15 replies on “Greatest Science Fiction TV Tournament – Final week”

  1. I know the numbers are what the numbers are, but really? These are our final two? Both are well-known and influential shows, but, the Greatest SF TV? Feels like an anti-climax.

    • I find myself agreeing with this too, and although the numbers are what they are, I can’t quite work out why we should have ended up with these two in the final round. I really liked B5 and do consider it one of the greats, but I certainly wouldn’t rank it in my top two were I pushed to draw up a list. I also enjoyed The Twilight Zone which, while it was hugely original and innovative, I’ve always suspected its cult reputation is more down to people rememberinng the better episodes and overlooking the not-so-great – and there were some *excellent* episodes. It’s not as if the statistical upsets culled all my best picks either, as just as many of my preferred shows went through on upsets as didn’t, and I didn’t feel that the popular anime series and Treks did as much damage as I was expecting either.

      Ultimately, I think we just have to accept that it was the luck of the draw, or rather the probability that determined the bracket layout. Different combinations would obviously produce different choices, different combinations of statistical upsets, and different tough calls that might have seen shows in the final few rounds here knocked out much sooner. Them’s the breaks with tie-breaker contests like this, I guess. Maybe something akin to the Soccer World cup with mini-leagues drawn at random before playoffs would work better – say everyone picks their top 3 from groups of 8 with the top four shows from each group going into the next round?

      • I’ve been thinking of better ways to do the tournament, through application of a combination of parametric and non-parametric item response theory, but my coding skills aren’t there yet. We do plan to do the fantasy TV and movie tournaments in the next two years, after which I plan to take the top 64 from each tournament and put them together into one massive bracket so everything goes head to head.

        We will also see impact from tastes. Science fiction comes in many forms. I’m pleasantly surprised that the top two are diametric opposites in many ways. Twilight Zone is one of the first sci fi shows on TV, and is about as episodic as any form of entertainment can get. Babylon 5 is relatively recent, and is about as serialized as you can get. TW had a lot of episodes that were just barely out of the ordinary, while B5 contained something we don’t have in our present world in every single frame. I wouldn’t be surprised if the vote here is a perfect match to a vote for episodic vs. serialized storytelling.

        • I have written software to track NCAA brackets. It could potentially be modified to take votes instead of brackets, and use those to come up with winners. This way each person fills out their entire bracket at once. Its then up to the system to arrive at winners for each round based on how popular each entry was.

          See http://bracket-tracker.sourceforge.net/

  2. You have got to be kidding me. I know I’m the resident Trek fanboy, but this feels pretty anti-climactic. There were so many other shows better than these two, not that these two don’t have merit.

    Maybe we need to revisit this tourney with different rules/metrics (as you’ve already said, Blaine) in a few years.

    I swear I can hear “The Price is Right’s” sad trombones playing when I pulled up the voting sheet.

  3. Yeah, somethings not right. As others have stated, I liked both shows but there are definitely shows out there that I liked better. I always sucked at math, so I leave the how up to people more knowledgeable than I.

    • The complicated math is determining how to rank them for round two. Everything after that is pretty direct, and we had very few ties: people choose one show or the other. When everything is said and done in a week, you’ll see the percentages behind every vote added to the brackets.

    • The big problem with this format (series of tie breaks) is that ultimately the only thing it can tell you is the *single* best show overall in the opinion of those who vote. That last part is important – the same bracket would obviously get much different results if it were to be run on an anime appreciation site, or a Trek fan-site, for instance. You also can’t even say that the runner up is second best because if the actual second best show might have met the winner in an earlier round, and that gets even more tenuous if you look at the “Top 4”, “Top 8”, and so on.

      I know that the math used to determine the brackets should prevent that from happening, but the number of statistical upsets we saw in the heats is also clear evidence that there is something amiss between the math and the voting. It might also be interesting to evaluate the impact of genre and see if there is some kind of demographical bias in the site’s readership. For instance I’m not really into superheroes or anime, so quite often when one of those came up against a show I’d watched and liked it it didn’t get my vote, even though it might actually have been a better show overall.

      Ultimately, I think the crux is the first round. Given sufficient data the bracket math should then be able to take over and ensure a fairer knock-out phase with less upsets, but you are still going to have some issues with too many top-rated shows potentially coming up against each other in the same heat of the first round. Perhaps if the first round is merely to gather ratings… e.g. a few weeks of rating a group of shows with a score from 1 to 10 or “No opinion”, with those with the highest total scores overall going into the tie break?

      • The big difference between the TV and movie tournaments is the number of “haven’t seen it” votes. In general, our readers/voters were more likely to have seen any given movie than any given TV show, making the first round more statistically significant. Genre definitely comes into play; anime series often had under four votes that weren’t “haven’t seen it” in round one, but could have 40 votes against them in round two.

        Mathematically speaking, this is the same model used for mission-critical votes for things like Hugo and Oscar winners. The difference is that we just didn’t have enough voters for statistically significant results, nor did we have a consistent number of voters each week in round one.

        Part of me wants to find the time to learn to program well enough to make sure you have to be logged in and vote the entire system on a “better/worse than” basis so each person enters their preferences and sorts things as he/she would sort them. With that process, we can look at correlations between votes and tastes and place things in subcategories, provided we have enough voters, so the system could say “according to your previous voting history, the TV series/movie you haven’t seen that you are most likely to enjoy is ______” and make it an effective engine for recommendations. I just don’t know that we can get enough voters to draw any significant conclusions.

  4. I know I haven’t been paying any attention to this whole thing, but frankly Babylon 5 (much as I love it) could not have existed as it did without The Twilight Zone. Heck, Morden is a direct reference to Rod Serling.

  5. I’d love to see a multiple choice of the last four entries: DS9, TNG, B5 and Twilight Zone. As Blaine said, Episodic vs. Serialized also describes TNG vs. DS9.

  6. I agree with half of this final. I count B5 as the best TV show ever. Last weeks matchup B5 vs TNG probably should have been the final.

  7. When Galactica 1980 and Come Back Mrs Noah got eliminated so soon, I lost all interest.

Comments are closed.