Scoring full points for the same sentences day after day

I think Clozemaster would be much improved if the ability for Pro users to earn 32 points repeating the same sentences day after day was removed.

I’m not suggesting that it shouldn’t be possible to practice the same sentences every day if you feel you need the revision, only that they should revert to minimal points until a reasonable period of time (3 months? 6 months?) has elapsed.

ETA: maybe 3 or 6 months is too long for the language pairings with small collections. Daily is still too short, however.

1 Like

Thanks for the feedback! Do you mean for users who set the max review interval to 1 day? At the moment you do score half the points for sentences played before they’re ready for review.

We’re also working towards adding leaderboards for number of sentences played and mastered, as well as a streak leaderboard, to allow for competition on other metrics besides points.

2 Likes

Yes, that’s what I mean. I think it should be possible to set reviews for very short periods if people want to, but not to earn maximum points every day for reviewing the same sentences.

2 Likes

What about changing the minimum amount of time for mastered sentences to appear in reviews to 7 days(for example)? So reviews don’t get mixed with mastered sentences ready to review and words you have already mastered 1 or 2 days ago.

1 Like

Personally I think being able to score full points for the same sentences every week is still much too frequently. I’d suggest no more often than every few months.

I don’t think this addresses the main issue, as those who are there for the points can simply delete the deck and re-add it again and play the very easy sentences again the very same day.

On the other hand, people who are working fiercely on shorter decks for less-common languages and who still benefit from those reviews will have much less incentive for working on those sentences. So they lose twice, for already having fewer sentences to work on, and for expecting to score the bare minimum after just a couple of months playing. Learners of the big languages will always be on top.

Don’t get me wrong, I have enough reviews on major languages for this not to be an issue whenever I decide to play.

1 Like

They CAN… but unless I’m missing something they’d be back to 8 points per question. They’d have to build back up to 32 over the following 4 days.

I’d agree that 7 days is too short for system rorters (let’s be blunt, that’s what we’re talking about here) but every few months would be to my mind like throwing out the baby with the bathwater. Yes, it’s true that using Clozemaster shouldn’t be about scoring points (or not exclusively anyway), but getting reduced points just because you’re reviewing within several months feels like a disincentive. I’d agree that it would be justified if people remembered perfectly something that they had seen multiple times… but in real life it doesn’t happen. For a genuine student there are thousands of words competing for your attention, plus everything else in your life competing for memory space. Granted my memory isn’t the best in the business but I’m sure I’m not the only one who has seen more than a few words which are “100% mastered”, but which we have no memory of. (Mostly these would be words which we have nothing to connect them to; no common etymology with known words, no sounds that relate to the meaning, and often which are important but not used frequently.) If you had to wait several months to review them… that would not be a good thing.

It’s one reason why I never play the Favourites collection; 2 points per question makes it feel like you’re slogging through thick mud as punishment. (Though I do get that the 2 points per question was (probably) implemented to avoid the kind of rorting that we’re talking about.)

The other reason that I don’t concern myself with Favourites any longer is that once you get a few hundred entries in there stretching back over months you can rarely remember much less find what you may have added 2, 3, 4 months ago. That’s not a criticism, incidentally, I do think that HAVING a Favourites group is both a good and necessary idea, it just doesn’t work for me. I’m finding that a better approach is to create custom collections broken up by topic.

On which point, by the time I hit 32 on one of those I’ve seen the question at least 4 times. I’m fine with not seeing it again for 3 or 4 weeks (which is the minimum range in which I set my mastered review anyway), but not 2, 3, 6 months from now. That would not be useful to me. And scoring fewer points were I to review it within that period would feel punitive and a disincentive.

Let’s be completely blunt; there are people who are rorting the system mercilessly to get their names into the leaderboard. I don’t think there is any way to stop that from happening. If this change were to be implemented I can think of at least one way around it, though it would require a lot of setup work. I do however agree that making it harder for them to do this is worth doing.

Sometimes the system-gamers niggle because it means that people who work hard at their learning rather than at gaming the system won’t have their day in the sun. But most of the time I think “so THAT’S what you want to do with your life? Game a system to get your name (or more commonly handle) to the top of a leader board that most of the planet will never see? Hmm. Interesting.” I find it impossible to take the leaderboards too seriously because it’s blatantly obvious how much some of the scores are about rorting the system rather than studying. I keep tabs of where I am on the overall standings, and who is ahead of me and their score, but that’s just a variation of the old army training technique; “Say to yourself ‘I’ll run to that tree, then I can rest’. And when you get to that tree, point to another one and say ‘I’ll run to THAT tree, and then I can rest’” That just keeps you going. However there’s really no incentive to aspire to being “Number 1”. Still, as I said… there’s no reason to make it easy for someone if they want to spend their days that way. It’s unfair to people who really want to learn.

2 Likes

You can set a sentence to fully mastered with a shortcut key the first time you play it, and then play it the next day, and the next, and the next, all for 32 points. I think having a leader board for mastered sentences will probably only encourage this.

1 Like

Ah, you’re right. I’ve never played around with the percentage mastered setting so I hadn’t thought of that. (Other than when I commit a typo and bang on enter twice too fast, at which time I hear the dreaded “Bleep-bloop” sound and it drops to zero… but that’s unintentional.)

I haven’t seen that suggestion, but agree that would be the effect.

1 Like

Interesting discussion, but I’m not sure that there will ever be a set of (flexible) rules where everyone who gets into the top ten on a leaderboard are felt to be ‘legitimately’ there, especially by those who worked hard to get there, and are displaced. Or might be displaced soon.

A possible solution? Perhaps there is too much focus in the clozemaster boards on the top ten. As the site gets older, it will be discouragingly harder for people to work their way into the top ten. Especially for those who don’t fiddle with parameters to amp up their point scores.

Another approach? See the leaderboard at Memrise. Here’s an example:
Leaderboard - Learn Romanian for Real! - by coelyoung - Memrise

Notice that it shows the top 50… and if you’re not there, you can click ‘more’ to drill as far down as you want. For beginners, you can see exactly where you are. And getting into the top 25 is a real accomplishment. (Speaking as 75th on the Clozemaster leaderboard…)

Think of this as an alternative to fiddling with the rules…

Romanophile

2 Likes

I think you’re really onto something here. I’m not sure about the top 50 thing; the only issue I see with that is screen space on smaller platforms. But at the moment we have the scoreboards for “All time”, “This Month” and “This week”. One disincentive to gaming the system might be to drop the “All Time” list completely - or more precisely shove it off into some “Hall of fame” page that people need to click through to - and have it as “last 7 days”, “last 30 days” and “last 365 days”. This would also stop that sinking feeling at the start of a week or a month when your score drops to zero and you have to start all over again. (Although admittedly that CAN be useful; Mondays and the first of the month are the only days that I have any idea how many points I scored for that single day. In theory, at least, since I’m actually 10 hours off UTC and usually do some questions long before the clock ticks over.)

It would encourage people to study more regularly since the effort that they put in every day contributes to their currently visible ranking scores. It would discourage people from gaming the system because as soon as they stop, their name will slide from view… and who would want to do that day in and day out for years on end? And it would encourage new students who would have a chance of a place in the sun rather than looking at the top of the leader board and thinking “what’s the point?”

The hall of fame could show (say) the top performers by year giving the more persistent students a permanent spot of glory. Here a top 50 would indeed be useful.

My bet is that you’d end up seeing far fewer new 8 digit scores that way.

Actually that’s an interesting side point. It’s impossible to say how many points you can score on Clozemaster and still be learning, because it depends on how you play it. (Review frequency, FFT vs collections, and let’s not even start on the impact of Cloze Reading.) But just after I reached 1 million points I had a pang of curiosity. I was at that time 30th in Italian and 521st overall. So give or take, 520 other people across all languages have hit a million. As of now I’m at a bit over 1.2 million, and am 416th overall. Yes, it’s likely that I’ve simply overtaken some people who are studying more slowly but I’ve also noted that a lot of the people that I passed beyond the million mark were just dormant accounts. I think it’s safe to say that the majority of people that I passed between 1 and 1.2 million aren’t with us any longer. 100 or so people isn’t a huge number but it’s still roughly a fifth of those who went over a million. I suspect (but can’t prove) that for a lot of people it was something of a psychological barrier; “Well, I’ve done a million points, that’s enough”. IF I’m right about that (and I don’t deny it’s a pretty big if), de-emphasising the end goal and re-emphasising the process by rolling leaderboards may, MAY, possibly, make that less of an issue.

2 Likes