TL;DR aka Synopsis. You can read the bullets below and ignore the rest.

1. Evolution: "A Frequency Dictionary of German" (a book published in 2006)-> original deck -> picture deck -> "Fluent Forever-ized" (this deck)
2. Biggest additions: 1) English translations of sentences (from Google translate) 2) German sentence audio (Google robot voice).
3. The Objections noted in comments in the parent decks, persist. Some pictures are sexist and many pronunciations were denounced as non-native.
4. Card types: words, pictures, fill-in-the-blank, word form, base word form, word order, articles, sentences, spelling
5. Follows minimum information principle
6. There are most-likely some errors in this deck.
7. Play with the templates to create the cards you want/need.
8. If you downloaded this deck before March 9th 2025 read the update below.

FAQs

Why are there so many cards?

Cards follow the minimum information principle I was using the predecessor deck and I found I could spend up to a minute on each card, which slowed me down and confused me sometimes. "What am I learning at this moment, the sentence, the word, the article, the word form, everything at once or a combination of these?" Now you know exactly what you're learning with each.

Why don't the cards show all the fields?

I initially attempted to make the cards as minimal as possible so they wouldn't be cluttered, but after working with it I added back on few fields based on my needs. Please modify the templates as you see fit (rather than down-rating this deck). E.g. If you'd like to see the picture somewhere it's not found, it's as easy as adding '{{picture}}
' to the card template.

Why is the English mostly only found in hint fields?

Because one of the recommendations found in Fluent Forever is to not translate. Through the use of pictures you can associate pictures with meanings of words without using English as an intermediary. Unfortunately, with some pictures not being obvious I found the English was necessary more often than I'd hoped and I made it visible in some cards by default. If you want the English to show up without having to tap every time, look for things like '{{hint:English}}' in the templates and replace it with '{{English}}'. Do the opposite where the English isn't hidden, if you'd like no English.

Why didn't you include prompts in the cards?

I did initially include prompts like "What's the base form of this word?", "What's a sentence that includes this word?", but I eliminated them in favor of badges at the top of the cards that indicate card type to make review quicker. After breaking down the cards further by sub-decks, the prompts really don't seem necessary but the badges can be nice, if you're working on multiple sub-decks within the same sitting.

What if some of the pictures don't accurately represent the words?

Use the English initially (which does require clicking the hint). Then, I HIGHLY RECOMMEND, at this point, that you create a story in your mind of how this picture represents the concept, if you can tie it to the example sentence, all the better.
For instance, an early note you'll encounter is 'Ein', it has a picture of the unicorn, ask yourself "how many horns does a unicorn have?".
Later you'll find "im Grunde genommen", with the sentence "Es ist im Grunde genommen ganz einfach." meaning roughly "It's actually/basically quite simple." The picture is of a Giraffe with it's tongue retracting on itself. You can tell yourself "Without hands, there's no way a Giraffe could clean it's face." and imagine a response coming "It's actually quite simple." If you tell yourself this story the subsequent times you see the picture you won't need to see the hint.

Where did '4250' come from?

The original book had 4036 words, the parent decks had 4214 notes (because notes were made for the phrases found in the book as well) this deck has 4250 notes, because I split the two sentence notes into two notes and this resulted in 4248 notes, so I added two more to make a nice round number.

Why did you organize the deck into sub-decks?

I figured this way, people of different levels could still make use of this deck. It's easy to just study the whole deck, but if you want to skip some initial decks or master one collection at a time, you can. The original cards were sorted into 'levels' with each level (except the last) having around 25 cards. I kept the levels the same, though now a few of these levels have more than 25 cards each. The decision to make 10 sub-decks was arbitrary, but 425 notes seemed like "a lot" to me on it's own, so I felt this made the overall deck more manageable. Finally in an update, I further divided the 425 sub-decks into sub-deck types, so people can chose which to review or skip some altogether. The proper way to make sub-decks by type is to use many card templates and then use filtering. I chose to not use the anki standard, since I could not share these filters, and I've been sharing this deck with many Anki novices (who wouldn't be able to make their own filters).

Do the example sentences rely on words only in this deck?

No, some of the example sentences have words that are not found in the deck.
From that group of words, these are the 17 words (followed by their translations) that occur 3 or more (up to 7) times in the example sentences.

• dritten (third), Bewerbung (application), Einbrecher (burglar), Opa (grandpa), Theaterstück (theater play), Studiengebühren (tuition fees), Spenden (donate), zweite (second), Erdbeben (earthquake), kaputt (broken/kaput), Hausaufgaben (homework), Verhältnisse (circumstances/situation), Angeklagte (defendant/accused), DDR (GDR), Obst (fruit), abgeschlossen (completed), Päckchen (package/parcel)

What's the difference between the 'German Sentence' and 'Sentence without Blanks' fields on the card?

'German Sentence' is the complete sentence while 'Sentence without Blanks' doesn't have the German word in it. The latter is used for the word order cards, i.e. 'where does this word go in this sentence'?

Update 2024-Nov-14

Following experience working with the cards, I decided to rearrange them to make sub-decks for each card template.
There are many upsides to this re-arrangement, but also a couple downsides. Downsides: 1. you can no longer just suspend one note for the words you already know, 2. there are 4250*9 (38,250) notes, rather than just 4250.
Upsides: 1. you have finer control and can simply ignore whole sub-decks. For instance, if you like to learn words through pictures just do those, if you're here for the example sentences do those, if you never really got the articles down (die, der, das), do those. 2. you can progress through multiple sub-decks at the same time and since they're in the same order they can reinforce each other, meaning you may be more confident clicking "good" instead of "hard" as you're learning the words, speeding up (perhaps ironically) your progression through the decks. 3. I think the sentence cards (as in "think of an example sentence for this word") are the hardest. So you can simply wait until you've mastered all the other sub-decks before doing this. Note, I had to name the sub-decks with numbers to list them in priority order

I also found the picture and word templates to be a bit too spartan so I modified them. They now have the example sentences (in German and/or English) on them (in some cases as hints).

Update 2025-March-17

First please note, that if you update to this deck you will lose all your progress (except for in the "fill in the blank" subdecks). This is due to a misunderstanding/oversight on my part regarding card unique id's when I broke the original deck further down into subdecks in November 2024. If you downloaded before that and didn't update, you can probably download this without losing your progress (but please back up first, just in case).

• I changed the templates a bit (they were too minimal for me and I was basically relying on the English and didn't want them all to be hints all the time).
• I manually edited some of the English translations (to improve them), and added a "Form Varies" field so that on the word form sub-decks there are only cards for those in which the word in the sentence varies from the German word.
• I added a "type" field, purely for unique id generation purposes, this will ensure future updates can be consumed by all without losing their progress.
• I added a new card type "Sentence Audio" so you can test your ability to interpret the sentence audio.

Experience doing the deck (March 2025)

I'm over 1275 cards into this deck and I thought sharing my experience may be helpful to others. I also devoted myself to getting through one of the sub-decks in about a month, in order to share my experiences.

• My biggest learning: You must do anki daily. I used to skip the weekends and the difference in ease/retention is enormous!

• To do a subdeck in 4 weeks you have to do ~15 cards a day.

• These notes track my experience with the third subset which I did in 24 days (I had 30 new cards for 4 days, then 20 for 4 days, then settled on 15 new per day).

• Time spent reviewing cards -> per day -> per card

  • First 24 days seeing new cards -> 45 mins -> 21 secs
  • Week 1 without new cards -> 32 mins -> 20 secs
  • Week 2 without new cards -> 18 mins -> 15 secs
  • Week 3 without new cards -> 13 mins -> 13 secs
  • Week 4 without new cards -> 9 mins -> 12 secs

• I'm learning exclusively with the "fill in the blank" (FITB) decks.

  • You have to already be around A-2 level for this to work with you or at least be on the second 425 words.
  • I'm using the "word" and "picture" decks as auxiliary tools to pull the info out of my brain in different contexts
  • I zoom through most of these at around 3-4 seconds per card, hitting "easy", where I get stuck is where I need work.

• Whenever you can, speak out loud, the whole sentence or just the word.

• I'm ensuring I understand every word in the FITB example sentences even if I have to use google translate for some of the words.

  • The time per card when I first see it can be >1 minute, this is the main reason, the time/card drops when they're no longer new.

• I mostly click "hard", I'm in my early 40's and my 20 year-old self would've done this in half the time or less.

  • I'm way faster in the morning and dog-slow just before bed.

• "Anticipation" - It's worthwhile trying to guess if you're seeing a new card that's related to other words you already know, e.g. the noun to an adjective or verb, e.g. Behandlung/behandeln.

• The pictures the previous author chose make a lot more sense in the context of the sentences (remember some will be offensive to some viewers).

• You will not recognize every word you learn "in the wild", some you will inevitably only learn within the context of the card.

  • It's probably the standard 80/20 rule, 20% of the words you'll have to re-learn in a real-world context. I don't think this negates this practice.

• Always break down the compound words and understand the bases, it becomes easier to tell yourself a story about things.

  • For example: Independent is unabhängig, which means “un”-dependent (abhängig). While hängig on its own means “pending”, it’s similar to hängen which means to “hang”, so you can tell yourself the story: “it’s independent, it’s not ‘hung-up’ on anything”.
  • Eindruck is “impression”, by the time you see this you already know drucken means “to press (a button)” and “Ein” is like “in”. You can picture a ball of clay that you’re “in-pressing” with your thumb which leaves an “impression”.
  • By the time you see Miteinander, which translates to “each other”, you’ll already know it literally translates to “with one another” (mit-ein-ander).

What I'll do moving forward:

• I've been avoiding studying most grammar. When I start studying grammar in earnest, I'll do the "word form", "base word" and "word order" decks.
• I think I'll skip the spelling decks (because I'm strong in that realm)
• I may never do the sentence decks or do those when I'm conversant in the language and click "easy" or "hard" for any example sentence with the word in it.
• I aim to spend less time on Anki and more time doing other German learning things, e.g. reading, watching, talking, working slowly through a grammar book.
  • I've already started the fourth deck at a rate of 5/new per day, this will take me 85 days to get through everything once. I think this is a sustainable rate for me and it'll mean I'll learn (or at least see) 1700 new words in a year.

Read below only if you care about the process or how the errors came to be.

This Deck may contain errors, some of which may be easier for you to explain to yourself if you understand the process I went through to create the cards.

Description of Process:

1. Exported the predecessor deck as text.
2. Wrote python code to import and export as CSV (Excel-like) file.
3. Used code/google translate to translate example sentences into English.
4. Used code/google translate to Convert sentences into audio (using the robot voice).
5. Used code to break up the cards with multiple sentences.
6. Manually checked the new card pairs created from single cards containing multiple sentences.
7. Used code/heuristics to find the words in the example sentences and replace with double-underscore characters (more below).
8. Manually replaced words in sentences with double-underscores for around 168 notes that I couldn't do automatically.
9. Used code to set a field to true or false if the word(s) in the sentences were different from the word(s) associated with the note.

Description of Heuristics used to create "Fill in the blank" sentence cards.

Finding and replacing the words in the sentences was a time-consuming process in which I tried one heuristic then manually checked for correctness before moving on to the next heuristic, manually checking that and repeating. Finally, I manually edited 168 cards that I couldn't match with heuristics (I mean I could have but, you know, diminishing returns...).
Note, I did not manually check exact matches for obvious reasons.

Before running each of the heuristics, I found words and alternates for each note. If it was a single word or a word with just an article in front, I'd return the word, if word had an ellipse, like "je...desto" I'd remove the ellipse and return the words as a set ("je","desto"). If the word had parenthesis, like "das heißt (d.h.)", I'd return the first part as the word and the second part as an alternate without the parenthesis, e.g. either "das heißt" or "d.h.". If the word contained a comma, I'd return the first part as a word and any subsequent words after commas as alternates, e.g. from "selbst, selber" I'd return "selbst" or "selber". When replacing words, I'd replace all of the "words" or one of the "alternates". E.g. If the words was "je...desto" I'd find "je" and then "desto" in the sentence and replace each with "__", but with "selbst, selber", I'd either replace "selbst" or "selber". I ran the heuristics in order, so later heuristics were only run on Notes that didn't match an earlier heuristic. Heuristics in order:

1. Exact matches or "alternates".
2. single words that matched any word in the sentence with a Levenshtein distance of one.
3. if there's a value in the "Plural and inflected forms" field parse the values by comma and match as alternates, e.g. from "hat; hatte; hat gehabt" find either "hat", "hatte" or "hat gehabt".
4. For just the words (not the alternates), add 'er', 'es' and 'en' as a suffix and see if a match is found in the sentence. - Matched 172
5. Break trennbar verbs (those prefixed with 'auf', 'an', 'aus', 'ein', or 'fern') into two words and match both in sentence - Matched 15
6. Split the "Plural and inflected forms" field by comma and run trennbar heuristic on these. - Matched 0
7. Run a match on single word alternates with a Levenshtein distance of one. - Matched 19
8. Do a match with a Levenshtein distance of 1 on "windows" in the sentence. I.e. Traverse each sentence, character by character, using the length of the word or alternate as the window of characters to compare against and replace anything that matched with a distance of no more than one. - Matched 28