Weak characters for the most part can only beat other weaker characters. This is true both in Saimoe and ISML. Huge upsets can happen, but they aren't that common.

Hinagiku. She was always strong but it's the sudden surge of popularity near the end of the main season that put her in the running for winning the contest.

Sadly, it's still far fewer posts than we got last year and the year before. I don't think Saki characters will fall off the map. They are too strong for that. However, I don't expect them to be as strong as they were this year.

You don't need huge upsets the seperate ISML and Bestmoe from Saimoe. Only moderate ones, and they happen a whole lot. I can predict 95% of all ISMl matches on average(well, during the regular season as relegation matches happening now were the only tricky predictions in the regular season), and was close to around 89% with Bestmoe. Saimoe, I'm lucky if I can reach 70%. The only thing tricky with Bestmoe is the fangirl factor, and it's what I hate most about it, anyways. And atleast Saimoe keep is moderatly close, usually not above 66% unless It's some one very weak vs. a top tier...70%+ rapes is commonplace in ISML.

Not that I'm dissing the other two contests, but point is, I like Saimoe's restrictions. It's complex, and takes time and effort to actually vote, compared to other contests where It's just vote and forget about it the next 24hrs. Or random people surfing the web come across ISML...and vote within the span of thirty seconds, only to never return again. And I think this is the main reason why Saimoe preferance is so much different than everyone else.

Japan Saimoe's way of voting entices more people to stick around. It actually utilizes effort and correct timing.

95%? I knew you were better at predictions that I was but I didn't think it was that much. No wonder I can't beat you. I was only 90.18% for the regular season. I don't follow Best Moe so I can't comment on why it's easier to predict matches there. However, a large part of the reason ISML has fewer upsets is because we have a lot more data. Even if a character makes the finals in Saimoe, we've only seem them in 6 matches plus a preliminary match. In ISML by the end of Aquamarine we've seen every single character in 9 matches. Also 60% of Saimoe matches are Round 1 matches. In those matches we have nothing but the unreliable preliminary matches to go on. For me, it was difficult to predict a lot of the matches since I didn't know a lot of the weaker characters. I think you watch a fair bit more anime than I do so maybe this wasn't as much of an issue for you. However, regardless of the reason why you are correct that Saimoe is a lot harder to predict. I just think it's because we don't have as clear a picture about the character's strength as opposed to Saimoe being particularly friendly to upsets.

The closer matches thing as an illusion caused by 3/4-way matches. If you only look at the votes recived by the top 2 in each match, there are still lots of huge blowouts. Not always by top tiers either. In H3, Yomi got more than twice as many votes as anybody else in the group. Only looking at the top two, Yomi broke 70%. C07, Nogizaka Mika nearly tripled anyone else in her group. Mika isn't close to being top tier. By the time we're down to 1 on 1 matches, almost all of the fodder weak enough to get 70-30'd have been eliminated. Considering Tacos vs Chiaki was a 70-30 match, imagine Tacos vs Futami Eriko.

But she already was DE material, I meant kinda like what happened with Negima last year. Sudden burst for a period and they all got out with the exception of Evangeline.
And in Hina case it wasn't a sudden small surge of popularity. It was cause by a one semester long series, in the middle of the contenst that had lot's of Hina.

just bumping one more time. Here are compilations of Saimoe material for this year:

Saimoe 2009 YGO Cards -> link
match posters and other stuff i created -> link
overviews and charts -> link
Anon's Saimoe posters, brackets and others (reupload without source material; i'll post the original RS link later) :) -> link

finally, my wrap-up for this year's Saimoe. again, although my favorite characters did not win, it has been a fun ride, everyone. thanks to everyone who participated. see you in 2010!

Well if anyone wants here is a download for the three files I made related to Saimoe, in openoffice format. They're the Manga main tournament bracket; translated list of all girls in anime saimoe prelims and their results in the same, sorted by series; and the same as the previous but with only the characters in the main tournament.