Deago said: NeoYami said: Congrats on this amazing thread, let me share my results: Manga results In the case of Skyhigh Hevan, its +7 because its not ranked, idk why Anime results The difference is usually useless, also the manga you mentioned has relatively too low voters to make a judgement, the mean is around 7, idk why is that. For you question, not To clarify, is the difference in score, the measurement used to figure out the Bias?(thats what I got from the original post) No, this has nothing to do with the difference, although high difference is usually related to 10 scores and 1 scores, which are usually fall into the clustered bias in the histogram. Most anime don't have positive bias, although can be noticable with "sequel effect" Most anime have negligible negative bias, the negative bias increases when popularity increases, in certian situations, negative bias also increases when the anime rank is very high; I guess because it is in "your face" which can bothers some people.[/quote] Oh ok, makes more sense now, I asked because the way I initially perceived it wouldnt be an accurate way to identify bias. Once again congrats on this work |
 |
TheFirmSword said: Although I'm not familiar with histogram, it's kinda pointless analyzing scores, because it varies from site to site. Detective Conan has a higher rating than Fate/Zero in IMDb, but it's the otherway around in MAL. So this may not work for every site My concern is this site IMDB votes are even worse in term of clustered bias, I had a hard time to find a histogram free from it. MAL is a good choice since it is the most popular anime website (I think), the more statstical power you have, the more accuracy. |
 |
| I dont wanna start next war but looking at numbers how LOGH can have higher rate than final AOT season? Just curious. |
| I don't like to make judgment of anime ratings from a website that is not dedicated for anime, I consider MAL accurate in terms of "bias" presentation, but let us compare it to other dedicated anime websites Detective Conan Histograms Comparison Apparent positive bias, very low negative bias Both slightly positively and negatively biased High positive bias, small negative bias No bias It seems AnimeNewsNetwork average score is the most "genuine" among the four, however, we must put into account the low statstical power into account before making judgements. |
|
Deago said: Bots are real problem when it comes to biased score in statistics, whether negative or positive. In my opinion, the solution is to develop and programme a more advanced algorithm for the weighted score. For example, lessening the effect of scores made by bots by utilising the mean score for every user, let's say people with high mean score (who rate only 10-9) will have less effect on the score than people who have average mean score. And the other way around, people with low mean score cause less effect with negative scores. I am not expert with algorithms, but I am sure you can have advanced ones that mitigate the issue of bots. This is why I sometimes find it useful to look at aniDB ratings, because their ratings do take into account each users mean score (they also provide the raw average as a separate score). They don't share their algorithm, but it does seem to at least do a decent job of mitigating fake scores as they claim. According to aniDB, "Votes are weighted in a manner that makes vote stuffing obsolete". I wish MAL would implement something similar and offer two separate scores like aniDB. Though MAL does claim to not count scores given from illegitimate accounts, how that works exactly I'm not sure. |
 |
ReegrezSNK said: I dont wanna start next war but looking at numbers how LOGH can have higher rate than final AOT season? Just curious. Almost every main charcater dying before the end of LOTGH and AOT despite being supposedly more grim and realistic still having characters with obvious plot armour? I am joking. More likely AOT fans being bad at setting up and using bots. |
| If you aren't at least somewhat biased, you likely have a robotic mentality and are a boring person. |
 ☆☆☆ "There's a huge difference between one and infinity. However, compared to the difference between existence and non-existence, one and infinite are nearly the same. I am the child destined to become the best witch... no... The greatest Creator in the world...!" -Maria Ushiromiya ☆☆☆ |
| Woah a high quality thread. Time to watch this thing called "Shitcom" it has caught my interest. |
Deago said: So with popularity, you made a chart binned by the median of every 20 pages (why 20?), and visually determined an elbow at page 19 (was this done with a formula as well? e.g. AIC, BIC). My problem is that seems supremely arbitrary. Notice how popularity rockets starting from page 39. We can assume any anime from page 1-19 is "popular" , or minimum requirement popular anime in MAL is about 140k members. EDIT* I think I can refine that chart by doubling the data and get a much better approximation of where the rocketing is, I may do that later, but I hope you get the idea. EDIT** chart is refined Is this the mean rating? Yes Nevertheless, you have a popularity cutoff based on the number of people who rated and a quality based on the mean rating. I'm assuming you did the same kind of subjective procedure to find the quality cutoff. Then, you applied a sort of Gaussian mixture model, k=2, where you assumed there to be one "genuine" label and one "fake" label, which you've associated with a term "fanboyism", which you have not defined. Up until the point the GMM was applied, we have some data that must be qualified. That is, it might have some use, although I'm unsure what exactly. After that point, you subjectively assumed that the entire "real" population falls into one distribution, and the entire "fake" population falls into one other. That seems very questionable. It would be more reasonable if the entire population, period, falls into one distribution with noise, or that a multitude of distributions is possible with the real population. There is no reason to assume that the distribution is necessarily bi-modal, with one real and one fake. Furthermore, if we center one such distribution and call the other distribution biased with respect to former distribution, it's unclear what that "bias" exactly entails. That does not mean that the voters of the show is "biased" or that any individual is "biased", nor that the anime score is "biased" -- and certainly not in the sense of being biased from impartiality, which is entirely a different concept altogether from the statistical definition. Hence I'm suspecting you're playing loose with words here, a semantic fallacy, by attaching this "bias" with "fanboyism". P.S. You are actually talking about skewness, not bias.  |
katsucatsDec 8, 2020 2:24 PM
| My subjective reviews: katsureview.wordpress.com THE CHAT CLUB. |
| haha you won't find me anywhere there because I'm unbiased asf with absolutely no 10's and no 1's!!! mainly because I haven't found my true love of a 10/10 anime and I don't think anything is really bad enough to be a 1/10- although I haven't added boku no pico to my list because I don't want the fbi on me |
  NYANPASU  whiskey tango foxtrot |
REJECT TRADITION, EMBRACE TIER LIST! |
No-a-nimeDec 8, 2020 5:35 PM
| heh. |
| I think everybody are all biased in some sort. I don't think I can rate series well so I'll just keep using 10s to help the series I think deserve boost 3:) :p |
 |
PrasantGrG said: ReegrezSNK said: that's due to botting issues currently and MAL algorithm trying to figure it out. This is an issue which will fix itself over time when votes increase as long as too many more bots aren't created.I dont wanna start next war but looking at numbers how LOGH can have higher rate than final AOT season? Just curious. Deago said: For "the missing positive bias" that might be the case for now, things change, many who are not AOT fans wait untill the season finishes, some people wait for the whole story to end, I know this phenomenon, I am 80% sure many genuine voters will vote later with 9-2 votes, and lower it's rank, and maybe you will see more 7 or 8 votes than 9, if that happened, then yes "the hidden positive bias" about 10/10 voters you mentioned will be real, but for now, I cannot detect it. Genuine? Even after genuine votes come in the early numbers suggest except 10s and 1s suggest it'll do similar to atleast the shows currently in the top 10. Wouldn't it be normal to have a very high 10 score percentage even after normalizing The finale will be the key for the score due to imo heavy bias towards the end of an anime in terms of overall perception and votes can often drastically go up or down depending on it especially something that airs nowadays weekly after its conclusion Look, anytime the 9 bar is lower than 10 and 8, this indicates external force rising the bar 10, it is untypical rare pattern you would only see in titles of the the categories I've mentioned: popular outside MAL with so much fanboys giving 10s, sequel effect where a clustered fanboys creating a secondary peak (10) away from the mean (8-7). Now, I am not attacking specific titles I dislike here, because obviously I cannot check every title's histogram in the database, I've searched random titles here and there that share the abovementioned trend, It turned out all of them, by coincidence, I, subjectively -duh- consider overrated, over-raved. There might be titles I like that falls into the same trend, who knows? everyone has the possiblity to have their favourite anime eat shit here. Logically, same thing should also apply to titles with the total opposite trend, where the 2s bar is lower than 1 and 3, which indicates "haters" and so on. Dunky said: Deago said: Bots are real problem when it comes to biased score in statistics, whether negative or positive. In my opinion, the solution is to develop and programme a more advanced algorithm for the weighted score. For example, lessening the effect of scores made by bots by utilising the mean score for every user, let's say people with high mean score (who rate only 10-9) will have less effect on the score than people who have average mean score. And the other way around, people with low mean score cause less effect with negative scores. I am not expert with algorithms, but I am sure you can have advanced ones that mitigate the issue of bots. This is why I sometimes find it useful to look at aniDB ratings, because their ratings do take into account each users mean score (they also provide the raw average as a separate score). They don't share their algorithm, but it does seem to at least do a decent job of mitigating fake scores as they claim. According to aniDB, "Votes are weighted in a manner that makes vote stuffing obsolete". I wish MAL would implement something similar and offer two separate scores like aniDB. Though MAL does claim to not count scores given from illegitimate accounts, how that works exactly I'm not sure. I thinks it's really good to take the user's mean-score into account, I've already mentioned that but I had never known they apply this in AniDB. This mitigate the effect of skewness due to everyone here uses a subjective scaling, the they typically consider the score of "mediocre anime" -in their opinion- is 7. Using such algorithm makes their score "less positively influencing". EndlessMaria said: If you aren't at least somewhat biased, you likely have a robotic mentality and are a boring person. I totally agree with you, being robotic is boring. I hope this gives us some some insight of our own bias that may creates such phenomenons. UsuallyExpected said: Woah a high quality thread. Time to watch this thing called "Shitcom" it has caught my interest. Have fun, hopefully if you loved it, think twice before scoring it a 10 instead of lower than that. katsucats said: Deago said: So with popularity, you made a chart binned by the median of every 20 pages (why 20?), and visually determined an elbow at page 19 (was this done with a formula as well? e.g. AIC, BIC). My problem is that seems supremely arbitrary. Notice how popularity rockets starting from page 39. We can assume any anime from page 1-19 is "popular" , or minimum requirement popular anime in MAL is about 140k members. EDIT* I think I can refine that chart by doubling the data and get a much better approximation of where the rocketing is, I may do that later, but I hope you get the idea. EDIT** chart is refined Yes Nevertheless, you have a popularity cutoff based on the number of people who rated and a quality based on the mean rating. I'm assuming you did the same kind of subjective procedure to find the quality cutoff. Then, you applied a sort of Gaussian mixture model, k=2, where you assumed there to be one "genuine" label and one "fake" label, which you've associated with a term "fanboyism", which you have not defined. Up until the point the GMM was applied, we have some data that must be qualified. That is, it might have some use, although I'm unsure what exactly. After that point, you subjectively assumed that the entire "real" population falls into one distribution, and the entire "fake" population falls into one other. That seems very questionable. It would be more reasonable if the entire population, period, falls into one distribution with noise, or that a multitude of distributions is possible with the real population. There is no reason to assume that the distribution is necessarily bi-modal, with one real and one fake. Furthermore, if we center one such distribution and call the other distribution biased with respect to former distribution, it's unclear what that "bias" exactly entails. That does not mean that the voters of the show is "biased" or that any individual is "biased", nor that the anime score is "biased" -- and certainly not in the sense of being biased from impartiality, which is entirely a different concept altogether from the statistical definition. Hence I'm suspecting you're playing loose with words here, a semantic fallacy, by attaching this "bias" with "fanboyism". P.S. You are actually talking about skewness, not bias. No, I am not talking about skewness, look, there's a clear gap between 10 and 9 even though the mean is 7.4.  Calling this a mere skewness is understatement. BadJoash said: haha you won't find me anywhere there because I'm unbiased asf with absolutely no 10's and no 1's!!! mainly because I haven't found my true love of a 10/10 anime and I don't think anything is really bad enough to be a 1/10- although I haven't added boku no pico to my list because I don't want the fbi on me Rating things 10 or 1 doesn't necessarily contribute to the clustered bias in histogram trends, but if that's the case with you, you're very good at avoiding falling with the groups who forcefully change the mean To help me and all other people who liked this topic (read comments above) to observe different types of histogram patterns that indicate certain categories I've mentioned, e.g sequel effect. It also gives awareness to which title the bots are targeting, and those who have difficult time differentiating between 2 & 1 score, and 10 & 9 scores, and how they may fall foul with downvoters and boosters, this not to lecture anybody nor for starting a movement, it is just observations for those who are interested, which you seem not, but many others are. If you don't care why bother? 2. All humans are biased no matter what. Yes they do, some are smart enough to be aware of their own bias, some are not. 3. Quality is subjective. Nobody said otherwise. Holy moly! MartinKetchum said: I think everybody are all biased in some sort. I don't think I can rate series well so I'll just keep using 10s to help the series I think deserve boost 3:) :p That's my point, some score to boost the rank, or lower the rank. At least you're aware of it and open about it. |
DeagoDec 8, 2020 11:58 PM
|
Deago said: Calling it an "external force" is just unfounded (external to what?), and moreover calling it non-genuine is unfounded. You count a major gap by cherry picking one anime to illustrate a point out of tens of thousands. What does that really mean in practice?No, I am not talking about skewness, look, there's a clear gap between 10 and 9 even though the mean is 7.4. Calling this a mere skewness is understatement. The answer is "just noise". But you don't seem to want to address those concerns. |
| My subjective reviews: katsureview.wordpress.com THE CHAT CLUB. |
katsucats said: Deago said: Calling it an "external force" is just unfounded (external to what?), and moreover calling it non-genuine is unfounded. You count a major gap by cherry picking one anime to illustrate a point out of tens of thousands. What does that really mean in practice?No, I am not talking about skewness, look, there's a clear gap between 10 and 9 even though the mean is 7.4. Calling this a mere skewness is understatement. The answer is "just noise". But you don't seem to want to address those concerns. Here, I've calculated the mean if 10 scores were equal to 9s, and 1 scores where equal to 2s. Current mean score (10 x number of voters + 9 x number of voters + 8 x numbe.......) / total number of voters 738440 + 145719 + 232352 + 194488 + 77208 + 28435 + 4131 + 2174 + 4700) / users 1,427,647 / 175,235 The mean score : 8.15. Normalised (161910 + 145719 + 232352 + 194488 + 77208 + 28435 + 4131 + 2174 + 1087) / 113,964 847,504 / 113,964 = 7.4 The mean score= 7.4. |
|
| Interestingly, I checked the histogram for "Ghost Stories", and this is what I've found: I guess this is one of the rare occasions where voters hadn't actually seen the same excact show. I assume the inflation in 10/10 scores for Ghost Stories is due to the English Dub, which turned the anime which many people think is "boring" into an adult dark humour show (and many loved that), creating an external force (the people who loved the dub), leading to a secondary peak away from the mean. For more detail about this, this phenomenon was addressed by @Oreomonogatari72 in his topic here: https://myanimelist.net/forum/?topicid=1692644 |
|
Deago said: What do you purport that tells us? lolkatsucats said: Calling it an "external force" is just unfounded (external to what?), and moreover calling it non-genuine is unfounded. You count a major gap by cherry picking one anime to illustrate a point out of tens of thousands. What does that really mean in practice? The answer is "just noise". But you don't seem to want to address those concerns. Here, I've calculated the mean if 10 scores were equal to 9s, and 1 scores where equal to 2s. Current mean score (10 x number of voters + 9 x number of voters + 8 x numbe.......) / total number of voters 738440 + 145719 + 232352 + 194488 + 77208 + 28435 + 4131 + 2174 + 4700) / users 1,427,647 / 175,235 The mean score : 8.15. Normalised (161910 + 145719 + 232352 + 194488 + 77208 + 28435 + 4131 + 2174 + 1087) / 113,964 847,504 / 113,964 = 7.4 The mean score= 7.4. The biggest case of confirmation bias |
| My subjective reviews: katsureview.wordpress.com THE CHAT CLUB. |
katsucats said: Deago said: What do you purport that tells us? lolHere, I've calculated the mean if 10 scores were equal to 9s, and 1 scores where equal to 2s. The biggest case of confirmation bias It shows us the effect of boosters unfairly increasing the mean score, leading to a misleading rank that does not represent the the most common opinion or rating of the show, which is is around 8-7, if the show wasn't truly overrated, then you would actually see more people also scoring it 9. |
|
| i read the title as based not biased and was waiting for explanation of which anime scores are based and which are not |
| I have more 1's than 2's does that make me biased? |
Roevhaal said: I have more 1's than 2's does that make me biased? No, this depends on what you've excatly rated 1, your personal histogram is only skewed; e.g, I don't see 2 scores lower than both 3 and 1, so you are good. EDIT* I see you've given FMAB a 1, yes, you've contributed to the negative bias within this title. Could you elaborate or justify this low score? Or do you consider yourself negatively biased against FMA brotherhood? |
DeagoDec 9, 2020 11:16 AM
|
Deago said: It sucks, it gave me no reason to care, it has ideas but it fails to present them in a way that makes you interested in the world and characters, everything just feels extremely bland. That and the constant gags gave me cancer.Roevhaal said: I have more 1's than 2's does that make me biased? No, this depends on what you've excatly rated 1, your personal histogram is only skewed; e.g, I don't see 2 scores lower than both 3 and 1, so you are good. EDIT* I see you've given FMAB a 1, yes, you've contributed to the negative bias within this title. Could you elaborate or justify this low score? Or do you consider yourself negatively biased against FMA brotherhood? |
| this was a great and informative thread |
 |
| Epic thread, but everyone is biased af lol |
| People who uses MAL forums are mostly bug people, take their speech with a grain of salt. Professional Supreme Gentleman - 恋人はあなたが選んだ兄弟です https://www.reddit.com/r/AnimeElitism/ (AnimeElitism Reddit) https://discord.gg/4kdGsFG (Dead Discord) 2静か先生88#7307 |
Deago said: These "boosters", as you accuse without evidence, are a part of the mean score, and the most common opinion is not the "right" opinion. It shows us the effect of boosters unfairly increasing the mean score, leading to a misleading rank that does not represent the the most common opinion or rating of the show, which is is around 8-7, if the show wasn't truly overrated, then you would actually see more people also scoring it 9. It's unclear what you mean by "if the show wasn't truly overrated" -- overrated according to whom? What is the reference being compared here? You took an arbitrary number, a peak between 7-8, and just assumed that's the score everyone thinks is right. Why isn't it the peak at 10 that's right, or the peak at 1? Why aren't they just all right with differing groups of opinions? What is your credentials in statistics, if you don't mind me asking? You're not using sound statistics to form a conclusion, you made the opinion a priori and cherry picked evidence to back it up. You say "No bias indicates acceptance of rank", but the rank takes into account every vote. You say high "quality" (i.e. mean score) indicate "negative bias" -- how? How does a higher mean score correlated to a lower peak? How are you assuming that the only "bias" occurs in the extremities? Anyways, it's clear you're not actually interested in these questions, so by all means keep impressing lemmings with your fake stats. There are so many ways to interpret these histograms. You chose one and presented it like it's objective fact, and it's not. Your jargon is all over the place. In one post, you used "healthy" to describe "variance". You also used "histogram skewness" to describe "bias" (the real usage of the word). Interestingly, you use "bias" to describe "skewness". Anyways, I'm done. lol |
| My subjective reviews: katsureview.wordpress.com THE CHAT CLUB. |
| Nice thread. But i think we should take this stuff for fun. Please dont be so serious. I liked it so much. |
RioAlDec 9, 2020 10:19 PM
 |
| tell me stat wizards, what does my rating scores tell you about my rating habits. |
I'm level  on mal-badges. View my badges.  |
Roevhaal said: I have more 1's than 2's does that make me biased? Man, there are around 600 anime you watched that you rated 4 or below. That's more than 70%. No offense but why do you watch anime when you don't like it? |
 |
Hillary_Clinton_ said: Can you analyze Boku No Pico's rating spread for us? Yes, these anime with this pattern are called anime with diverging opinions. They have both positive and negative biased 10's ans 1s, cultural or emotional reasons may create controversial opinions, look at Twilight for example. There are also "troll" 10 votes, while 1 votes are probably "negative emotional overreaction". Notice how it is different from Pupa's or Shitcom's histograms, the reason for this is "Boku no Pico" is a living meme, just like Twilight, very popular even outside MAL, leading to an external force (anime casuals with accounts with few famous anime e.g DBZ, AOT, SAO, Pokemon, and Boku no Pico "because memes") increasing the frequency for 10's and 1s. My interpretation is Boku no Pico doesn't deserve either 10 or 1, it is just an average yaoi hentai so 5/10 (which is the closest to the mean when normalised by equalising 10's to 9, and 1 and 2 to 3.) Shitcom and pupa are both shit anime that anyone rating them a 10 is probably biased or just a troll, while 1 scores are mostly unbiased. Pupa however has lower 3 scores compared to 4 and 2 The mean is closer to 3-4, not 1-2, hence, compared to shitcom, 1-2 votes in Pupa are more biased than 1-2 votes for Shitcom. doihavepotential said: tell me stat wizards, what does my rating scores tell you about my rating habits. You have a skewed histogram with a mean score higher than average, this mean you are more prone to "postive bias". You have scored an "overrated" special episode for Keroro Gunsou a 10 Meaning, you fell into one of the positive bias categories; "Sequel Effect", or in this case "Prequel Effect" lol.  |
|
| This is a really good thread, good job on making it! |
 |
Slawadia said: iirc fmab only has that many 1/10's because users wanting ishuzoku reviewers to reach the top in MAL ratings proceeded to mass low-score the competition earlier this year. Interspecies Reviewers situation was hilarious. It's a shame it ended so quickly |
 |
| Very interesting read thanks for the post |
Deago said: Same goes for "critically good" anime, some people call this "haters", "haters" only target anime that is popular around let's say the forum, LOGH for example is associated with "elitism" which has a bad connotation, hence it received more negative biased scors than, let's say Monster. I think it's more like fanboys who want to see their FMAB or AOT dominate the charts rather than "ew dumb space anime from the 80s no one cares about." |
 This is not your grave...But you are welcome in it :3  |
Cyanice said: Deago said: Same goes for "critically good" anime, some people call this "haters", "haters" only target anime that is popular around let's say the forum, LOGH for example is associated with "elitism" which has a bad connotation, hence it received more negative biased scors than, let's say Monster. I think it's more like fanboys who want to see their FMAB or AOT dominate the charts rather than "ew dumb space anime from the 80s no one cares about." Not true, the mean is close to 10, the histogram follows a logical pattern as in Imagine 60 members scoring: 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10 9, 9, 9, 9, 9, 9, 9, 9, 9, 9 8, 8, 8, 8, 8, 8, 8, 8 7, 7, 7, 7, 7, 7, 7 6, 6, 6, 6, 6, 6 5, 5, 5, 5, 5 4, 4, 4, 4 3, 3, 3 2, 2 1 This means there more non-fanboys who consider FMAB at least good anime, when you compare to something "overranked" like this: Positive Bias 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10 9, 9, 9, 9, 8, 8, 8, 8, 8, 8 7, 7, 7, 7, 7, 7, 7 6, 6, 6, 6, 6, 6 5, 5, 5, 5, 5 4, 4, 4, 4 3, 3, 3 2, 2 1 FMAB / AOT actually have negative bias, something like this: 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10 9, 9, 9, 9, 9, 9, 9, 9, 9, 9 8, 8, 8, 8, 8, 8, 8, 8 7, 7, 7, 7, 7, 7, 7 6, 6, 6, 6, 6, 6 5, 5, 5, 5, 5 4, 4, 4, 4 3, 3, 3 2, 2 1, 1, 1, 1, 1, 1 If they were really considered "overrated" by the majority of the users -except the fanboys who always give 10- then you would see less people scoring them 9. |
DeagoDec 10, 2020 4:23 AM
|
I'm pretty sure FMA has positive bias only |
No, the mean is close to 9, you cannot call scores between 10-4 biased because they represent the majority of MAL users, inclusing those "critics" and "non-fanboys". Bias happens when a clustered group exist whose opinion is far from the general opinion, whether crtical or non-critical. |
|
Deago said: no, it follows a natural decrease starting from one untill all fanboys, bots and trolls get involved.No, the mean is close to 9, you cannot call scores between 10-4 biased because they represent the majority of MAL users, inclusing those "critics" and "non-fanboys". |
Roevhaal said: Deago said: no, it follows a natural decrease starting from one untill all fanboys, bots and trolls get involved.No, the mean is close to 9, you cannot call scores between 10-4 biased because they represent the majority of MAL users, inclusing those "critics" and "non-fanboys". Fanboys, trolls, bots, don't score with 4, 5 ,6, 7, 8, not even a 9. Why? because doing so actually decreases the anime's ranking. |
|
| ^^LMAO that's the kind of trolling I like (^: |
Prophetess of the Golden Era  |
Deago said: They do because they are smarter than the ones from Ishuzoku Reviewers incident.Roevhaal said: no, it follows a natural decrease starting from one untill all fanboys, bots and trolls get involved. Fanboys, trolls, bots, don't score with 4, 5 ,6, 7, 8, not even a 9. Why? because doing so actually decreases the anime's ranking. |
Roevhaal said: Deago said: They do because they are smarter than the ones from Ishuzoku Reviewers incident.Fanboys, trolls, bots, don't score with 4, 5 ,6, 7, 8, not even a 9. Why? because doing so actually decreases the anime's ranking. Then they are doing the rank boosting wrong. And even if that's the case, you are basically saying thousands of strangers were organised enough to shape the histogram that almost nobody cares about? let alone trying to fool few statisticians? And if they were "smart" and organised as you said, they would try to fill the gap in 3 score bar before getting caught by by you right? |
DeagoDec 10, 2020 6:15 AM
|
Deago said: No they aren't, they've managed to get first place and go unnoticed, how can you call it wrong with those result? The organised efford is indeed impressive, I'll give you that. It might've been a bit to much to fill up even the 3's or maybe they are currently working on it, but I'm sure they figured it could be explained by claiming the haters are more prone to rate 1's and 2's than 3's.Then they are doing the rank boosting wrong. And even if that's the case, you are basically saying thousands of strangers were organised enough to shape the histogram that almost nobody cares about? let alone trying to fool few statisticians? And if they were "smart" and organised as you said, they would try to fill the gap in 3 score bar before getting caught by by you right? |
RoevhaalDec 10, 2020 6:22 AM
More topics from this board
» Real People and Live Action scenes in animeTheBlockernator - Yesterday |
15 |
by WatchTillTandava
»»
10 seconds ago |
|
» Anime musicnishant0 - 9 hours ago |
16 |
by GANCH
»»
2 minutes ago |
|
» Waifu War V5 (Anniversary-Edition!) (Round 1) ( 1 2 3 4 5 ... Last Page )Minkalex - Sep 28 |
367 |
by SoumyaUkil73
»»
6 minutes ago |
|
» ☀️ Anime Summer 2025 Male & Female Characters Tournament ( 1 2 3 )ISeeLifePeople - Oct 5 |
107 |
by GrumbleDango
»»
11 minutes ago |
|
» 🍷 AD Summer 2025 Best Girl Contest ( 1 2 3 4 5 ... Last Page )Shizuna - Sep 28 |
373 |
by GrumbleDango
»»
17 minutes ago |