Tabulated review threads sorted by average score

Discussion in 'Trek Literature' started by Sho, Jan 13, 2012.

  1. Defcon

    Defcon Rear Admiral Rear Admiral

    Joined:
    May 9, 2003
    Location:
    Germany
    @JWolf: I'm just using the system that has been used here for quite some time now, and for the purpose of what I want to try to achieve, filling up the ranking site, it has to be done that way. (What you propose would mean a lot of work not only for the poor soul who would have to repost the threads, but for the moderators, would flood the board with "old news" at the risk of drowning out current discussion and would have the heavy downside of losing many votes)

    @Zarkon: Thank you for your comments.

    BTW the review thread for The Belly of the Beast has been posted.
     
  2. Thrawn

    Thrawn Rear Admiral Premium Member

    Joined:
    Jun 15, 2008
    Location:
    Washington, DC
    If it's good enough for Amazon and Goodreads, it should be good enough for us. And the validity of several years of ratings to compare with is much more valuable than the added "accuracy" (which, if we're talking about opinions, is kind of a poor concept anyway; you're really talking about "granularity", which isn't the same thing).
     
  3. Csalem

    Csalem Commodore Commodore

    Joined:
    Mar 28, 2006
    Location:
    Dublin, Ireland
    How about slotting in Ship of the Line at some point since it pops up in threads quite frequently. Might be good idea to have one thread dedicated to discussing and reviewing it?
     
  4. JD

    JD Fleet Admiral Admiral

    Joined:
    Jul 22, 2004
    Location:
    Arizona, USA
    I really don't see the need for more options either. Honestly, I think what we have now works fine. All adding more options would do is just make things more complex, without really giving anymore information that five options doesn't already give. I think it's pretty easy to figure out if people liked the book/comic/whatever based off of what we have now, and really, that's the purpose of reviews.
     
  5. ronny

    ronny Fleet Captain Fleet Captain

    Joined:
    Mar 21, 2006
    Location:
    San Francisco, CA
    Why stop at 10? Let's go with 1000. Then we can see what people really think...

    Seriously, I think Thrawn has the killer argument for leaving it as is, if it's good enough for Amazon and Goodreads there's no reason I've heard that 5 doesn't work. And no reason for Sho to do all the work involved in updating his scripts and no reason for the mods to spend time creating new polls and combining the old ones.

    We're rating tie-in books, not putting a man on Mars. 5 stars seems to be doing the trick.
     
  6. JWolf

    JWolf Commodore Commodore

    Joined:
    Oct 9, 2005
    Location:
    Massachusetts, USA
    I've been reading on Goodreads that it's not good enough for a lot of people and I have to agree. For example, a 7-star rating is much more accurate then a 3- or 4-star rating.
     
  7. Thrawn

    Thrawn Rear Admiral Premium Member

    Joined:
    Jun 15, 2008
    Location:
    Washington, DC
    More granular. Not more accurate. There is no independently measurable quantity being assessed.

    What the hell difference does it make if something averages 3.2 on a scale of 5 compared to 4.5 on a scale of 7?

    Besides, oddly enough, I'm a teacher and I've done some research on survey and assessment design. At least for the average high school and college students, more granularity does not result in any higher statistical correlations based on preferences; the official recommendation of the design document I read literally was that, when surveying people about likes and dislikes, a scale of 1-5 be used, as anything more was overly complex and produced no more useful data.

    Which is probably why the biggest and most competitive sales companies in the world, which rely on user reviews to sell things, do exactly that.

    So perhaps you are so granular that you have precisely plotted your opinions to the nearest seventh, but the vast majority of humanity doesn't give a shit, and that's been statistically proven.
     
  8. JeBuS

    JeBuS Lieutenant Commander Red Shirt

    Joined:
    Sep 4, 2013
    I hate posting with a simple "This!" but really, what else is there to say? Your post had everything necessary, and this line is an excellent summation.
     
  9. Stevil2001

    Stevil2001 Vice Admiral Admiral

    Joined:
    Dec 7, 2001
    Yeah, my university recently went from a 10-point student eval system to a 5-point one, for reasons like what Thrawn mentioned.
     
  10. JWolf

    JWolf Commodore Commodore

    Joined:
    Oct 9, 2005
    Location:
    Massachusetts, USA
    It's not about the average. It's about a more accurate rating system. As I said before, rating something 7 of 10 is more accurate then rating something 4 of 5.
     
  11. zarkon

    zarkon Captain Captain

    Joined:
    Mar 24, 2011
    Well, it seems from this and previous threads that you're the only one who favours /10, everyone else either prefers /5 or doesn't care enough either way to bother posting

    probably time to stop tilting at windmills dude
     
  12. Christopher

    Christopher Writer Admiral

    Joined:
    Mar 15, 2001
    As you've been told, that's a misuse of the word "accurate." You're making the common mistake of confusing accuracy with precision. Accuracy is about whether you have the right answer, while precision is about how narrow the margin of error is. For instance, if you say that the value of pi is somewhere between 3 and 4, that's accurate but not at all precise. But if you say that the value of pi is 3.9658751356884309, then that's extremely precise but not at all accurate. Increasing precision does not necessarily increase accuracy.

    Indeed, too much emphasis on precision can work against accuracy. For instance, the common belief that normal human body temperature is precisely 98.6 degrees Fahrenheit is quite wrong. In fact, it varies from person to person and fluctuates over the course of the day. What medical science actually determined was that human body temperature has an average value of approximately 37 degrees Celsius. And then, when that was converted into Fahrenheit, someone made the mistake of taking that accurate but approximate figure as a precise one -- as exactly 37 C, which converts to exactly 98.6 F. It would be more accurate to say that body temperature is about 99 F, give or take, but the presence of the decimal point created a false perception that it was a precise and unchanging figure, rather than something that naturally varies. Thus, too much precision created a belief that was inaccurate.

    In this case, we're talking about personal opinions. It's problematical to talk about "accuracy" for something like that, because there is no objective "right" answer for matters of taste. Heck, I've often read a book a second time and found I liked it substantially better or worse than when I read it the first time. Even a single individual's tastes are variable enough that a broad and imprecise set of ratings is more likely to be accurate -- i.e. to encompass the right answer -- than a more granular rating that may include your reaction on one reading but exclude it on another. So your belief that more precision equals more accuracy is pretty much backward. Part of being accurate is acknowledging the margin for error.
     
  13. JD

    JD Fleet Admiral Admiral

    Joined:
    Jul 22, 2004
    Location:
    Arizona, USA
    Honestly, I think even only two options, I liked it/I didn't like, would work. IMO when comes to stuff like this that's really all you need to know. Everything more than that is simply degrees of like or dislike.
     
  14. Defcon

    Defcon Rear Admiral Rear Admiral

    Joined:
    May 9, 2003
    Location:
    Germany
    I have a question:

    I have noticed that the Classic Review threads get the vast majority of their votes and comments in the fist few days of being posted, with very few votes (if any) or discussions coming in later. Do you think I could/should tighten the posting speed a bit from one every two weeks to one every week?
     
  15. zarkon

    zarkon Captain Captain

    Joined:
    Mar 24, 2011
    Once a week sounds fine to me.
     
  16. Skywalker

    Skywalker Admiral Admiral

    Joined:
    Feb 24, 2005
    Yeah, I don't think it would be a problem. There are a lot of books to get through, and a weekly schedule as opposed to a biweekly one will move a lot faster.
     
  17. Defcon

    Defcon Rear Admiral Rear Admiral

    Joined:
    May 9, 2003
    Location:
    Germany
    So one thread a week (most likely on the weekends) it is.

    Updated schedule:

    Week 9/2014: TOS: The Pandora Principle *Carolyn Clowes (most likely later today or tomorrow)

    Week 10/2014: TNG: Resistance * J.M: Dillard

    Week 11/2014: DS9: Avatar, Book One * S.D. Perry

    Week 12/2014: VOY: String Theory, Book Two: Fusion * Kirsten Beyer

    Week 13/2014: ENT: What Price Honor? * David Stern

    Week 14/2014: I.K.S. Gorkon: Honor Bound * Keith R.A. DeCandido

    Week 15/2014: Titan: The Red King * Andy Mangels & Michael A. Martin

    Week 16/2014: New Frontier: House of Cards * Peter David

    Week 17/2014: TOS: Federation * Judith & Garfield Reeves-Stevens

    Week 18/2014: TNG: Ship of the Line * Diane Carey

    Week 19/2014: DS9: Avatar, Book Two * S.D. Perry

    Week 20/2014: VOY: String Theory: Evolution * Heather Jarman

    Week 21/2014: ENT: Surak's Soul * J.M. Dillard

    Week 22/2014: I.K.S Gorkon: Enemy Territory * Keith R.A. DeCandido

    Week 23/2014: Titan: Orion's Hound's * Christopher L. Bennett

    Week 24/2014: S.C.E.: Fatal Error * Keith R.A. DeCandido
     
  18. Sho

    Sho Fleet Captain Fleet Captain

    Joined:
    Sep 8, 2006
    Location:
    Berlin, Germany
    I think another reason to keep the polls simple is that it makes it a bit easier to parse the bar graph in the table at a glance. I'd like it if the ranking table would incite folks to become curious about the explanation behind a certain ratings profile and click through to the thread to find out, and the five bins make it easy to see if a book is e.g. controversial by having spikes at the front and back of the profile. Sure, you could derive the same viz from more granular data, but a granular poll doesn't force people to make their mind up as much so I think it'd act to even out/dilute the distribution. So I think five options also contribute to making the results more interesting actually.

    I think this sets the ranking table apart from just a generic voting site; it's intimately coupled with a discussion forum and you can dig deeper into actual discourse about a book in a way that comment systems on voting or shopping sites usually don't allow. None of us have our book talk at Amazon.
     
  19. trampledamage

    trampledamage Clone Admiral

    Joined:
    Sep 11, 2005
    Location:
    hitching a ride to Erebor
    Agreed - it's the comments I take more value in, the voting is just a starting point.

    Also, I can say for certain that there's no way we're recreating all the polls :eek: Learn to love the five point system, it's here to stay :techman:
     
  20. Defcon

    Defcon Rear Admiral Rear Admiral

    Joined:
    May 9, 2003
    Location:
    Germany
    I always forget to add this to my schedule posts:

    If you want to "nominate" some novels for the Classic Review Threads feel free to do so here. :)

    As you can see I already included Ship of the Line, as Czalem suggested.