6 ms·

> Is that the intended behavior? Yeah, I'm afraid so. At the start, selection two would always be the same as the algorithm doesn't know anything about the sta

> Is that the intended behavior? Yeah, I'm afraid so. At the start, selection two would always be the same as the algorithm doesn't know anything about the standing of any episode; it's trying to decide where in the array to place it (above or below selection two). It might be worth seeing if I can randomise the episodes displayed, if only so it doesn't feel so repetitive.

bravura 3y ago
Make it slightly fancier and assign a “similarity” between users. Start with uniform similarity, and as preferences are made, adjust them. Then you have personalization.
- pocketarc 3y ago
  I'm not sure what you mean by this - are you saying that the episodes that get displayed would be based on what is likely that other users would've picked for the same show?
  - bravura 3y ago
    I’m saying that instead of one objective ordering, you have a subjective per user ordering. The decisions made by other users get a weight assigned to them, which is individual to each logged in user. So every users viewpoint is personalized. (Apologies for the short explanation, I’m on mobile. If you want to ask more you can email me.)
boomboomsubban 3y ago
Why does selection one get randomized if you say selection two is better, but then jump back to the original sequential if you say selection one is better again? That behavior felt bizarre.
- pocketarc 3y ago
  Selection one jumps between the start and the end of the show repeatedly until you make it to the middle (selection two), after which point it'll move on to getting you to rank the best episodes, and then the worst episodes. Definitely looking into seeing if I can come up with something better though! The problem is making sure that whatever algorithm is picked remains as close to O(n log n) as possible. Randomising options in a way that makes require a lot more comparisons would be far worse.
  - ysavir 3y ago
    > Randomising options in a way that makes require a lot more comparisons would be far worse. For the algorithm, but not for the people taking time to do the rankings. Which do you want to prioritize?
    - pocketarc 3y ago
      I was prioritising for taking less time total, but you’re right, that doesn’t matter if the person gets bored and leaves. I’m tinkering with it now and I think I have a good solution to the problem. I’ll be deploying it soon!
      - gamerDude 3y ago
        Maybe you can pull in episode rankings from another site to seed a first ranking and then make an algorithm to find where you disagree with the norm.
bena 3y ago
I've done something similar. You're essentially doing A/B comparisons across the entire set. It looks like you have it so you're basically setting where "B" is before moving on to the next item. This isn't strictly necessary. You could just generate a novel pair every time and ask the user to choose between them. The thing is that you'd need a way to track a user. So you can make sure that user hasn't seen a certain pair already. Once you've exhausted all the pairs, you'll know exactly how to sort the array. You'll have an idea before. You might have an issue with circular lists though. People are fickle. You could have someone who says that A > B > C > A. In this case, I'd allow for repeat pairings after a certain amount of time. To allow the person to reevaluate essentially. You could also take the comparisons across all users and compile a general sort of "best of" ranking.
- pocketarc 3y ago
  > The thing is that you'd need a way to track a user. The page you're on already knows what comparisons you've made (otherwise how it would move forward), so this is entirely possible! I've come up with a way to randomise it (by just picking a random element from an array of comparisons that haven't yet been made), that's the next step. I deployed it earlier, but there was a bug with it so I've had to rollback until I can look into it.

bravura 3y ago

Make it slightly fancier and assign a “similarity” between users. Start with uniform similarity, and as preferences are made, adjust them. Then you have personalization.

pocketarc 3y ago
I'm not sure what you mean by this - are you saying that the episodes that get displayed would be based on what is likely that other users would've picked for the same show?
- bravura 3y ago
  I’m saying that instead of one objective ordering, you have a subjective per user ordering. The decisions made by other users get a weight assigned to them, which is individual to each logged in user. So every users viewpoint is personalized. (Apologies for the short explanation, I’m on mobile. If you want to ask more you can email me.)

boomboomsubban 3y ago

Why does selection one get randomized if you say selection two is better, but then jump back to the original sequential if you say selection one is better again? That behavior felt bizarre.

pocketarc 3y ago
Selection one jumps between the start and the end of the show repeatedly until you make it to the middle (selection two), after which point it'll move on to getting you to rank the best episodes, and then the worst episodes. Definitely looking into seeing if I can come up with something better though! The problem is making sure that whatever algorithm is picked remains as close to O(n log n) as possible. Randomising options in a way that makes require a lot more comparisons would be far worse.
- ysavir 3y ago
  > Randomising options in a way that makes require a lot more comparisons would be far worse. For the algorithm, but not for the people taking time to do the rankings. Which do you want to prioritize?
  - pocketarc 3y ago
    I was prioritising for taking less time total, but you’re right, that doesn’t matter if the person gets bored and leaves. I’m tinkering with it now and I think I have a good solution to the problem. I’ll be deploying it soon!
    - gamerDude 3y ago
      Maybe you can pull in episode rankings from another site to seed a first ranking and then make an algorithm to find where you disagree with the norm.

bena 3y ago

I've done something similar. You're essentially doing A/B comparisons across the entire set. It looks like you have it so you're basically setting where "B" is before moving on to the next item. This isn't strictly necessary. You could just generate a novel pair every time and ask the user to choose between them. The thing is that you'd need a way to track a user. So you can make sure that user hasn't seen a certain pair already. Once you've exhausted all the pairs, you'll know exactly how to sort the array. You'll have an idea before. You might have an issue with circular lists though. People are fickle. You could have someone who says that A > B > C > A. In this case, I'd allow for repeat pairings after a certain amount of time. To allow the person to reevaluate essentially. You could also take the comparisons across all users and compile a general sort of "best of" ranking.

pocketarc 3y ago
> The thing is that you'd need a way to track a user. The page you're on already knows what comparisons you've made (otherwise how it would move forward), so this is entirely possible! I've come up with a way to randomise it (by just picking a random element from an array of comparisons that haven't yet been made), that's the next step. I deployed it earlier, but there was a bug with it so I've had to rollback until I can look into it.