Opinion Average Info

[Otohiko]

I think what you need to consider is that everyone has their own approach to Oping vids. It seems to me that there may be 3 approaches to op'ing videos: 'Innocent until proven guilty', 'crappy until proven otherwise' and 'average until proven either way'.

I'm the of the 1st category. I'll give a 10 in a category unless I find a flaw or a detraction; then based on the number and obviousness of detractions, I'll reduce the score accordingly. Not surprisingly, my average given is relatively high (although personally, I blame it more on the fact that most videos I op are wonderful little bits of work).

I haven't really met many people of the 2nd category. I'd imagine their scores would be low.

The 3rd is relatively common, although there's no agreement as to what is average - some go with 5, some go with 7, some go with 8...

And I don't think the org should go out and tell them, 'no, this is NOT your average'. Let them decide what their average is.

As for me, I'll still go with my approach, thank you. High scores? Maybe. But you can't accuse me of not being detailed/justified in my ops, either

(which reminds me that I still have over a dozen of ops I've been owing for a while )

[Zarxrax]

It basically all comes down to this: 5 is the average score possible. Whether or not people actually USE that as their own personally average is up to each individual.

Look at it this way:
Lets pretend for a moment that this is a site that collects video game ratings from various magazines and websites.
There could be one magazine that posts very fair reviews of every game that comes out. Now you go and look at the ratings from "Nintendo Power". Even the shittiest game probably wont get less than a 6. Their "average" is obviously NOT a 5. But they can still give scores as they see fit. Now then, what if a bunch of other magazines rated on the same sort of scale as Nintendo Power. Should the fair magazine be told it needs to readjust its rating system?

[storyteller]

quadir

Actually your behavior exemplifies one of my concerns.

Lets assume that in a normal distribution chart, aka the Bell Curve, there is a standard deviation of 0.75 points with a mean of 8.0 (8.0 being a simplification of the actual global average data. Standard deviation of 0.75 being a guess.)

We could predict that:
2.3% of scores would be less than 6.5 points
13.6% of scores would be between 6.5 and 7.25 points
34.1 % of scores would be between 7.25 and 8 points
34.1% of scores would be between 8 and 8.75
13.6% of scores would be between 8.75 and 9.5
2.3% of scores would be greater than 9.5

We could summarize these predictions as:
68.2% of scores would be between 7.25 and 8.75
AND:
95.4% of scores would be between 6.5 and 9.5

Now compare that with your figures. (Keeping in mind that a total of only 41 opinions may be too small of a population size to be normally distributed.)

I am going to simplify and say that from you, (and others with the same practices) the average score given is 5. So when a person gets an opinion from you there is a much higher probability that their score will be around five.

From your population we simplify here to a mean of 5.0

With the same standard deviation of 0.75

We could predict that:
2.3% of scores would be less than 3.5 points
13.6% of scores would be between 3.5 and 4.25 points
34.1 % of scores would be between 4.25 and 5 points
34.1% of scores would be between 5 and 5.75
13.6% of scores would be between 5.75 and 6.5
2.3% of scores would be greater than 6.5

We could summarize these predictions as:
68.2% of scores would be between 4.25 and 5.75
AND:
95.4% of scores would be between 3.5 and 6.5

With a broader standard deviation of 1.5

We could predict that:
2.3% of scores would be less than 3.0 points
13.6% of scores would be between 3.0 and 4.5 points
34.1 % of scores would be between 4.5 and 5 points
34.1% of scores would be between 5 and 6.5
13.6% of scores would be between 6.5 and 8.0
2.3% of scores would be greater than 8.0

We could summarize these predictions as:
68.2% of scores would be between 4.5 and 6.5
AND:
95.4% of scores would be between 3.0 and 8.0

Aiming for five, when hardly anybody else does, creates an artificial low when compared to the entire population.

[quadir]

Aiming for five, when hardly anybody else does, creates an artificial low when compared to the entire population.

And this is why when you look at movie reviews, they do not go `3 star average', they list particular reviews.

Sites like rottentomatos will give you a select high and select low ones, and site the sources.

Some people are known for just always rating things poorly. Others are known to be fair. Your problem is that to create a proper average, I need to BE average already. This is poor, since it doesnt take into account the credibility of the source.

But that is beyond the scope of the system, so I think you will just have to live with the current system as it is, people giving others ratings they feel are fair. Even if you includes an option to pick your `grading theory', people would pick the harshest, and rate the highest, just to boost the video up more. There is no real way to win, and it is certainly not by making people who rate average=5 to change their ways.

[Zarxrax]

Hmm... How about instead of changing the displayed average, we just limit the numbers that people can choose from to 5 and up!

That way, the lowest score anyone can possibly give is a 5, and you don't run into such problems as people using 5 as their average score. It makes perfect sense!

[storyteller]

It basically all comes down to this: 5 is the average score possible. Whether or not people actually USE that as their own personally average is up to each individual.

How is five the average? We have a set with a even number of values [1,2,3,4,5,6,7,8,9,10] The median could not be 5. It would be 5.5 taking the two middle values and taking their mean would result in 5.5
As 5.5 is not a selectable value for an individual, the number curently gets rounded down to 5.0

Is five the mean? If so, then what is being displayed in the "global average" chart?

Is five the mode? Is it the most frequently occurring value?

Look at it this way:
Lets pretend for a moment that this is a site that collects video game ratings from various magazines and websites.
There could be one magazine that posts very fair reviews of every game that comes out. Now you go and look at the ratings from "Nintendo Power". Even the shittiest game probably wont get less than a 6. Their "average" is obviously NOT a 5. But they can still give scores as they see fit. Now then, what if a bunch of other magazines rated on the same sort of scale as Nintendo Power. Should the fair magazine be told it needs to readjust its rating system?

What is fair? Equality is fair. If everyone got treated exactly the same as everyone else by everyone else. Inequality is unfair.

Are scores inflated? Yes. Will scores continue to be inflated? I would predict yes.

Can a few people change this trend? I would predict that no, they cannot.

Luckily, an opinion is more than just a number. Hopefully it is also valuable written feedback to the creator.

[storyteller]

quadir

Once you switch the topic to professional movie reviews then you enter a whole new territory. Professional movie reviews involve money and businesses. Professionals get paid to write their reviews and movies make millions upon millions of dollars. Anime music videos are just pure fan entertainment. There is no profit involved here.

I don’t understand the following, could you please clarify what you mean:

Your problem is that to create a proper average, I need to BE average already. This is poor, since it doesnt take into account the credibility of the source.

I am not trying to change people. What I had originally suggesting was changing a line of text. If people decide to change their behavior in response to a change in a line of text, then that is their business.

[Garylisk]

I think there's another problem here, and that is that people tend to feel really offended at a score of 5. They think that it is such a low score, when really, it's AVERAGE! So, in order to keep from making people mad, lots of people don't even go as low as 5 unless the video actually sucked. It's like, 5 is 1 or 2. or might as well be, to these people.

See what I'm saying?

I always base things aroudn 5 as an average, but I tend to rate higher.

[Otohiko]

Well, as I said in my previous post, I think marking is still very much dependent on individual practices.

As far as getting marks, I'm usually much more concerned with how the reviewer justifies them rather than the numbers themselves. Frankly, it's often the case that a 7 from a reviewer who is obviously critical and meticulous ends up being much more of an achievement in my eyes than a 10 from one who just enjoyed the video and left me a positive grade - not to say I appreciate either one more than the other, but the motive and justification behind the marks really plays the major role in what I look at as a good op.

[DrngdKreationz]

Storyteller, First off you said it yourself this is fan based entertainment... In my opinion that means not to take things TOO seriously (like you are right now about this average thing...) In order to demonstrate I'll show you how I grade my videos. each level takes 2 numbers on the scale (for normal, or feel strongly about it)

1-2 = absolute shit. no matter how I look at it I couldn't possibly say anything positive about this video the only "nice" thing i could do for you is not leave an opinion at all based on the "if you don't have anything nice to say don't say anything" moral value.

3-4= Below Average so you tried to learn how to edit, but you suck. you're rythym management is like a retarded kid trying to perform stomp. if you had skill this idea you have could take a better shape/ or you showed potential to do something good if you tried a little harder, i might come across harsh in my commentary but i'm just being honest about how this could have improved.

5-6= you are an average video. so you made a few mistakes editing, extra frame hold here and there. could have worked on your video quality, in short good job, but its nothing special. you might get a cookie...

7-8 = above average video, now these videos have shown a good amount of work in them well timed/synch great video quality. good concept I actually felt what you were trying to do. great song selection etc... kudos to you

9-10= fucking superb!!! HOLYSHIT WAS THAT AMAZING, WILD/CRAZY, (toss in whatever really positive and strong adjectives here) it blew my mind. i had to loop it because i just wanted to watch it again... hats off to you,

thats how the "average of 5" works for me. and looking at this breakdown i find it rather fair don't you?