Bumping this up in case there's still anyone with some feedback...

Interesting to give a probability to appear to the bad rating quotes even if this probability is low. Thanks.

The algorithm could adapt the quotes in a smart way using one or more of these ideas:
- The size and difficulty of the quote according to the number of accept or reject of the user or adding in the profile some preferences about the size of the quotes preferred.
- This can be also used if you have some kind of meta-tag or keywords associated to the tags. (I.e. one user can love to make quotes about literature, or famous quotes, or quotes from films, jokes,...)
- Collaborative Recomendation. If one user "similar to me" (similar because of coincidence of evaluations) liked one quote, this one can be proposed to me.

Only some ideas to grow up this interesting tool.
Best Regards
David Bueno

I appreciate the ideas David!

That's good to know that something has changed. I hadn't known about such an algorithm. I applaud your leniency for quotes which receive "bad" markings because the methods by which people make those judgements is quite personal. Quotes which are difficult to accomplish and don't give me the sense that I'm a fast and accurate typist appear to me actual "good" quotes because they reveal to me places that I struggle with. I would surmise that keyhero does not exist to pet people's egos about obtaining fast speeds.
So I'm happy to hear that. I have however, not very often given a quote any marking because my own personal biases may sway decisions. I will mark quotes as good which provide me the opportunity to type uncommon words or combinations of words.

I saw one person provide a quote which was requesting that there be less grammatical errors in the quotes, and that actually seems counter productive to me. On the one hand, if a person is trying to learn the formal grammatical rules of English then it makes sense of course, but part of typing, and by extension, writing, is to create creative formulations of phonemes and thoughts symbolized by words. I would wager that quotes with some incorrect grammar could cause people to notice and readjust their preconceptions about what constitutes expression, and if meaning is something which always must be bound to formal rules. In a sense I would say we use language as a vehicle to condense our inner imagistic and emotional landscapes into a hopefully communicable form. Although, with some experience, one might figure out that knowing the same language does not necessarily procure one understanding of others, or vice versa.

I have absolutely no idea of how any of this works, but would it be possible to switch to star ratings rather than like/dislike? Or add a neutral face vote that makes the vote move closer to 2.5 stars regardless of its position?

I appreciate the feedback. @neveronground, I like the idea. Being able to choose 1-5 stars would allow greater accuracy. Stay tuned...

We are moving forward with the ability to rate quotes on a 1-5 scale rather than 0 or 5. This should allow bad quotes to be filtered out of the test's pool quickly while allowing ones that are initially mediocre to be given an opportunity to be included in tests. The end goal is higher quality quotes without automatically dismissing bad quotes.
The one thing that remains the same is that *we need you to keep rating quotes*. Rating each quote that you test on is important for weeding out bad quotes and ensuring the best quotes show up most frequently.

I'm not seeing the buttons to rate quotes anymore. I used to see them before, but they have disappeared for me. Is it only me, or did it happen to others too?

Edit: Nevermind - I found that now we click the stars to choose a rating

Glad this site is in active development :)

Thank you for the encouragement!

I like It and it seems good.

Honestly the more I use the site, the more I realize I keep getting a lot of the same quotes. It's like I'm getting maybe 20-30 quotes and they are just cycling.

When did you first notice this happening? We did make a change in the past week that increases the probability of receiving a highly rated quote, but it shouldn't allow you to cycle through 20-30 quotes. More like 2500 quotes.

I've just been trying to find out some way to block out quotes that I don't want to type. :D

Hey sorry for long time to reply. If you are still interested: I may have exaggerated a bit in that comment. I don't know if it's really just 20-30 quotes that get recycled. Idk what I was on with that.

I think I mainly get new quotes. But there's a handful of quotes that I see every time, or every other time, or if not then fairly often.

I'd say maybe 45 (very rough guess) percent of my time taking tests, is on tests that I've seen a bunch of times before.

Could have something to do with the fact that I'm a very sporadic user. I'll do like 10-20 tests once every now and then. And I skip a lot. So maybe the algorithm hasn't had much time to figure me out? If it is some kind of machine learning situation anyway. Idk much about anything so don't mind me if I make no sense lol.

Hey, thanks for the feedback. Makes perfect sense actually. I'll look into it.