I might be confused here, but as I was reading through a recently discovered review on one of my games on IFDB (Milliways), I noticed what appears to be a miscount in the number of ratings. The game lists two 2-star, three 3-star and five 4-star ratings - but if I count the actual number, it’s (including off-site ratings and an author review, which I think is probably fair, although I can remove the author rating) really two 2-star, five 3-star and seven 4-star ratings.
But this isn’t even the end. If you exclude all off-site and author ratings, the number is still off: two 2-star ratings, four 3-star ratings and four 4-star ratings. So even if you go for the basic rating numbers, it’s still off. And I think it’s miscounting in the overall rating.
The reviews on Milliways? The other two are going to reviews on Intfiction. The top link always sends you to the hosting website’s homepage. The “see full review” later on sends you to the actual link.
The review I did was a summary of my postmortem. I put a rating, not thinking it would give a rating. I don’t think it does. Maybe it does? (You actively have to select “By the Author” which seems really weird. Anybody could click that, and the author doesn’t have to click it. Which is weird. Unless it’s because of Cragne Manor …)
I believe on iFDB you can “rate” a game - click a number of stars - without reviewing. You can write an IFDB review by clicking to review and writing it on IFDB as a player - which I think is different from an “off-site” review that’s plugged in by editing the IFDB page where it’s submitted as part of the game entry rather than submitted separately without editing the entry.
I don’t know for sure, but I wonder if it’s counting curated “off-site” review links to another page differently from a review written on IFDB not linking anywhere.
Similarly, there might be ratings that are not counted toward the total - say if a troll was spamming one-star reviews. There also I think is a box that reviewers can check that don’t count their rating toward the average - like if they were a tester.
I would agree and it would all be fine, except that it’s not that maybe one person did that, but actually if it were just normal reviews, it’s counting one too many normal reviews (specifically, a 4-star rating that isn’t there). It might turn out to all be okay, but I feel slightly confused. Because also then, why does it ask the off-site review to give a star rating? It would not get counted, which I think is a little bad.
Only other thing I can think of - there’s an “embargo date” setting where someone can set a review to appear on a specific date. Maybe that rating counts even though the review is not visible? [1]
These are just guesses based on my experience with IFDB. hopefully you’ll get better input from the IFDB admins… ↩︎
The code for IFDB is open source. I’m poking around to see if I can figure out what’s going on (I bet it’s probably your own author review showing up in one view but not the other) but you can look at the pages here:
and here:
among others.
Oh and this page:
seems to me like the external reviews are kept in a separate list and not added to the overall ranking.
First off, note that there’s a difference between “ratings” and “reviews.” Some written reviews have no rating; some ratings have no written review. We show two separate counts at the top of the page. Millways has 9 ratings and 8 reviews.
But that’s not all. There are five reasons we might exclude a rating from the count.
Written reviews without ratings don’t count, naturally.
Off-site reviews don’t count. You can create off-site reviews by editing the page and adding them as links.
“Special” reviews don’t count. You can mark a review as “special” by editing the review with a “Special Review Type,” which can be either “From the Author” or “External.”
Embargoed reviews don’t count. You can embargo a review by edting your review and setting an embargo date in the future. When the date passes, your review will be unembargoed, posted publicly, and counted as a review.
“Omitted” reviews don’t count. You can omit a review by editing your review and checking the “Omit my rating from the game’s average” box.
There are 15 reviews rows for Milliways in the database.
There’s 1 review with no rating.
There are 3 off-site reviews.
There’s 1 “special” review, your “From the Author” review.
There are 0 currently embargoed reviews. (Several of the reviews are past their embargo date; they count.)
There’s 1 “omitted” review. It’s this one. “Note: this rating is not included in the game’s average.”
15 - 6 = 9 ratings.
The histogram includes special reviews; I think that’s probably a bug. But, as a result, it’s taking the 9 ratings and adding in your own 4-star review, giving a histagram of 5 4-star reviews, 3 3-star reviews, and 2 2-star reviews, for a total of 10 ratings. https://github.com/iftechfoundation/ifdb-suggestion-tracker/issues/471
There are 12 written reviews in the database. 3 of them are off-site reviews, and 1 of them is your “From the Author” review, and so, there are 8 written member reviews.
You can pry my one-star reviews of my own juvenilia from my cold, dead hands.
Well, I’m kidding, of course. But I don’t think it’s unreasonable for an author to be given the chance to provide their own assessment of their own work—it’s just one voice in the crowd.
Oh, I don’t remember having the option to designate a review as “from the author” when I applied those ratings. I wonder if this feature was added after I had done so? I might need to take another look at my profile.