I think it does change a lot, with my arguments being the following:
1. The number of votes between Saber and Asuna is around 700 VD. Even without this 230 votes that you claimed was irregular, Saber will still have a solid win over Asuna for 400+ VD.
2. Unlike the number 600 you first claimed, 230 VF is perfectly normal if you consider other hourly raise at well (in case you haven't checked it, it's around 100+), this is EXTREMELY likely to happen if this 230 raise belonged to the most convenient timezone to ISML voters in general.
3. When Chibasa mentioned social media, it is not a claim without clues. We observe several changes from campaigning in social media and take it into consideration.
I mean, i appreciate the suggestions in general. But imo, a suggestion based on some numbers that you didn't even try to doublecheck is not really useful in any circumtances.
Chibasa wrote:Social media are very various and can have different impacts. You may think it would have an impact on long term but that's not how most social media works (except if you pay).
Social media economic system is actually based on fact you need to pay so every people following you can see what you post (targeted ads working same way). If you don't pay, only a part of your community see your post (except if they go check your own profile specifically). And the part seeing your post is mostly people connected during 2-3 hours after you posted. (It also helps people to have approximately clear timeline, posts they can see are not totally random: partly based on their interactions with posts from different people and pages in the past). So a bump like this is totally possible.
The impact on long term is way smaller but you can see it. If you check between 12 and 24, Saber is having more votes than Asuna. (1028-947 for Saber)
While Asuna were little stronger during 4 first hours. (588-556 for Asuna)
It could be caused because people voting Asuna votes earlier in general too, it's sure arguable, but if you can't be sure by looking at this graph that the campaign had a small lasting effect, it's still more plausible than arguing it didn't had any.
I would like you know that I rechecked my messages and my informations several times, I'm just a bit too quick when I give explanations and that is why I simplify a little the reallity.
The bump stay strange If I analyse lot of previous result and I never see a bump that hit the most parts of characters after the 2-3 first hours. And here, it's the case, the bump of 5 to 8 hit several specific characters.
"The number of votes between Saber and Asuna is around 700 VD. Even without this 230 votes that you claimed was irregular, Saber will still have a solid win over Asuna for 400+ VD."
-> It's not really true because few months ago Asuna lose against Illya but after a recheck Asuna win hundreds of votes, results can be very variable in function of what manner you check them. Therefore result are never solid on Internet. And i don't claim that votes are irregular, but just that we can ask questions about the significant bump of 5 to 8 that have aproximatively the weight of the strange advance of Saber against Asuna, and perhaps the vote are irregular.
To go back about the Chibasa explanation about social media, I don't agree 2 or 3 hours are not sufficiant. I have a rather popular Twitter and I get the most part of my "Impressions" and "Engagement" in the 12 first hours and not only the third.
"When Chibasa mentioned social media, it is not a claim without clues."
-> I will say you that I already said to maglor, I think the functionnement of ISML is too opac and a bit more of transparency about where and how ISML finds there informations can reassure lot of people.
"It could be caused because people voting Asuna votes earlier in general too"
-> After the habitual bump of the start vote that is common of the most part of characters in each round, previously Asuna win a lot of votes few hours before the end of votes, we can saw its curve increasing. (But not today and it's also a bit strange). (cf: Ruby rounds 1,2,3)
"The impact on long term is way smaller but you can see it. If you check between 12 and 24, Saber is having more votes than Asuna. (1028-947 for Saber)"
-> I never say that Saber can't win against Asuna (because I have also vote for Saber to defeat Asuna becaus I think Asuna is an opportunistic character this year) but not with a so much higher result. But as you point out I think that is the true effect of the campaign but not for the bump.