Anki for CCNA: Spaced Repetition That Beats the Forgetting Curve

The Forgetting Curve, and Why Anki Beats It

Learn a fact once and your recall of it decays fast — the classic Ebbinghaus forgetting curve. Without reinforcement, a list of port numbers you memorized on Monday is mostly gone by the weekend. Cramming produces a sharp spike of knowledge that collapses just as sharply, which is why a night of frantic review before the exam feels productive and rarely holds.

Spaced repetition works with that curve instead of against it. Each time you successfully recall a fact, the curve flattens and the memory lasts longer, so the next review can be pushed further out — a day, then three days, then a week, then a month. Anki tracks this per card. It knows that you find EIGRP's default timers easy and the administrative distance of external EIGRP hard, and it schedules each one independently.

The scheduling is driven by a spaced-repetition algorithm (historically a variant of the SM-2 algorithm; recent Anki versions also offer the newer FSRS scheduler — check Anki's manual for the current default and how to enable it). Mechanically, you see the front of a card, try to recall the answer, reveal it, and then grade your own recall: Again, Hard, Good, or Easy. Miss a card and it comes back within minutes; nail an easy one and it disappears for weeks. The payoff is leverage — a few minutes a day maintains hundreds of facts, because Anki only ever shows you the ones you're about to lose.

What Belongs in Anki: the Rote Layer of CCNA

Anki is for atomic, factual recall — the things that are simply true and have to be at your fingertips. CCNA is full of them, and they cluster into a handful of categories worth building out deliberately.

Administrative distances: directly connected 0, static route 1, eBGP 20, internal EIGRP 90, OSPF 110, RIP 120, external EIGRP 170, iBGP 200. Well-known port numbers: FTP 20/21, SSH 22, Telnet 23, DNS 53, DHCP 67/68, TFTP 69, HTTP 80, NTP 123, SNMP 161/162, HTTPS 443, syslog 514. Subnet boundaries: the mask-to-block-size table, where /26 is 255.255.255.192 with a block of 64, /27 is .224/32, /28 is .240/16, /29 is .248/8, and /30 is .252/4 — the numbers you need instant, not derived, when you're subnetting under time pressure.

Protocol defaults are another rich vein: OSPF hello/dead timers of 10/40 seconds on broadcast and point-to-point links, its default reference bandwidth of 100 Mbps, and the cost formula (reference bandwidth divided by interface bandwidth); EIGRP hello/hold timers of 5/15 seconds on high-speed links and 60/180 on low-speed multipoint, with its metric based on bandwidth and delay by default; BGP's 60-second keepalive and 180-second hold timer. Command syntax fits here too — the exact keywords for things like a floating static route or an OSPF network statement — as long as each card tests one specific piece.

What does not belong in Anki is anything that requires reasoning across steps: troubleshooting a broken adjacency, deciding where a summary route should live, interpreting the output of a show command in context. Those are skills, and flashcards can't rehearse them. Trying to force multi-step logic into a card produces cards that are either too vague to grade honestly or so long you memorize the shape of the answer instead of the substance.

Write Your Own Cards, and Keep Them Atomic

Write your own cards. This is not busywork you're skipping by grabbing someone else's deck — the act of making a card is itself a study session. To create a good card you have to locate the fact, confirm it's correct, decide what the cue should be, and phrase the answer precisely. That encoding effort is a large part of why the fact sticks, and it's exactly the work a pre-made deck does for you and therefore robs from you.

Keep every card atomic: one fact, one question, one answer. A card whose front is "List all the administrative distances" is a bad card — you'll memorize the order of the list rather than the values, and grading yourself is a coin flip. Break it apart: "Administrative distance of OSPF?" → "110". "Administrative distance of eBGP?" → "20". Now each fact is scheduled on its own difficulty, and the ones you actually struggle with get the repetitions they need instead of hiding inside a list you half-remember.

For tables and syntax, cloze deletions are your friend — hide one value in a line of context, like a subnet mask blanked out of its /prefix, or a single keyword removed from a command. Phrase each front as a precise question with exactly one unambiguous answer, and resist stacking multiple facts on the back. If you find yourself writing a bulleted answer, that's a signal to split the card.

Review Daily, Vet Shared Decks, and Getting Anki

Do your reviews every day. The whole schedule assumes you show up when a card is due; skip a few days and the due pile compounds while the timing that makes spaced repetition efficient quietly breaks. Ten or fifteen minutes daily beats a two-hour session once a week, and the habit is the point — consistency is what actually flattens the curve.

Shared CCNA decks do exist, mostly on AnkiWeb, and they're tempting as a shortcut. Treat them with suspicion. Community decks vary wildly in quality and routinely contain errors, ambiguous phrasing, and facts that were true for an older exam blueprint. A wrong card you drill daily is worse than no card at all — you're paying to memorize a mistake. If you use one, verify every card against your official cert guide or current Cisco documentation before you trust it, at which point you've done most of the work of writing your own anyway.

On cost and platforms: Anki desktop is free and open-source on Windows, macOS, and Linux, and AnkiDroid on Android is free as well. AnkiWeb gives you free syncing across devices (and basic review in a browser). The one paid piece is AnkiMobile, the iOS/iPadOS app — a one-time purchase whose proceeds fund the project's development. Check the App Store or Anki's site for the current price. Practically, you can build cards on the desktop, sync through AnkiWeb, and knock out reviews on your phone during the small gaps in your day.

Anki Handles the Facts; Labs Build the Skills

Here's the hard boundary. Anki can make you recall "OSPF cost equals reference bandwidth over interface bandwidth" in under a second, and it can make eBGP's administrative distance of 20 as automatic as your own phone number. What it cannot do is teach you to diagnose why an OSPF adjacency is stuck in EXSTART, or to configure and verify a working topology under a ticking clock. Those are skills, and skills are built by doing, not by recalling.

That's the two-layer reality of passing CCNA and, more importantly, of being able to do the job afterward. The rote facts are necessary but nowhere near sufficient — the exam and the work both test application. So run two loops in parallel: Anki for the rote layer, and hands-on practice on real IOS for the skill layer. On this site that means graded Cisco Modeling Labs, where you build the config yourself and get told whether it actually works. It's the same 'beat the forgetting curve' idea pointed at two different targets — one keeps facts in memory, the other keeps skills in your hands.

The two loops feed each other. When a lab surfaces a fact you fumbled — you couldn't remember the timer, or you had to look up the port — make an Anki card for it on the spot. The fact sticks, and the next time you're in a lab you spend your attention on the reasoning that matters instead of burning it trying to recall a number. Facts in Anki, skills in the lab, and neither one waiting for the other to catch up.