Eye of Ender in Minecraft: Craft This Many Before You Hunt the Stronghold (Each Throw Risks a Real 20% Break)

The first time I went stronghold-hunting, I hauled 30 Eyes of Ender into a two-hour trek and only ended up needing 13. Every guide I’d read said “craft extra, just in case,” without ever explaining what “extra” means in actual numbers. That vagueness is exactly why players either burn a whole Nether trip over-preparing, or run dry three throws from the stronghold and have to turn back for more Blaze Rods.

The Eye of Ender has a fixed, documented 20% chance to shatter on every throw — not a vibe, a real probability confirmed on Minecraft’s own wiki [1]. Once you know that number, how many eyes to craft stops being a guess and becomes basic math. This guide walks through that math, the triangulation steps that make each throw count, and the mechanic almost no other guide mentions: a surviving eye isn’t gone, it’s sitting on the ground waiting to be picked back up.

Quick Start: What You Need Before You Throw a Single Eye

Before you head to the stronghold, get this sorted first — skipping any of these is the single biggest reason players run out of eyes halfway through a hunt.

• Collect at least 12 Ender Pearls (Endermen drops) and 12 Blaze Powder (crafted from Blaze Rods, which only drop in Nether fortresses)
• Craft your Eyes of Ender in a crafting table — 1 Ender Pearl + 1 Blaze Powder makes exactly 1 eye, no bulk discount
• Decide your target count using the player-type table below before you start — don’t just wing a round number
• Enable coordinates (F3 on Java, the debug HUD on Bedrock) so you can log throw positions
• Pick two starting spots roughly a few hundred blocks apart, ideally at a sharp angle rather than a straight line, before your first throw
• Bring a stack of blocks and a pickaxe — once you’re triangulated, you’re digging straight down, not sideways

If you haven’t made it to the Nether yet to farm Blaze Rods, our Minecraft Beginner’s Guide covers the early progression that gets you there.

Gathering the Ingredients Without Extra Trips

Endermen spawn in the Overworld at night or in dark areas, and far more densely across every biome of the Nether — hunting them on the same trip where you’re farming Blaze Rods is faster than a separate Overworld pearl-hunting session. There’s no farm-free shortcut in vanilla survival either; piglin bartering and trading don’t produce pearls, so direct kills are the only source.

Blaze Rods only drop from Blazes, which spawn from a fixed spawner inside each Nether fortress. Blocking off the spawner room’s entrance so Blazes funnel into a single choke point is faster than chasing random spawns across the fortress. Two Blaze Rods craft into four Blaze Powder, so one fortress trip yielding even 10-15 rods comfortably covers the 24-eye completionist target with powder left over for Potions of Strength or Fire Resistance later.

What an Eye of Ender Actually Does

An Eye of Ender is one Ender Pearl combined with one Blaze Powder on a crafting table [1]. That combination isn’t arbitrary — it’s the game forcing a specific progression path. Blaze Powder only comes from Nether fortresses, so you can’t reach the End without a Nether trip first, and the pearl is the same item Endermen use to teleport, which is why the finished eye can sense a fixed point in space (the stronghold) instead of just a random direction.

Once thrown, the eye flies roughly 12 blocks in a straight line toward the nearest stronghold, trailing purple particles, then either drops as a pickup or shatters mid-air [1]. You get the direction reading either way — breaking doesn’t erase the information you just saw, it only costs you the physical item.

The Real Break-Chance Math (Not Just “Craft Extra”)

Every guide says “craft extra eyes.” Almost none show why a specific number is enough. The actual mechanic: each throw has a fixed, independent 20% chance to shatter and an 80% chance to survive [1]. That’s not a per-session average, it’s the odds on every single throw, reset each time.

Run the math forward and the picture gets clearer. Across N throws, your expected number of permanently lost eyes is simply N × 0.2:

The number that matters most here is the one nobody mentions: an eye that survives its throw doesn’t vanish — it falls to the ground as a collectible item you can walk over and pick back up [1]. If you’re already walking in the direction you just threw, recovering it costs you nothing extra. Treat every throw as a guaranteed loss and you’ll over-craft by double; walk the line and pick up what survives, and your real attrition rate is closer to that 20%-per-throw table than to “one eye per throw, period.” On that first 30-eye hunt, I didn’t know this and never doubled back for the ones that landed instead of shattering. Once I started walking through each landing spot on the way to the next throw location, my actual eye consumption for the same number of throws dropped by roughly a third.

How Many Should You Actually Craft? (By Player Type)

The right number depends on how many throws you’ll realistically make and whether you bother recovering survivors — not on a flat rule of thumb.

Triangulating the Stronghold: Step-by-Step

One throw only gives you a direction, never a distance. You need two throws from two different spots to get a real fix — the two lines only cross at one point [3].

• If you can’t clearly see the eye’s arc (fog, night, rain) — don’t guess and walk anyway. Re-throw. Each throw only costs a small, known percentage; a wrong 300-block walk costs far more.
• If the eye flies upward and disappears over the horizon — you’re still more than 12 blocks from the stronghold. Walk 300-400 blocks in that direction, ideally shifting about 45° off your first line rather than walking dead straight, then throw again [3].
• If the eye flies downward instead of up — stop walking. That means you’re already within 12 blocks of the stronghold horizontally [1]. Throw one more eye to confirm, and if it drops straight down at your feet instead of flying off, you’re standing on top of it [3].
• If your two triangulation lines cross at a shallow, near-parallel angle — don’t trust the fix. Two readings taken close together at similar angles produce lines that are extremely sensitive to small errors; move further apart and re-throw before committing to a dig site.

Expect roughly 50-100 blocks of error on a first pass if the stronghold turns out to be 500+ blocks away — that’s normal, not a sign you did something wrong. Take a second, closer reading about 50 blocks from your estimated point to tighten it up before you start digging [3].

Here’s what that looks like with real numbers. Say your first throw at (0, 0) reads a bearing of 300°, and after walking about 400 blocks northwest you throw again at (-280, -290) and read a bearing of 260°. Feed both readings into a triangulation calculator and you get an estimated crossing point — walk to it, throw one more eye about 50 blocks out to confirm, and adjust before committing to a dig [3]. The two-throw minimum is a hard floor, not a suggestion: a single reading only ever gives you a direction, never a location, so there’s no shortcut that finds a stronghold off one throw alone.

Filling the Portal Frame: The Part With Zero Risk

Once you’re inside the stronghold and find the portal room, the risk calculation changes completely. Placing an eye directly into an empty frame socket by using it on the block isn’t a throw — there’s no shatter chance at all [1]. Every eye you’re still holding at this point goes in safely.

The frame has exactly 12 sockets, and each one independently had a 10% chance of already containing an eye when the stronghold generated [1]. On average that’s about 1.2 sockets pre-filled, though it’s pure luck how many you’ll actually find — a portal that generates fully lit and ready to use is a 1-in-a-trillion occurrence [1], so budget for needing close to the full 12 rather than counting on a head start.

Common Mistakes That Waste Eyes (and Time)

• Treating every throw as a permanent loss. Walking through where a surviving eye landed costs nothing if you’re heading that way regardless — skipping this doubles your real attrition rate for no reason.
• Throwing both triangulation readings too close together. A shallow crossing angle turns a small yaw-reading error into a stronghold estimate that’s off by hundreds of blocks.
• Digging straight down before a confirmation throw. The “flies downward” signal means you’re within 12 blocks horizontally, not standing exactly above the portal room — one more throw at the exact spot tells you whether to dig here or shift slightly.
• Assuming a fan-quoted stronghold distance without checking the edition. You’ll sometimes see figures like “600-1,200 blocks from spawn” repeated across guides — that number doesn’t match Mojang’s own generation data for either edition (see below), so don’t plan your travel time around it.

Java vs. Bedrock: What Actually Changes

Verified against Minecraft Wiki data for Java Edition 1.21 and Bedrock Edition 1.21 — the crafting recipe, 20% break chance, and 12-socket portal frame are unchanged since their original introduction and are identical across both editions.

What does change is where strongholds generate. Java Edition has a fixed total of 128 strongholds spread across 8 concentric rings; the first ring holds 3 strongholds between 1,280 and 2,816 blocks from world origin, so your very first eye throw in a fresh Java world is guaranteed to be a long walk [2]. Bedrock Edition instead generates strongholds without a hard cap, with a minimum distance of just 160 blocks from origin (three are guaranteed at 453+ blocks) [2] — meaning a Bedrock world can hand you a stronghold dramatically closer to spawn than any Java world ever will on the first ring.

Practically, that means a Java player should budget for a 1,000+ block round trip on a fresh world — an elytra, boat river route, or a Nether highway (1 Overworld block equals 8 Nether blocks) all cut that travel time down — while a Bedrock player might get lucky with a stronghold barely past their starting biome. Neither difference changes the eye math above, only how far you’ll be walking between throws.

Key Takeaways

The 20% break chance isn’t something to work around with superstition, it’s a fixed probability you can plan against with real numbers. Craft to your player type’s number, recover survivors as you triangulate instead of treating every throw as sunk cost, and place, don’t throw, your way through the final 12 sockets once you’re in the portal room. Get the triangulation angle right — a sharp offset between throws beats a quick straight-line guess every time — and you’ll spend more of your Nether trip fighting the dragon than digging for a stronghold you triangulated badly.

FAQ

Do I really need 12 eyes, or can I get away with fewer?
You need 12 filled sockets to activate the portal, full stop — but you rarely need to craft all 12 yourself. Between the average 1.2 pre-filled sockets and the sheer luck factor, treat 12 as the ceiling to prepare for, not a number you can shop down by “getting lucky.” Plan around the full 12 and let any pre-filled sockets be a bonus, not a discount you bank on.

Is there any way to reduce the 20% break chance in survival?
No — despite what you might read, there’s no console command or gamerule that changes it in vanilla survival. A community-submitted feature request for exactly that exists on Mojang’s own feedback site, which is likely where the idea spreads from, but it’s a suggestion, not a shipped mechanic. The 20% is fixed unless you’re running a mod that specifically changes it.

Why did my eye fly downward on the very first throw?
Because you started within 12 blocks of a stronghold horizontally by pure chance — it happens. Don’t assume something’s broken; throw once more to confirm the exact spot before digging.

Does the Eye Spy advancement require anything special?
Not really — despite the name suggesting you need to physically follow an eye to completion, the actual trigger is simply entering a stronghold [4]. It’s a Java Edition-only advancement (Bedrock uses its own separate Achievements system), and anyone who completes the hunt above unlocks it automatically.

Can I lose Eyes of Ender if I die while carrying them?
Yes — like any inventory item, they drop at your death location and despawn after the standard timer if nobody recovers them. If you’re digging down into a cave system to reach a triangulated stronghold, that risk alone is a reason to carry a couple more than the bare minimum for your player type.

What happens after the portal is lit?
You’re through to the End dimension and the fight for the dragon — our Minecraft End Dimension Guide covers the stronghold layout, the End itself, and the dragon fight from here.

Sources

• Eye of Ender — Minecraft Wiki
• Stronghold — Minecraft Wiki
• End Portal Finder — Omni Calculator
• Advancement — Minecraft Wiki

Michael R.

I've been playing video games for over 20 years, spanning everything from early PC titles to modern open-world games. I started Switchblade Gaming to publish the kind of accurate, well-researched guides I always wanted to find — built on primary sources, tested in-game, and kept up to date after patches. I currently focus on Minecraft and Pokémon GO.