How to Review a Particular Anki Deck During Exam

How Much to Learn with Anki

24 minute read

When you're starting out with Anki or another SRS, yous might wonder how much y'all can look to learn with this newly efficient method of studying. Considering spaced repetition follows a mathematical formula, information technology'due south possible to do a decent chore of estimating answers to these kinds of questions – certainly much ameliorate than you can manage for any other report method I know of. In this post, we'll answer three common questions posed by new (and experienced!) users of spaced repetition systems: how much you can learn and retain in a given amount of written report fourth dimension, how much time you'll need to learn a given amount of information, and how long it will have to learn and retain a single card.

Notation: If you lot use the estimates in this post, I would honey to hear from you later on you've been studying for a while and assemble more data about how well they work in practise for a variety of people! Please consider emailing me at contact@sorenbjornstad.com. I haven't made anything up in this postal service to my cognition, only more of information technology than I would like is estimates and anecdotes on top of estimates, and the bottom estimates oftentimes have just vague theoretical backing. Because results can vary greatly depending on what you study and how you lot study it, and spaced-repetition algorithms are based primarily on ascertainment, all of this stuff still involves a meaning corporeality of "guess and check" piece of work.

How much can I learn in 10 minutes a twenty-four hour period?

The near bones question you can ask is, given a certain corporeality of fourth dimension you're willing to spend doing flashcards every mean solar day, how much new material should you exist learning? The verbal value depends on many things, similar the difficulty of your flashcards and how consistently you review. Simply you tin can come up with a decent estimate; you only need two pieces of data:

1. How many minutes a 24-hour interval can you lot devote to reviewing? Remember, for all-time results, you want to review every day, for the rest of your life. Non the first two weeks when yous're excited about spaced repetition. Not four days a week. Not the days when you feel like information technology. You might want to take the number you call up is good right now and cut information technology in half! Of course, there'south nothing illegal about getting rid of some of your cards or not using spaced repetition anymore, if y'all later determine yous take other priorities. But the more consistent and continuous yous are, the better results y'all'll go, and the more than realistic you are with yourself now, the more consistent and continuous you lot'll be. With SRS, consistency beats volume in terms of results.
2. How many seconds does it accept you to review each card, on boilerplate? This usually comes to 5–x seconds, with meliorate cards yielding shorter times. If you have some review history behind you, yous might want to look and run across how long it has taken you. You lot can find this effigy underneath the Reviews graph in Anki's statistics if you tick the "Time" checkbox above the graph. I tend to sit merely nether 7 seconds. If this number exceeds 10 seconds, you definitely need to think near improving your cards.

A unproblematic formula

Now we can calculate how many cards we can review (R) in a day. Hither'south the formula, where \(t_\text{solar day}\) is in minutes and \(t_\text{card}\) is in seconds. (If algebra isn't your thing, in that location's a widget a scrap farther down the page that runs the numbers for yous.)

\[R = \frac{ 54 \cdot t_\text{24-hour interval} }{ t_\text{card} }\]

As an example, suppose we desire to study approximately 10 minutes per mean solar day and get through a card every 5 seconds. Nosotros get R = 108, so that's about the number of reviews nosotros can maintain in ten minutes a mean solar day.

I give 54 as the constant factor in the formula instead of 60, the actual number of seconds per minute. I add a x% penalty to your \(t_\text{mean solar day}\) because information technology's unlikely yous'll really be studying cards for 100% of the fourth dimension you've set aside to study: on a typical day, yous'll probably stop to fix one of your cards, or expect something up, or get make a cup of tea. If yous're bad at concentrating in the digital age – like near of us! – yous might desire to reduce the number a bit further.

Warning: If you want to check this gauge confronting how long you actually spend studying, exist aware that Anki won't always include your unproductive time in your statistics – if you aren't looking at a card, it's non counting time, and if you lot spend time looking at a card, become to another screen, and come back, the fourth dimension you spent earlier going elsewhere won't be counted. Similarly, if you undo a review (considering you chose the wrong rating) and option again, the time y'all spent earlier undoing won't be counted. That's why this time should be considered separately. If Anki says you studied for 15 minutes, information technology might have taken something similar 17 minutes of clock fourth dimension.

A normally bandied-about dominion of pollex is that your reviews will, over time, mountain to roughly 10 times your new cards per mean solar day. I've plant this to exist tolerably accurate. And so our figure of 108 means we can afford to add ten-11 new cards per day if we don't want to exceed 10 minutes of study.

Here's a little JavaScript widget that does the math described higher up for you:

Implementation details

If 10 new cards per twenty-four hour period sounds small, it is and information technology isn't. If we assume that you lot create two to three cards about every useful thing you learn, you're durably learning about 4 carefully selected things every day. (If you're not sure why you'd create more than than one card, be sure to read up on creating precise cards and the Minimum Information Principle.) You can become quite a lot of mileage out of that; afterward all, most people don't select whatever items to learn every day, they merely recollect whatever they happen to remember. To put it another fashion, later a year of learning 4 new things per day, you'll know about 1,500 new things of your choosing – not just any one,500 things, 1,500 things that you know will be useful in your daily life and that you won't forget. Of course, if you tin study for 20 minutes every day, you'll learn twice equally much, which is great… merely there isn't really a minimum amount of useful study. Consistency beats volume.

Now, but because yous can simply add ten new cards per day doesn't mean you literally have to click the "add" button and type in exactly 10 cards each twenty-four hours. You lot can add them in whenever you call up of them or have time to batch-add together some content; your SRS volition accept care of queueing them up and only introducing the number you lot ask for every day to go along your workload steady. (If yous end up with a big backlog of new cards, any good SRS will offer tools to reprioritize them.) To modify the number of cards introduced per solar day in Anki, visit your deck's options (gear icon in the deck listing), click the New Cards tab, and adjust the number of "New cards/day".

Anki doesn't introduce new cards if you lot miss a mean solar day of studying – for instance, if you miss 5 days of study and y'all've chosen to add ten new cards per twenty-four hours, on the twenty-four hour period y'all come back you will but get 10 new cards, not 60. If you're trying to get through a big number of cards in a certain amount of time and expect to miss a few days here and at that place, it'southward important to take this into account. Of course, you can always cull to add together extra new cards in, but Anki won't do that unless you tell it to.

Lastly, don't forget to consider the time it takes to create cards. This is ordinarily small compared to your review fourth dimension, but not insignificant – it'due south reasonable to imagine it might accept xv seconds to a minute to create each carte du jour, depending on what you're studying and how much experience you have. Yous'll also have to spend some fourth dimension editing and maintaining the cards later. If y'all written report x minutes per day, and y'all're creating all your own cards, you might terminate upwards spending more like 12 to 15 minutes a mean solar day in Anki.

Rule of thumb: 1 for one

You might notice that, in our example, nosotros said x new cards per solar day would require 10 minutes of daily review. With the simple formula in a higher place, information technology'south reasonable to extrapolate the approximate and assume that 20 new cards per solar day will take 20 minutes, and so on. In general, over the long term, 1 new card per 24-hour interval equals one minute of daily review.

Of form, this is a dominion of thumb on peak of a rule of pollex, so expect it to be wildly off on occasion. In detail, the formula/widget volition give a noticeably improve estimate if your cards take longer than 5 seconds to report or your productivity value is much off of 90%. Nevertheless, this rule is wonderfully like shooting fish in a barrel and not totally imprecise for an off-the-gage approximate.

How much work will information technology be to fix for an examination?

That covers how many cards you lot should study if yous have a specific amount of time to spend. Just in some circumstances, you may want to know the opposite: given a sure corporeality of fabric you have to chief before a certain time, how much new cloth should you exist introducing and how much time is it going to accept to learn it? The following calculator does its best to respond those questions without any knowledge of your material.

The formula hither is merely a combination of the one higher up and some obvious dimensional analysis (eastward.g., the number of cards you lot have to learn is equal to the number of facts you accept to learn times the number of cards per fact), so I won't go through it in detail.

Important notes about the pregnant of the variables hither:

• I separated "cards per fact" and "facts to learn" for convenience if you don't have all the cards created alee of time. Perhaps you have a question banking company of i,500 facts you want to learn, and in the initial fix of cards yous've created, you averaged 2.5 Anki cards per question. (If you're not sure why you lot might need to create multiple cards per question, be sure to read up on creating precise cards and the Minimum Data Principle. If you don't do that, you'll likely be spending more time studying than estimated hither.) If you already have the cards created, simply leave this at 1 and fill the number of cards in for "facts to learn."
• Reducing the number of "study days per week" from 7 will mean no new cards are introduced on the days you exercise non study, and the widget will business relationship for that. Your study time will be much higher the mean solar day subsequently you skip one or more days, since some cards volition be overdue; the "converged study fourth dimension/day" field does not average this in, it continues to listing the estimated amount of time you will spend reviewing on those days when y'all don't have any overdue cards. In my experience, you tin expect to spend a wee bit less than twice the time the day after you skip.
• The word "converged" in "converged reviews/day" and "converged study time/mean solar day" reflects the fact that it will accept some time after y'all begin studying – probably several weeks – to reach this level of written report fourth dimension. In the days prior, it will be lower. You can attempt to flatten this out by introducing more than cards than "new cards per mean solar day" at the showtime, merely be careful nigh overshooting: the due reviews can mount up fast! Y'all absolutely don't want to add as many new cards as information technology takes to fill out a normal written report session, or at least not for long; if you add together 150 for only three days, suddenly you'll accept 450 cards due for review within a 5-day period or so.
• The new cards per day figure is exact (well, information technology'south rounded to two decimal places) and a matter of bones arithmetic; if you lot study at least this number of new cards per 24-hour interval, equally long every bit your other figures are correct, specially the ones nearly how many cards you cease up creating, you are guaranteed to get through all of the textile before the exam. The converged reviews/day and converged study time/mean solar day are estimates based on the model explained earlier in this postal service, resting on the 10x rule of pollex.
• The computer does not have into account the time you'll be spending initially reading and understanding the fabric (if y'all haven't done then already), creating cards (if you lot're creating your own), and editing cards (which you'll want to do regardless).
• The reckoner will take you right up until the test, then y'all'll still be learning new material the day before. It's advisable to add an additional calendar week or so, in instance you study fewer days than yous expect, and so you lot have a fiddling bit of time to consolidate your knowledge and mayhap take a practice test or 2.
• On that note, and nigh importantly, since this is an estimate, it'south wise to build in a margin of error and periodically review your progress so you can adjust if it proves to be low, specially if you lot haven't created any cards yet and don't necessarily know how many will exist needed to main the material. Even if you've washed this earlier, every topic is unlike. I am not responsible if you employ this calculator and aren't ready for your exam!

How long will it take to learn one carte du jour?

A final useful thing to quantify is how long it takes to learn one card and maintain it for the residual of your lifetime. This might help you decide whether information technology's worth learning a detail piece of information, for example. The figures should be the same for someone of just about any age, every bit near of the effort happens early on. If yous're at or above retirement age, you might be able to decrease 10%.

tl;dr: 2-5 minutes, more likely on the low side, depending on how optimistic, determined to learn everything, and careful at creating flashcards you are.

Cards vary wildly in how much review time they require. Leeches might have 20 minutes of your time and still get out yous ignorant; the easiest cards with straight like shooting fish in a barrel ratings might take 30 seconds over your unabridged lifetime. (On the default scheduling settings, a card with direct like shooting fish in a barrel ratings progresses through the sequence 4 days – 13 days – 1.5 months – 5.4 months – 1.7 years – half-dozen.ix years – 29 years – 100 years; at 5 seconds per review, that's forty seconds total. The last interval would be 128 years, merely Anki caps intervals at 100 years by default.)

However, people come with wildly different boilerplate estimates as well; for example:

• SuperMemo'southward theory folio suggests you can learn 200–300 items/year/infinitesimal (i.e., if you expend one minute studying every day for a year, yous'll learn 200-300 items, with their lifetime review costs included). If you run a couple of conversions on that, information technology comes to ane.2–1.ix minutes per item.
• Gwern comes to a similar decision of one.8 minutes, explaining how this figure can exist derived using a more circuitous formula on the aforementioned SuperMemo page, but then decides to more than double the number to 5 minutes, thinking SuperMemo'southward model may evidence too optimistic in real-globe use.
• Michael Nielsen finds Gwern'south estimate too optimistic in plow and opts for 10 minutes, using an advert-hoc but sensible-feeling reasoning process and his short-term review history.

That's a total range of 1.ii–10 minutes, spanning an entire order of magnitude! Who do we believe?

I've been using Anki for 10 years, albeit sometimes on and off. Let's have a look at how these estimates accept held upward in practice for me. I dug into my collection and took a look at those cards that have an interval of greater than i yr (northward = 18541, mean interval = 6.01 years). Cards that accept reached 1 twelvemonth have generally accumulated a majority of their lifetime review cost already because intervals rapidly ascent exponentially beyond a human being lifetime from that point. Hither's a quick summary of the current total review times on these cards. Times are in minutes:

• Hateful: 0.944
• Median: 0.561
• Q1–Q3: 0.30–1.14

You might notice just from the difference betwixt the mean and median that the data are strongly correct-skewed – that is, the extreme values are all on the top finish. If you're into statistics, here's a nice box plot showing that skew:

Hither'due south some other neat way to await at it. This shows the current interval of each carte versus how much full time I've spent reviewing it, merely colored by the ease (the ease values are shown in per mille, or parts per m – 10 times the normal percentage value):

We tin can see that both the total review time and the interval cluster towards lower values, and also that the more difficult cards tend to sit down at lower intervals and higher review times. The tail to the upper-left visually demonstrates the benefit of trashing the most difficult material in your collection – and this graph is afterwards I've already trashed the very worst from mine; an untrimmed collection would be even worse!

At any rate, we can see that the bulk of my cards have accumulated less than one minute of total review time after attaining an interval of at least a twelvemonth. As mentioned earlier, the menses of i year is meaning because at this point, unless you forget the card, but a handful of repetitions remain in your lifetime, less than half of the full repetitions for the card.

Just how many cards do we forget? If we forget one, nosotros have to start the progression over from the commencement, or at least close to it. Here's the summary for lapses:

• Mean: 0.82
• Median: 0.00
• Q1–Q3: 0.00–1.00

In other words, fully half of my cards have never lapsed at all, even with interval of over a yr. 75% of them take lapsed once or never, and even taking into account the outliers (some of which have absurd values like sixteen), the average carte lapses less than once. I run into no reason to believe that this number will increase much further for the oldest cards as time goes on; others have been doing spaced repetition for long plenty that if the spacing consequence broke down at that point, nosotros would know about it by now. Farther, assuming the spacing effect does hold out over the lifetime of a card, chances are that most lapses occur in the earlier stages, considering most reviews practice; that ways fewer cards are probable to be forgotten past an interval of 1 year, where the near rework is required.

(I haven't taken a rigorous await at information technology, only I remember this is how Nielsen gets then far afield from the other estimates: he uses the average calendar time between lapses per carte du jour to quantify how often cards will lapse and have to be relearned from the first. This doesn't aroma correct to me, simply because in that location are far fewer reviews to lapse on when the intervals become high. Since the algorithm aims to create an equal take a chance of forgetting the card for every review, if you imagine 1 year is a fleck more than halfway through the review schedule with 7 reviews earlier and 5 afterwards, by the time you striking ane yr, your chance of ever lapsing again is already less than over the previous 7 reviews, even though there may exist 30 times the corporeality of calendar time to exercise it in. This would seem to consequence in an overestimate of the average lifetime lapses per bill of fare. Nielsen also says he rarely uses the "hard" or "like shooting fish in a barrel" ratings in Anki, which could reduce his efficiency somewhat, but I have no data on how much efficiency this leaves on the table.)

Seeing this information, I feel pretty comfortable saying my average lifetime cost is going to be somewhere in the neighborhood of the value predicted by SuperMemo. 2 minutes seems similar an entirely reasonable estimate. 3 minutes is a squeamish upper leap even if I missed a few things in this analysis. 5 minutes is bourgeois and an easy round number, and so it's non a bad rule of thumb when considering if it's worth learning some item fact, as Gwern uses information technology – but information technology'south loftier plenty that it might atomic number 82 you astray if you're trying to make up one's mind whether information technology's worth learning, say, 20,000 cards.

Now, as alluded to earlier, I do practice the removal of highly hard material in my collection, though it probably amounts to one or 2 pct, null like the 10 percent SuperMemo theory considers. Reviews for cards that are no longer in my collection were not considered in my dataset. That means, outset of all, that my average review time is somewhat greater than what I've quoted, since I've spent time reviewing harder-than-average cards I later deleted. I think there's notwithstanding enough margin between my numbers and the SuperMemo numbers that this won't touch my assay. Secondly, information technology means that if y'all insist on keeping all material, even the virtually hard and to the lowest degree important, you may come up noticeably worse than me (remember that tail in the upper-left of the graph!).

Similarly, if you don't write practiced flashcards, more of your cards could go in the management of 5 minutes. But it's worth noting I haven't been religiously post-obit my Rules for Designing Precise Anki Cards for all of the past ten years. I've created my fair share of mediocre cards, and many are in there however. So my performance isn't hopelessly skewed abroad from what you could promise to get because I'm a Spaced Repetition Expert.™

I also certainly have some cards that are very easy in my collection, which would perhaps lower the boilerplate lapses and review time a bit, merely this is a good thing that you should do too: piece of cake cards cost inappreciably whatsoever fourth dimension to review, while the price of forgetting something "easy" is high.

A more technical concern: I have a contempo long hiatus in those 10 years of statistics, for several years after I got out of higher, and I haven't yet defenseless up on all of the overdue reviews, which means information technology's reasonable to suppose I might exist somewhat underestimating the review time since I haven't yet told Anki which of those cards I've forgotten and started the procedure of relearning them, costing additional time. This said, I took a quick wait at the cards I have caught up on and those I haven't and plant that the cards I haven't show upwardly consistently easier on every measure than those I take, including measures which definitely wouldn't be affected by being backside, like the average review fourth dimension per menu. Likewise, even with reviews delayed by 3 years in some places, I'm still recovering l-75% of the memories, depending on topic, which is much less than the ninety% promised by on-fourth dimension reviews but not bad at all given the circumstances. That means the cards I recollect go an actress boost because they were manifestly easier than their statistics would have suggested (else Anki would have scheduled them further in the future in the first place), which helps to counter the extra review time for the ones I've forgotten. So, overall, I remember this event is going to be weaker than I would accept guessed. If my results change significantly over the next couple of years, I'll exist sure to make an update.

Scheduling notation: Most halfway through my spaced-repetition career, I switched Anki's scheduling algorithm to reduce lapsed cards to 10% of their former interval rather than the default of 0%. I oasis't done whatever rigorous research into the effects of this, but my impression is that it reduces the pain of failing a card without making it substantially more probable to be forgotten over again immediately. Damien Elmes, the main programmer of the Anki algorithm, has said in the past that the only reason he put the default at 0% is that it's less confusing for new users that way. I include this bit simply considering there'due south a chance it could brand my learning slightly more than (or less!) efficient than others'.

Using your own data

If you lot have a bit of vanquish and data analysis know-how and want to play with this data from your own collection, here'south how to get the data out. This assumes a Linux-like or MacOS system, but you could adapt information technology for Windows besides.

Place the following SQL query in a file chosen query.sql:

                              SELECT                Cast                (                SUM                (                revlog                .                fourth dimension                )                AS                Float                )                /                60000                every bit                t                ,                notes                .                id                ,                notes                .                tags                ,                decks                .                proper noun                ,                cards                .                reps                ,                cards                .                lapses                ,                cards                .                gene                ,                cards                .                ivl                FROM                revlog                INNER                JOIN                cards                ON                cards                .                id                =                revlog                .                cid                INNER                JOIN                notes                ON                notes                .                id                =                cards                .                nid                INNER                JOIN                decks                ON                cards                .                did                =                decks                .                id                WHERE                cards                .                ivl                >                365                -- 1 year                Grouping                BY                cid                ORDER                BY                t                DESC                ;                          

Then make sure Anki is airtight (or your collection will be locked and unavailable for querying) and use the following beat out ane-liner to create a pipe-separated values file:

              sqlite3                "~/.local/share/Anki2/Soren Bjornstad/drove.anki2"                < query.sql |                awk                -F                '|'                'Begin { OFS = FS; impress "ReviewTime|NoteId|Tags|Deck|Reps|Lapses|Ease|Interval"; } { gsub(/^_/, "::", $4); print }'                >                records.psv                          

For the ^_ after gsub, that is a literal "unit separator" ASCII graphic symbol – to type it, in your final printing Ctrl+V, then Ctrl+underscore.

You'll need to have SQLite installed, and you'll want to replace Soren Bjornstad with the name of your Anki profile (if yous don't know what a profile is, yours is probably called User 1). If you're on a Mac, y'all'll have to change the ~/.local/share/Anki2 bit too – encounter file locations in the manual. Once you get the i-liner to run, just import the records.psv file into your favorite analysis software or spreadsheet.

christisonthaparme.blogspot.com

Source: https://controlaltbackspace.org/memory/how-much-to-learn-with-anki/

Share this post