I will no longer support reports from user info. So far userinfo had the feature for a user to report a user who they think is a cheater (this feature was developed by the authors of the original leaderboard), however for now there is no criteria for banning or restricting users so there are no users who should be banned, so if you find an offensive user please simply hide them by the Hide button. (If any regulations are needed we need to discuss and make a guideline.)
3 Likes
Instead of having the league end be based on the users local time, can it be changed to a common time? The issue is that once my reviews stop counting, I can not climb the leaderboard, but I can fall, because other people havenât reached the end time. Thus, somebody, say 30 ranks from the cutoff, can fall down all the way past the cutoff (both the promotion and demotion one) if enough people study after my time is over and before their time is over. This becomes a larger issue for people in timezones which are significantly ahead of GMT and âfavoursâ being in a timezone behind GMT.
Also, has the close window with Esc and Cmd + W not been implemented yet?
2 Likes
Technically it is possible, but isnât it inconvenient to use a common time because the end time would be different in each country? e.g. the end time may be about 12:00 noon in some countries.
This function is still under development, please wait.
The start time would compensate enough for that.
1 Like
More precisely I mean that there may be unfairness if users want to compete on the last day.
e.g. If users want to compete on the last day, in some countries they have to stay up from midnight to 5:00 AM. In countries that end at 21:00 they can review as usual and it is easy. In countries that end at noon they cannot participate in the first place, plus the calculation is from noon of the previous day to noon of the next day, not the whole day today (12:00 to 12:00).
In the case of local time this problem does not occur, and all users can review at their usual time, and it is calculated with todayâs cards (0:00 to 0:00).
Iâm suggesting using something like the reset time but globally.
If the reset time is GMT, then people ahead of GMT will review for the next day the day switches on GMT and people behind GMT will review for the previous day.
Effectively, the reviews shouldnât even change, considering the people ahead of GMT start earlier but also end earlier and vice versa for the people behind.
If this was the case, then the users rank shouldnât change, yes?
Yesterday (league end), I was at 265 in Gamma league when the league ended for me, at a comfortable 25 ranks above the cut-off.
Today, when I check, Iâm at 284 in Gamma league, just barely above the cut-off.
1 Like
Well I agree that there is that problem, but isnât it almost the same problem anyway?
e.g. In one userâs country the closing time is 12:00 noon, and they last reviewed at 0:00. Other countries review between 0:00 and 12:00 so the rank changes during half a day. Users cannot review as they can in other countries so they feel it is unfair.
I added an option to open the leaderboard window to another add-on BreakTimer. e.g. after 100 cards reviewed it will auto open the leaderboard window. If you use this need to update the leaderboard to the latest version. add-on:
Break Timer - After 10 cards, take a 3 minute break (Created by Shigeŕś)
Users have again send me direct messages requesting that we strengthen the anti-cheat system for users who are reviewing too fast. (e.g. 0 to 1 sec/card)
Basically my development prioritizes user requests, in the Leaderboard the number of requests for such tightening has been the largest so far, and it is very rare to receive requests that such restrictions are not necessary.
If you have any ideas or requests about whether these restrictions are necessary or not, or about changes in the way XP or other calculations work, please post them in this thread, and I will use them for development reference.
So far there are features for anti-cheat, fair competition, or no distractions.
- Hide users by option
- Show
and warn users with average review of 2 seconds or less - Force exclusion of reviews under 0.001 seconds
- Calculate overall by league XP
- Hide medals by option
- Sort global leaderboard by user preference
- Ban users and make leaderboards inaccessible
I donât think problem is when you review cards individually for less than 1 second. I have many simple cards that I can review in under 0.2 sec.. Especially some cards in my language decks.
What Iâm concerned is all about AVARAGE. When you review 5k+ cards 14 days straight with less than 2-second avarage. (Some cases for people in top ladder itâs actually 0.5sec avarage per day).
Technically I also probably can do that by keep reseting same old super easy deck 1000 times in row but that doesnât actually teach me anything so itâs complete waste of time.
Alternatively I could also use addons like Speed Focus and watch cards like slideshows (Which again is complete waste of time) because just watching something is passive learning.
I donât recommend banning anyone, but in current situation how everything works right now people should be more rewarded XP for time investment.
1 Like
There are several possible rational reasons for the large number of quick reviews that occur in less than one second.
-
For medical students who are aiming for a high score on their exams, there is a way to review all the cards using a filter deck before the exam. This is faster because most of the cards are already known, and more reviews will increase the exam score.
-
If the learner has a very high desired retention rate, or if they press Hard a lot, the learning workload increases, so they will review more than the typical learner. The same is the case if there are many Leech cards.
-
If the learner is learning from the basics, there could be a lot of easy cards. (e.g. The learner is learning a new language)
These are effective learning methods for better exam scores and retention, so I think there is no problem.
There are two types of Anki users, quick reviewers and long reviewers. So far there is no clear consensus on which is better.Ideally we would like to develop a system that is better for the long reviewers and a system that is better for the fast reviewers.
And there is another problem with changing the XP calculation method, if we change the calculation method and an unexpected bug occurs, the league is likely to be unusable. So it would be ideal to keep the current league calculation method as it is and add new system. In this case, if the new feature is broken by a bug, the basic feature can still work without any problem. (e.g. adding a new medal system or a new league) The gamification mode is developed in such a way, this feature can be optionally disabled, so even if a bug occurs the leaderboards will still work fine.
1 Like
Yeah I suppose it comes down to different learning methods. I admit I have no understanding of this spam method 
Like personally I also revise stuff I already know. But I also need to constantly study new information which is reason my avarage is like 20 sec per card. I have to study everyday something new. Mostly itâs just hard cards and have to process and do backrounds on them with help of gpt/google/books etc. I canât just focus revising nothing but old cards because of my study schedule.
1 Like
Maybe 20 seconds is a minority in Anki because Anki manual basically recommends 10 seconds or less for each card, so I think the average of typical Anki users is around 5-7 seconds. As far as I know the mainstream opinion is that if more than 10 seconds are needed there is too much info on the card. However as you say some learners may need more than 20+ seconds, e.g. in difficult cases such as law. So the best time actually depends on the contents of the study.
- As I said, a tier based system that seperates people by their ability is a good idea. Allow people to choose to which category they belong to (medical student, law student, language learner, casual learner, etc.). Add flairs.
Also further recommendations
-
Introduce teams competing with each other. Currently there are groups, but they donât mean anything. Allow teams to compete against each other like a sports league.
-
Introduce chat. So that people could communicate with each other. I donât know how viable this is for your server though.
2 Likes
I do not think such restrictions are necessary. We all use anki in different ways, and I think there are valid reasons why some people might review cards quickly. Coming from someone in delta/gamma leagues.
*
If restrictions are implemented, could you allow the following police car light to be optional?
Show
This seems reasonable
Force exclusion of reviews under 0.001 seconds
1 Like
I think the 0.001 second should be made 0.1 seconds. Itâs the fastest that a human can react to something, and I mean that in theory. In practice, you would take more, and if you are doing Anki certainly you should take at least 0.2 seconds. Regardless, 0.1 sounds good too.
For perspective, athletes trained to react fast to visual stimuli with predetermined set of movements still canât reach 0.1 seconds. And we are talking about Anki, so 0.2 is a bit more reasonable. Or are there people who take less than 100 ms to read out their answers?
2 Likes
Adding these functions will need a new database and UI, so development is planned but will take some time.
I received several requests for a chat function so Iâm looking into it. The issue for now is to get rid of annoying users. (e.g. trouble between users, banning offensive users, banning hate speech.) So I think it is necessary to develop functions to block users and ban annoying users as well as chat.
So far it seems that there are not so many requests for anti-cheat measures compared to the number of users, so there may be no need for them.
This is possible but there is not much space in the optional UI, so it has not been added yet.
Perhaps this threshold value is different for each user. In my opinion 1 second seems reasonable, 0.001 and 0.1 are almost the same so either one looks fine. If needed it might be a good idea to survey alpha and beta users. But so far there have been very few reports of problems from alpha and beta league users, so it may be of little importance, they may not care much about other users scores.
2 Likes
aim400kg. com/game/3
Yeah you can test with yourself. Out of 1k+ people top20 is around 0.17-0.19.
Mine is 0.22
Iâd say fastest in anki is around 0.3 maybe after processing info.
So 0.001 sec is too low. 0.1 sec I donât believe. Just compare reaction times to that site of top10 players who are skilled at it because they play fps games.
0.3 for quickest sounds about right even though thatâs insane for me too. (You would also need press space 2 times, because counter ticks after first space.
But I defienetly can get 0.4 on easiest words on my english vocabulary deck.
Same for med. You just see picture and can register it with one word.
2 Likes
Answer screen counting can be disabled in the deck option, so it is possible that some users with the option set may have about half the number of seconds as the average users.
It doesnât really affect your stats. Danika changed the tooltip to express that more clearly:
Whether to stop the on-screen timer when the answer is revealed. This doesnât affect statistics.
(emphasis mine)
2 Likes