Here is a fine example of my disagreement over data and documentation, I'll not bother to identify the user or the title it is irrelevant and unknown in the case of the user.

We recently had a Contribution that Contributed a bunch of data like some we know and love. His initial Contribution contained the following Notes

Studios and crew info came from movie credits. Looking at the list, I believe that the director should be George(first name) Roy (middle name), Hill (last name).

No DOCUMENTATION at all,  kind of like some other users, I voted no explaining that what he believed was irrelevant provide documentation and his Contribution was justifiably declined.

He resubmitted the Contribution with the following whine for notes and alll the same data

I am re-submitting this entry.  All info came from movie credits, opening and closing.  It is frustrating when I spend the time to go through the opening and closing credits, and sometimes watch part of the movie to obtain the info for DVD Profiler and it is rejected.

This time it was accepted, and I have to take MY time to go back and fix the erroneous data for George Roy//Hill, simply because this user did not provide documentation and was accepted based upon "I believe it so, so it is correct." Guess what, sports fans, the data was completely erroneous and i provided lots of documentation to back it up. My point is that it should NEVER have been accepted to begin with, regardless that he submitted a bunch of other data. This applies to all of us if you want to put data into the database because you say it is so, you can always expect a No vote from me, and when the day comes that the erroneous data is uncoverec I will call the previous Contribution notes to task for it. Incorrect data is incorrect data no matter how much good data there might be, undocumented data is bad data and I don't care how much other correct data there may be, I will always vote NO!!!!!!!!!!!!!!!!!! This is a team EFFORT and if you think forcing me or some other user to go back and unscramble a mess you(generic) have created is not being part of the team.

I have said before that I will NEVER ask any user to vote on my data simply on my so, that would arrogant and selfish on my part, and is arrogant and selfish on the part of any user who chooses to do so, any comments from Ken or gerri on this issue aside, I do not appreciate any user forcing me take MY time to correct bad data simply because he was so arrogant that he refused to provide documentation  and hid behind ken & Gerri to get bad data into the database.

Skip

Although I agree about documentation, I also agree with DJ's post. I've also looked at various sites and can't see anything that suggests G R//H.

Skip, right to rant granted 

The basic issue is, we require cast and crew to be taken from the actual movie credits. There is little we can do beyond believing the contributors statement that that's what he/she did, barring verifying each contribution by painstakingly comparing it by viewing one's own copy. For 10 to 50 (peak) contributions per day... I mean, not the easily disputed parts as BY's which require documentation, but every name in the cast and crew lists. This is a full evening's work, day in day out for everyone trying to review contributions. Me thinks, unreasonable.

The other issue is name parsing. Since the program standardises on USA conventions on first and middle names, any name consisting of three words is usually parsed as a/b/c. Only after a while the truly knowledgables adjust to the correct parsing (which in this case, seeing Karsten's scan, looks like Roy should be part of the last name).

The disputed contribution was basically flawed because of the "I believe" statement. Point is, for three word names no documentation is AFAIK required for the parsing (yet). Only errors with well known actors or crew are caught early. If he/she had omitted the "I believe" statement, nothing would have happened.

So, do we really want all three word names fully documented regarding to parsing? My guess is, we'd be overshooting then.

Edit: mdnitoil posted when I was still typing. I agree with his statements. I think we're both saying perfectionism is nice but unworkable. Please do not read this as a recommandation to be sloppy. That, it is not.

I think you guys are missing the forest for the trees.  It hardly matters if it was one profile or a thousand profiles.  Starting yet another multi-page thread on how to parse a particular person's name will only catch the 10% of users who even bother to read these threads...and then you've only addressed one name.  I think this gets back to Ken's more general statement of minor errors being ignored.

Personally, if I'm entering data from the end credits of a US movie and I see three names such as "George Michael Jones", I will always assume that it is "George/Michael/Jones" if there's no entry in my local cast/crew table.  Asking people to research every parsing is too much to ask for.

However, if I do see something like "George Michael/Jones" or "George/Michael Jones" or "George Michael Jones//" in my local cast, I'll research to see if my local cast/crew listing (which is downloaded from Invelos) is invalid or not.

If someone contradicts that parsing in a voting result, I will research it.  Looks like he did and still came up with A/B/C.

Quoting Forget_the_Rest:

Quote:
I actually tried to find a contribution that had been worded like that but couldn't find it.

It's A Little Romance (012569200128) and this all happened a month ago!
A user made a mistake and it got into the database - like that doesn't happen god knows how many times every day! 

Quoting mdnitoil:

Quote:
I think you guys are missing the forest for the trees.  It hardly matters if it was one profile or a thousand profiles.  Starting yet another multi-page thread on how to parse a particular person's name will only catch the 10% of users who even bother to read these threads...and then you've only addressed one name.  I think this gets back to Ken's more general statement of minor errors being ignored.

Fully agree. Really.

That said, since the rant made a big thing of the "I believe" related to name parsing, I went in a bit deeper. Note that I wasn't really going into parsing details, but I was trying to substantiate that it is unrealistic to either double the investigation going into the (any, for that matter) contribution and also that it is unworkable to supply documentation for anything and everything (some others sadi that too by now).

Get real please, a contribution must be as good as it can, of course, but it must not be a university level thesis with footnotes and references for every bit statement...

editted