ( Here, I really don't think your randomized scheme can accomplish this. Would love to see you prove that your scheme can achieve this 90+% of times . Again, if you lack sufficient stat knowledge ask someone for help. I can do it, but my Statistician's honor say I should not since I am in position of defending the current scheme )
4) For the methods that works for (2) and (3) above, what might be other added benefits and what potential problem exist?

As for trusting the staff, I do hope most of you trust us, but it only takes a few disgruntled people to spread discontent. If we don't have means to prevent this in place, there will be a time when things get blown way out of proportion. I have experienced how distrust of just a few random people eroded voting level in Korea's Best Moe Tournament, first hand, and I think having insufficient safe guards coupled with questionable attempts to remedy the situation was big reason why this happened.

Finally, do present for us a schedule made by your method. Let us study it. Perhaps it might solve one of my dilemma with currently proposed schedule by pure random accident.

[Momokutenshi]

This isn't exactly related to the schedule, but I wonder if the Necklace Showdown would work better if it was just winner take all instead of being weighed by the SDO of the period. SDO would still be used a the tiebreaker to determine which characters with 6-1 period records make the showdown. It would also have the benefit of having a simpler and easier to understand scoring system in the necklace period. Also I think people would rage less over the results since they are more willing to accept a loss by pure vote count than a loss due to how score is calculated (see Kanade in Aquamarine).

If that's implemented now (as opposed to next year), I smell the possibility of revival of rage-following-Aquamarine-necklace..
That said, I think this has merits, although it would epically suck for girls who overcame a horribly difficult schedule to 7-0 but lose by votes to someone who had a real easy schedule...

@the things Zorya's talking about: I don't really understand it, all that stuff is too complicated for me and flies over my head, but I agree re: the linked chars in one period. Although it's good for predicting matches in Round 7 xD

I will simply say, that by having linked char pattern, it is now possible for a 7-0 contestant to have SDO as high as 7 * 6 * 3 = 126 as it now is possible for all 7 of your opponent to go 6-1. If you don't have the link, the maximum possible SDO will be lower. Of course this means that a 7-0 contestant can have SDO of 0 as well, but that would either mean total misplacement of ALL 7 opponent's tier , or extremely persistent large group hell-bent upon pumping up SDO for certain character. I seriously doubt if any group can achieve this, because they have to keep up their effort for almost 3 weeks. It is almost impossible to keep thousands of people on same trivial objective for more than a day. 2008 group worked because their number were less and even they had on and off days.

But it sure sucks for the girls in that single infinite loop of crappy SDO in that period, ne? Especially the wildcard who's stuck in every infinite loop of crappy SDO in every period. *coughhinagikucough*

Not that I'm biased or anything...

All that said, you're doing an amazing (and mostly thankless so it appears) job, it sure as hell can't be easy for you so thanks a bunch for setting it all up ^-^

[maglor]

This isn't exactly related to the schedule, but I wonder if the Necklace Showdown would work better if it was just winner take all instead of being weighed by the SDO of the period. SDO would still be used a the tiebreaker to determine which characters with 6-1 period records make the showdown. It would also have the benefit of having a simpler and easier to understand scoring system in the necklace period. Also I think people would rage less over the results since they are more willing to accept a loss by pure vote count than a loss due to how score is calculated (see Kanade in Aquamarine).

If that's implemented now (as opposed to next year), I smell the possibility of revival of rage-following-Aquamarine-necklace..
That said, I think this has merits, although it would epically suck for girls who overcame a horribly difficult schedule to 7-0 but lose by votes to someone who had a real easy schedule...

@the things Zorya's talking about: I don't really understand it, all that stuff is too complicated for me and flies over my head, but I agree re: the linked chars in one period. Although it's good for predicting matches in Round 7 xD

I will simply say, that by having linked char pattern, it is now possible for a 7-0 contestant to have SDO as high as 7 * 6 * 3 = 126 as it now is possible for all 7 of your opponent to go 6-1. If you don't have the link, the maximum possible SDO will be lower. Of course this means that a 7-0 contestant can have SDO of 0 as well, but that would either mean total misplacement of ALL 7 opponent's tier , or extremely persistent large group hell-bent upon pumping up SDO for certain character. I seriously doubt if any group can achieve this, because they have to keep up their effort for almost 3 weeks. It is almost impossible to keep thousands of people on same trivial objective for more than a day. 2008 group worked because their number were less and even they had on and off days.

But it sure sucks for the girls in that single infinite loop of crappy SDO in that period, ne? Especially the wildcard who's stuck in every infinite loop of crappy SDO in every period. *coughhinagikucough*

Not that I'm biased or anything...

All that said, you're doing an amazing (and mostly thankless so it appears) job, it sure as hell can't be easy for you so thanks a bunch for setting it all up ^-^

Well, Hinagiku hasn't shown Tier 1 strength, so I don't think her average SDO is suffering. Being caught in a loop means your SDO will be 63 , no less, no more, if you go 7 - 0 in the loop. ( Please refer to some of lihuazou's posts for this if you want to check why ) . This can be a blessing in disguise if the other 7 caught in the loop are weak on average compared to others in the same tier. Hinagiku is currently 13th in overall SDO, which is about where she should be given her wins and losses so far. In Amethyst Necklace period, she was 16th. In Topaz, she was tied for 16th as well in SDO. In Aquamarine she was 10th. So, if you calculate her average necklace standing, it is 42 / 3 = 14th, so very close to where she should be. I doubt Hinagiku's SDO is suffering because she is caught in the loop. The undefeated in the loop always gets 63 in SDO. That would place her at tied for 4th in Aquamarine, 3rd in Topaz, and 5th for Amethyst for average standing of 4th, right in the middle of necklace group match for now, but soon will be the top of the necklace group due to some characters above her not entering. In this Maximize SDO scheme, being caught in the loop can be blessing in disguise , perhaps mean being the leader of the necklace group from Sapphire period on.

[Miku]

I don't know until today that to maximize the SDO by means of specific tied-up-character is for purpose to avoid tactical voting. For randomized schedual tends to prevent it by unknown and unpredictable, while now tends to tell clearly that's impossible or it's too hard to cheat. >_>

I think people probabaly make more complainments about the things they judge as "crazy" obviously, like the currently happening crazy SDO for one or two characters every period. I'm not supposing a randomized schedual making things better. It's based on the purpose to average the SDO for the characters in their own tier. Maybe that looks fairer, which some people prefer to see. Despite the problem of "which is better to prevent the misassignment of character's tier", the current assignment leads to prossible happend case that one character gets into the necklace round with poor SDO. Considering they are usually the most powerful ones in the League, they take large amount of votes but nearly impossible to win the necklace due to their pitiful SDO. So it may bring the question "Whether it's meaningful to allow them to enter the necklace round." They contribute to discombobulate the result rather than for their own chance of getting necklace. I can't provide any practical solution to this problem, but it's worthy to think about.

[Kordosa]

Good: Mio, Rin, and Tomoyo winning
Bad: Nagi, Kuroko, Mafuyu, Lingyin, and Winry losing
Extended Comments:

• I'm really sad that Holo lost, but at least Fate won.
• Hinagiku was close. It would have been nice if she had won to get an SDO boost, but these results are fine.

Picture: OMG, neko Yuki clones!
Video: A very nice video. And there were several series I recognized in it (some I wouldn't have expected), and some I didn't know.

Spoiler for Series I recognized; the last six I haven't watched:

Cowboy Bebop (the movie), Eureka Seven, InuYasha, Tales of Symphonia, Noein, Koi Kaze, Fullmetal Alchemist: Brotherhood, Shuffle!, K-ON!, Hayate no Gotoku!, Air, Kanon, Clannad, Fate/stay night, Fooly Cooly, Minami-ke, Tengen Toppa Gurren Lagann, Code Geass, Neon Genesis Evangelion, Zero no Tsukaima, Toradora, Yosuga no Sora, Toaru Majutsu no Index, Toaru Kagaku no Railgun, Shakugan no Shana, Black Rock Shooter, Angel Beats!, Ah! My Goddess (either the second season or the Tatakau Tsubasa anniversary OVA), Read or Die (OVA), Rosario+Vampire, Kaichou wa Maid-sama!, Utawarerumono, Beck: Mongolian Chop Squad (I think), Yumekui Merry, Full Metal Panic!.

LULZ: Ayase 5-2 with an SDO of 3

I don't much care for Ayase (especially after she beat Holo), but how do you manage to beat five opponents and only have an SDO of three? They all must have been the very-bottom tier.

Holo, what happened? How did someone like you of third tier drop this this much? What happened to your supporters?

I'm right here, but there's only so much I can do by myself... T_T


As far as the whole scheduling discussion that's taken place (yes, I actually read through all of it even if I didn't understand half of it), this is why I've decided to focus more on the post season and the tiara and less on the necklaces. Period SDO is based solely upon the schedule which isn't even randomized. The schedule is made so that there are exciting matches each period (though I know this isn't the only consideration when the schedule is made). In my mind, a more random schedule would make things interesting because it might allow for contestants not bound for the post season to have a chance at a necklace. But I'm not on the staff so I have no room to complain.

At the very least, though, would it be too much to ask to post the list of contestants by tier after the seeding matches conclude? Because I don't follow all the statistics talk, and I only have a very general idea of which girl is which tier. And as far as maintaining trust with the community, wouldn't releasing a tiered-based list of contestants be a good compromise for not posting the entire schedule? And since strength will rise and fall during the tournament, simply add a disclaimer indicating that the list is how strong each girl is projected to be at the time of the seeding tournaments and doesn't reflect those changes in strength over time. Heck, put it in a spreadsheet so voters can download it and modify it as the tournament goes on to indicate those changes (all for personal use, of course).

As far as the necklace match is concerned, I admit I facepalmed when I found out that period SDO would still play a part in determining who would get the necklace. It kind of defeats the purpose of the match, doesn't it? I mean, if period SDO still affects the outcome, then why not just go back to the original set-up of the highest period SDO contestant in round 7 pick up the necklace (with the addition of the one necklace rule)? Otherwise, the "second chance" it affords the lower period SDO contestants isn't really a second chance at all. Instead, it's more like another chance for a contestant to again fail to win the necklace. So, while I don't necessarily think the necklace match is a joke like someone previously stated, I do think it still needs a little more tweaking. Again, I'm not on the staff, so staffers don't take offense.

[Eater-of-All]

I think, as people are getting into the numbers of ISML or perhaps getting too comfortable in seeing patterns, that they are forgetting these necklace periods are first and foremost meant for excitement.

If you want the ultimate balance, look no further than the main statistics portal, i.e. the overall regular season rankings, because therein exists no such thing as "unbalanced scheduling" or "unfair SDO distribution" when everybody faces each other. But is it exciting to watch that table move glacially for 5 months? No, because it's predictable, at least much moreso than period rankings. Why then, would you ever want to schedule it so that the period rankings will become as similar to the overall rankings as possible?

I'm no statistician, so I can't explain the nitty gritty bits of how the current system works as well as maglor can, but I can defend (and remind everyone of) the philosophy of it. SDO is not supposed to be the root of all evil and distortion in this world - it's meant to give the girls who fought hard and vigorously in a period their due credit. As Momokutenshi mentioned, it wouldn't be fair for "girls who overcame a horribly difficult schedule to 7-0 but lose by votes to someone who had a real easy schedule". In terms of pure statistics, yeah sure, maybe the girl who wins the necklace round is the stronger one, but if we're defining the necklace as a prize to be given to only the strongest, and thereafter the next strongest, ad infinitum, why would we need period rankings at all? Why make girls go through all these ordeals in a period when you can achieve the same effect simply by going down the overall rankings and awarding the top 7? Most people don't realize it, but a necklace is in reality a product of luck, effort, and strength. Luck because you gain the opportunity for high SDO from good scheduling, effort because you "worked hard" to grasp that opportunity and fought through those difficult opponents, and strength because you need it to complement that opportunity with votes in the necklace round. Kanade in Aquamarine had strength, but she didn't have luck, and therefore the effort she put out in doing Angel-Beatdowns on her weak opponents can hardly be compared to Mikoto's. You might argue that Kanade was never given the opportunity to begin with but... the fact remains that Mikoto grasped her opportunity and worked harder for the necklace.

Besides, what fun, rage, and exhilaration would there be if no luck is involved, and Kanade predictably wins the first necklace?

I rest my case.

EDIT: fffff partially ninja'd by Kordosa ;_;

At the very least, though, would it be too much to ask to post the list of contestants by tier after the seeding matches conclude? Because I don't follow all the statistics talk, and I only have a very general idea of which girl is which tier. And as far as maintaining trust with the community, wouldn't releasing a tiered-based list of contestants be a good compromise for not posting the entire schedule? And since strength will rise and fall during the tournament, simply add a disclaimer indicating that the list is how strong each girl is projected to be at the time of the seeding tournaments and doesn't reflect those changes in strength over time. Heck, put it in a spreadsheet so voters can download it and modify it as the tournament goes on to indicate those changes (all for personal use, of course).

Every Amethyst fantasy thread has a link to the list of contestants by tiers. For convenience, here it is (Hinagiku is the wildcard).

[lihuazou]

Too long, I can not fully understand

If the aim is to increase the SDO variance,now the schedule is almost the best.But the big gap between the SDO will lose that part of the necklace game of suspense? Such as topaz

[xcrossfacekillahx]

Spoiler:

I'm no statistician, so I can't explain the nitty gritty bits of how the current system works as well as maglor can, but I can defend (and remind everyone of) the philosophy of it. SDO is not supposed to be the root of all evil and distortion in this world - it's meant to give the girls who fought hard and vigorously in a period their due credit. As Momokutenshi mentioned, it wouldn't be fair for "girls who overcame a horribly difficult schedule to 7-0 but lose by votes to someone who had a real easy schedule". In terms of pure statistics, yeah sure, maybe the girl who wins the necklace round is the stronger one, but if we're defining the necklace as a prize to be given to only the strongest, and thereafter the next strongest, ad infinitum, why would we need period rankings at all? Why make girls go through all these ordeals in a period when you can achieve the same effect simply by going down the overall rankings and awarding the top 7? Most people don't realize it, but a necklace is in reality a product of luck, effort, and strength. Luck because you gain the opportunity for high SDO from good scheduling, effort because you "worked hard" to grasp that opportunity and fought through those difficult opponents, and strength because you need it to complement that opportunity with votes in the necklace round. Kanade in Aquamarine had strength, but she didn't have luck, and therefore the effort she put out in doing Angel-Beatdowns on her weak opponents can hardly be compared to Mikoto's. You might argue that Kanade was never given the opportunity to begin with but... the fact remains that Mikoto grasped her opportunity and worked harder for the necklace.

Besides, what fun, rage, and exhilaration would there be if no luck is involved, and Kanade predictably wins the first necklace?

I rest my case.

The top contestants at the end of the regular season last year were the ones to win a necklace. The one necklace rule controls the number of times a contestant can win a necklace.
The good thing about the necklace round is that the characters who didn't have enough luck but had the strength to compete can win the necklace by having x more votes to the current leader. How about the 7-0 girls who compete but have slim to none chance of winning a necklace? It just backfires them on that round which is bad. Since SDO carries over after 7 matches, it does not bode well for them if their SDO is very low and if they are not strong enough. What happens most of the time is it boils down to 2,3,4 people even if it's a 7 way match unless the SDO diff. is close enough for everyone. But as for who wins the necklace, it's always the top 7 none the less.

[loudmouth]

I kinda like the schedule this year. Sure SDO gap between first and second is pretty huge (my only complaint, but I don't know how you'd fix that and it's not a big enough problem to look into) but I like the schedule. Less hate for Shana

[Zorya]

1) why is maximizing the SDO variation bad? some case studies will be nice as well.

it's bad.
because in that way the deciding of necklace is MOSTLY decided by schedule. it's just ridiculous.
and i'm tired of this, i've said what I wanted to say.

[Kordosa]

At the very least, though, would it be too much to ask to post the list of contestants by tier after the seeding matches conclude? Because I don't follow all the statistics talk, and I only have a very general idea of which girl is which tier. And as far as maintaining trust with the community, wouldn't releasing a tiered-based list of contestants be a good compromise for not posting the entire schedule? And since strength will rise and fall during the tournament, simply add a disclaimer indicating that the list is how strong each girl is projected to be at the time of the seeding tournaments and doesn't reflect those changes in strength over time. Heck, put it in a spreadsheet so voters can download it and modify it as the tournament goes on to indicate those changes (all for personal use, of course).

Every Amethyst fantasy thread has a link to the list of contestants by tiers. For convenience, here it is (Hinagiku is the wildcard).

Doh.... That's what I get for not bothering with Fantasy. Thanks.
So is the left column updated at the start of each period to account for strength changes, or is it just the match opponent portion (which I don't necessarily need)? And I assume the match opponent portion is filled in as the period goes along and not at the start. Or else we'd essentially have a period schedule available.

[Eater-of-All]

Looks like my case can't rest in peace yet.

The top contestants at the end of the regular season last year were the ones to win a necklace. The one necklace rule controls the number of times a contestant can win a necklace.
[...] But as for who wins the necklace, it's always the top 7 none the less.

You touch upon a point that I want to clarify: I don't condemn the fact that strong girls win necklaces. Strength, after all, is one of the key components I mentioned in securing a necklace. Luck, as I argued, is also a factor, but not so much that someone such as, say, Konata, Saber, or even Charlotte can steal a necklace right off the bat from the likes of Kanade and Mikoto.

So as I understand it, your argument here is that luck is useless since "top contestants at the end of the regular season last year were the ones to win a necklace" anyways. Two points I'd like to make:

1. What happened last year happened because there was no one-necklace rule. Notice that Shana won three necklaces, and notice that by winning necklaces, none of the top hitters were barred from winning any more necklaces. It was, indeed, very difficult for any non-top-7 to win any necklace. It gave too much of an edge to the strength part of necklaces, which I suppose is the reason why the staff put in place the one necklace-rule this year. Sure, someone from the top 7 will likely win the first few necklaces this year, but can you honestly neglect the increasing likelihood of a non-top-7 winning the latter few necklaces now that those who've already won can longer compete? Personally, I think it's still a tad bit hasty to assume that this year will end up the same as last year.

2. SDO, and thus scheduling, makes things more unpredictable. If you determine necklaces purely through strength, even with the one-necklace rule, what would certainly happen? Aquamarine: Kanade, Topaz: Mikoto, Amethyst: Yuki, Sapphire: Shana, etc. The SDO factor obviously impacted the first two, and for the third one, it's not entirely certain whether Yuki or Eucliwood would win it. For the fourth, Shana better pray that she gets a good schedule to fulfill that prophecy (although people who want Shana to get Ruby will pray otherwise). As such, it keeps things exciting. Alot of people raged at the Aquamarine result, and that's part of the excitement too (imagine a totally predictable Kanade-Aquamarine win, ugh).

The good thing about the necklace round is that the characters who didn't have enough luck but had the strength to compete can win the necklace by having x more votes to the current leader. How about the 7-0 girls who compete but have slim to none chance of winning a necklace? It just backfires them on that round which is bad. Since SDO carries over after 7 matches, it does not bode well for them if their SDO is very low and if they are not strong enough. What happens most of the time is it boils down to 2,3,4 people even if it's a 7 way match unless the SDO diff. is close enough for everyone.

If you have a 7-0 girl who has a low SDO, that means she fought easy opponents. Which one is more honorable: a 7-0 girl who defeated girls with barely 0 to 2 wins, or a 7-0 girl who defeated high caliber opponents with 4 to 6 wins? Who had the toughest route and deserved the necklace the most? Under the old (2008) system, the first girl would win, purely by easy VD. Under the previous (2009-2010) system, the second girl would win, purely by virtue of SDO. Under the current system, both girls would be able to duke it out one last time, and if the SDO difference is small enough ("the first girl put out effort close to that of the second girl") or if the first girl is strong enough, the result can be overturned. Incidentally, this current Yuki-Eucliwood situation for Amethyst 8 is similar to this; you can see it as an upset in both ways. It can be a big upset if Eucliwood overcomes her 6-1 record and conquers the necklace from Yuki, or it can be a small upset if Yuki overcomes her SDO deficit and conquers the necklace from Eucliwood. What allows this? Necklace round.

PS: I don't get what you mean by backfiring; I can't detect any injuries or long-term health issues from entering a necklace round and not winning.

At the very least, though, would it be too much to ask to post the list of contestants by tier after the seeding matches conclude? Because I don't follow all the statistics talk, and I only have a very general idea of which girl is which tier. And as far as maintaining trust with the community, wouldn't releasing a tiered-based list of contestants be a good compromise for not posting the entire schedule? And since strength will rise and fall during the tournament, simply add a disclaimer indicating that the list is how strong each girl is projected to be at the time of the seeding tournaments and doesn't reflect those changes in strength over time. Heck, put it in a spreadsheet so voters can download it and modify it as the tournament goes on to indicate those changes (all for personal use, of course).

Every Amethyst fantasy thread has a link to the list of contestants by tiers. For convenience, here it is (Hinagiku is the wildcard).

Doh.... That's what I get for not bothering with Fantasy. Thanks.
So is the left column updated at the start of each period to account for strength changes, or is it just the match opponent portion (which I don't necessarily need)? And I assume the match opponent portion is filled in as the period goes along and not at the start. Or else we'd essentially have a period schedule available.

The left column, I assume, has been staying and will stay the same (or else someone must be really horrible for not updating Holo's tier after all these recent matches). The opponent portion is used for fantasy stuff, so you can ignore that if you only want the tiers. And yeah, that portion is only filled out as the period goes along.

[loudmouth]

At the very least, though, would it be too much to ask to post the list of contestants by tier after the seeding matches conclude? Because I don't follow all the statistics talk, and I only have a very general idea of which girl is which tier. And as far as maintaining trust with the community, wouldn't releasing a tiered-based list of contestants be a good compromise for not posting the entire schedule? And since strength will rise and fall during the tournament, simply add a disclaimer indicating that the list is how strong each girl is projected to be at the time of the seeding tournaments and doesn't reflect those changes in strength over time. Heck, put it in a spreadsheet so voters can download it and modify it as the tournament goes on to indicate those changes (all for personal use, of course).

Every Amethyst fantasy thread has a link to the list of contestants by tiers. For convenience, here it is (Hinagiku is the wildcard).

Doh.... That's what I get for not bothering with Fantasy. Thanks.
So is the left column updated at the start of each period to account for strength changes, or is it just the match opponent portion (which I don't necessarily need)? And I assume the match opponent portion is filled in as the period goes along and not at the start. Or else we'd essentially have a period schedule available.

The left column, I assume, has been staying and will stay the same (or else someone must be really horrible for not updating Holo's tier after all these recent matches). The opponent portion is used for fantasy stuff, so you can ignore that if you only want the tiers. And yeah, that portion is only filled out as the period goes along.

The left column is this year's seeding based on the preseason seeding rounds' results so it doesn't change. lihuazou has a working estimation of all the contestants' strength though
http://www.internationalsaimoe.com/forum/viewtopic.php?f=30&t=1898

[xcrossfacekillahx]

^ It's much better that way than the one before

[KawanCido]

an analyzing war(discussion)...you are all smart*** here~
if i may make a simple conclusion here, i'd like to say: "Major probability for 1st-Tier characters own the Necklace, include Heavenly Tiara."