Yuku free message boards

MasterThief wrote:

---

scepticA, I'm not saying a 2 is bad, I just think she was more focused in on, more screentime, involved than say, Grant, but the numbers say otherwise. She is still a very strong CPP regardless of whether it's a 2 or 3. Besides I thought visibility was a relative kind of thing. We don't have to follow the vis chart to the T every time, but for example, it would be wrong to give someone a 5 who was more like a 3. Would it really be that wrong to give Andrea a 3? (Serious question, I've only done one season of edgic before this)

This is very true... but as I argued a few episodes ago, I think that we probably should unless there's a REALLY good reason. I'll sum it up so no one has to dig through the thread (and maybe improve the explanation):

Visibility is a relative thing (and I'm speaking relative as in one person's visibility being relative to that of the other contestants). When the numbers are calculated, they are based on not only the footage for a character, but also the footage all other characters received. If we make an adjustment to one character's visibility, the relative nature of the ratings is diminished.

A great demonstration of this actually came up this episode with the chart. The first version had an extra confessional for Krista. In that version, I believe Phillip was in the 5 range. Change Krista's visibility factor and the basis of comparison has changed. With that, Phillip is now a 4-5. (In all honesty I can't remember if this was Phillip I was looking at. I do remember noticing a change in other people's ratings after the chart was fixed for Krista). For that reason, I think that manually adjusting the ratings should make us consider what that means in a relative-to-other-players sense.

My other argument regarding this is that manually adjusting visibility puts players on an unequal playing field. Here are the components to visibility from my perspective:
Sentences*
Confessionals*
SPV
Time spent on screen
Size of the player's image on the screen during footage (full screen shot vs. hanging out in the background)

Edgic quantifies Sentences and Confessionals to produce a visibility chart. In this regard, every single player is considered equally. No extra attention is paid to one character's sentences or confessionals. It's consistent for everyone. The calculation leaves things out, but these things are pretty much all highly correlated. If you have a high number of sentences and confessionals, chances are good you also had high amounts of the things not being measured.

The other 3 are not calculated directly but they are often cited when we have an issue like this. I see two problems in the way this generally proceeds:

- This consideration is not applied equally across all players. In this recent situation, we are presumably thinking that Andrea's sentences and confessionals do not adequately capture her visibility (perhaps the time spent on screen is what's driving this). However, is anyone worried about these same factors for Julie? Stephanie? Rob? By only examining a select number of characters, it biases our ratings. If Andrea were not a winner pick, would we be having this conversation? I think it's a problem when the hypothesized winners start getting special treatment during data collection - do the ratings fit the right pattern because they truly fit or is it because we have subtly/unconsciously MADE them fit?

- The relative nature of visibility is not accounted for in these cases. If we were to measure time spent on screen for all characters, assuming we had time to do it, would the final numbers result in a net gain for Andrea relative to other characters? So she had a lot of time on screen. Well so did Rob - does he get a boost too? Natalie wasn't on screen very much so being lower-than-average on this new factor should lower her visibility, right?

Before I'm comfortable deviating from the chart, I would want to answer these questions for myself. Am I only doing this because I think the character might win/lose? Is the new information I'm bringing in actually going to change the relative position of the players' visibilities? While I wouldn't expect someone to actually do the hardcore visibility measuring in order to make this argument, I'd feel much more convinced if they did something along these lines. If someone can point out that, relative to other players at the same visibility level, this character got significantly more SPV or more screentime or what have you, I would be more likely to buy it. I'd want to explore the issue and see if adding that factor changes other people's ratings too. On the other hand, if people just have a "feeling" that a certain character deserves a boost, I'm not biting.

In the end people will do whatever they feel comfortable with. People can rate visibility however they want. I'm just saying that I think the chart works great and mucking with it is potentially hazardous without a good reason.