Talk:Tutorials/Iron golem farming

From Minecraft Wiki
Jump to: navigation, search

Bedrock experiments[edit]

Experimental Bedrock 1.11 village converted to iron farm

According to YouTube videos like this one, in Bedrock Edition 1.11

  • Only beds and villagers are needed to spawn iron golems
  • Villagers don't need workstations and don't need to sleep in the beds

In my quest to figure out how to convert an existing village to an iron golem farm in a way that's feasible in survival mode, I found that those two facts aren't enough. So far I have learned:

  • The iron golem often spawns during the mingling period of the day, but also at other times.
  • The spawn location appears to have no relationship to the placement of beds.

In the image to the right, my farm (with 24 beds symmetrically arranged in the buildings around the water trap) seems to spawn golems mostly between the well and the trap. The one I observed spawning firsthand appeared about at the crosshairs in the image, near the bell under the well roof. I suspect that the spawning is either in a region around the bell, or in a region around a specific gossiping villager like in the Java Edition.

I wish I could figure this out. Maybe next I'll suspend the bell over the center of the farm and see what happens. ~ Amatulic (talk) 06:52, 12 May 2019 (UTC)

Update: I destroyed the two small buildings on either side of my trap in case they extended the spawn area off to the sides. Now I have just two parallel rows of beds with the water trap in between and slabs on the ground outside. That didn't help. The iron golem still spawns near the well.
So I removed the bell from the well, and suspended it above the lava pit in the trap, and created a fenced pathway 3 blocks above the water trap so the villagers could get to it by stairways. Sure enough, they started mingling there, or around the stairways... and a golem spawned again, but between the trap and the well.
Then I broke all the doors in all buildings except my sleeping buildings, so the doors are symmetrical around the trap. The next 3 golems spawned closer to the trap, but still outside of it. The spawn rules seem buggy; occasionally a golem spawns halfway stuck in the ground.
I suspect that in Bedrock 1.11, the golem spawns around some center that isn't determined by beds, the bell, doors, or any specific villager. ~ Amatulic (talk) 17:49, 14 May 2019 (UTC)
It seems the YouTube video for Bedrock isn't valid anymore, so I deleted the related content from the article.
In my experiments in the flat world, iron golems don't just spawn in a range away from the head or the foot of a bed. They can spawn all around. Strangely, no matter what orientation I use for the beds, the golems never, ever, spawn on the south side of the farm. They're heavily biased toward one side. I think the Bedrock golem spawn mechanics are buggy still. ~ Amatulic (talk) 04:37, 17 May 2019 (UTC)
Turns out spawn center is determined by beds, or a bell, sometimes one or the other. It's unpredictable and hard to control. I updated the article and added a couple of screenshots to MCPE-41886. Not much else to add to the Bedrock part of this article until Mojang addresses the problems. Golems spawning half-buried in the ground, and spawning on slabs, and unpredictable or constantly-changing spawn centers will make AFK farms impossible, requiring constant monitoring and intervention. ~ Amatulic (talk) 21:00, 14 June 2019 (UTC)

Cat Cap[edit]

In Amatulic's edit, the cat cap was included. But is it proven that the cat cap actually exist just like on Java? I would like a clear answer because some people have stated that the cat cap doesn't exists as proven in their tests. If we want an efficient iron farm, we need to know it's spawn requirement, spawning volume, and limitations as well.--skylord_wars (talk) 10:03, 27 July 2019 (UTC)

@Skylord wars:, the positioning of that paragraph may have been an error on my part. I think I saw the information about a cat cap in a Bedrock context. If someone can verify that the cat cap doesn't apply to Java, then the section should be moved inside the Bedrock section. ~ Amatulic (talk) 05:53, 1 August 2019 (UTC)
@Skylord wars: As you yourself commented on the cat cap is in Bedrock. I'm putting that paragraph back in the overall introduction for now. ~ Amatulic (talk) 23:30, 3 November 2019 (UTC)

The above user seemed to suggest that the cat cap is a bedrock-only feature, however in the article it is tagged as a Java-only feature. Can anyone confirm which is correct? 10:58, 2 June 2020 (UTC)

I summoned 12 cats in the same room as the villagers in a Java Edition iron golem farm and the farm still functioned normally. -- 00:32, 14 January 2021 (UTC)

There's no such thing as a cat cap in either Bedrock or Java Edition. This may be from an old version - I don't know. But it hasn't existed since at least 1.14. I'm removing the references. 10:15, 24 March 2021 (UTC)
Turns out I can't remove a section. Could you do it for me Amatulic ?

Village center[edit]

Skylord wars - Thanks for fixing things up last month. In this edit you changed the definition of village center to "Iron golems can spawn at a 16x5x16 around the village center — 8 to the north and west and 7 to the south and east horizontally, 1 block above and 4 blocks below the bottom of the center block."

Last time I checked this (in Bedrock 1.11), iron golems were spawning up to 8 blocks horizontally from the center block (which is a 17x17 area). I'll grant that the center might be considered a corner of a block and the spawn area is 8 blocks around that corner, which would be 16x16. Maybe I got the directions confused.

The thing I had tested more extensively, though, was the vertical spawn range. I observed the feet of the golem would spawn no lower than 3 blocks below the bottom of the center block, and 3 blocks above (including the center block meaning the feet could be in 2 of the blocks above the center block). Has this changed? I haven't had a chance to fire up Minecraft in a while. ~ Amatulic (talk) 06:21, 23 September 2019 (UTC)

Concerning village popularity[edit]

If I build a standard farm that uses lava to kill the golems, but I live (and the golem farm is located) near a village, will the system harm my popularity with the village I live in? --2057clones (talk) 00:24, 19 April 2020 (UTC)

It shouldn't. Your popularity goes down only if you do the killing, not if the iron golems die from "natural" causes like running into lava.
Kills made with a trident killer may be counted as your kills, though. I am not sure. Trident killers are more trouble than they're worth for an iron golem farm, in my opinion. ~ Amatulic (talk) 01:47, 21 September 2020 (UTC)
Trident killers with impaling 5 reduce the time required to kill a golem which can increase rates. It's 12 seconds with lava and about 6 with lava and tridents (don't quote me on that last figure). So especially in 10 villager farms, it does have an effect on rates. Less so if you have 20 or even 30 villagers. 10:11, 24 March 2021 (UTC)

Education Edition[edit]

Even though Education Edition seems to be run on the same thing as Bedrock 1.14 I can't seem to get Iron Golems to spawn at an artificial village or even in a real village, help much appreciated.

Bedrock Video Links[edit]

The first two links currently, by Rufus Atticus are excellent and accurate. They're quite heavy going for the more casual player, but I think they should stay.

The second link by "Rah" is a nicely made tutorial, but has some problems in the mechanics it describes. It's incorrect about how golems and cats are detected in the village and about cat spawn mechanics in general. It also doesn't account well for village center shifting which could reduce the area of available spawning platform. Finally, it claims rates of 350/h when the maximum it can produce is 325/h (even if the village centre is perfectly aligned, which it won't be the way it's built in the video) so is misleading in this respect. The primary benefit of this farm seems to be that it only has one platform, which saves the player laying down one more platform of blocks for a rates increase of about 50/h. I'm OK with it, but what does the community think about correcting some of the things said in the video on the wiki page?

The next video is by Old Guy with a built in trading hall. I'm fairly certain he wouldn't recommend it as reliable. His newer, "Mjollnir" iron farm (without trading hall) is much better - and has been the best and only reliable iron farm tutorial for Bedrock for a long time. It should be on the wiki. The final video is by GruvaGuy. It is misleading about the rates it will achieve (400/h when in fact it will be less than 380/h). It has an arrangement of beds and villagers such that villagers are unable to connect to a bed if they ever unlink - which will happen as a minimum on game updates. So this farm is guaranteed to break... There are better farms out there.

I'd also like to propose adding this video (full disclosure: it's mine). As it includes a new mechanism for ensuring Bedrock Iron Farm reliability, by deliberately un-linking and re-linking villagers to workstations. It's extremely effective, and unlike Old Guy's method of achieving reliability (multiple pods, with different professions in each pod and every villager has to be traded with) it's simple as it only uses a single pod of 20 un-traded villagers. It also has optimised platform size with no drop chute and no impact if the village centre shifts. I don't want to fall foul of any self promotion rules, so want views from the wiki folk. Eggfur (talk) 13:45, 24 March 2021 (UTC)

 Support: Your video would be a great addition to the rest of the video's, as it is a different way of making an iron farm.
 Support: The video of Old Guy, I would say replace it with the newer one. I don't know about the misleading part of the rates as I did not try this farm recently enough.
 Oppose: I'm not sure where you get your information about the rates and reliability of the iron farm in my tutorial.
The reliability: I've tested it extensively and it never breaks down, it always works. I've never had any problems with the village center shifting as you say and by my testing: as long as it is in the air and not within 100 blocks of other village center contraptions (as stated in the video as well) you'll have a decent amount of iron.
The rates of 350/h are based on 10 separate hours of test running in a normal world and 350/h is the average of that.
And could you explain to me what exactly is incorrect about how the golems and cats are detected in the village and about the cat spawn mechanics? As far as I know, I'm describing exactly what the mechanics are, as also stated here on the wiki. The RAH (talk) 10:25, 25 March 2021 (UTC)
Your farm misses some spawn attempts because the spawn location can be 8 blocks from a bed pillow, and there are several locations like that beyond the edge of the farm. The village center may be shifting a lot, but that isn't a problem, it will still get most of the golems. Amatulic (talk) 17:42, 26 March 2021 (UTC)
 Oppose both videos as failing criteria I outline below (they aren't practical and aren't thought experiment exercises). Amatulic (talk) 17:42, 26 March 2021 (UTC)
@Amatulic:I am baffled why you think an Etho hopper clock is not practical! In return, you get a single pod that you can load all your villagers in rather than having to distribute them, you eliminate the risk of villagers linking to the wrong workstations. And as a build, it's really simple as I'm using the floor as my first platform, the second platform is optional (as described in the video). Couldn't be a much simpler build - and trust me: I know how to over-complicate it to get marginal improvements in rates ;) Eggfur (talk) 18:09, 30 March 2021 (UTC)
@The RAH:Firstly let me say I thought your tutorial was really well produced and I'm not suggesting it be taken off...
check Rufus Atticus' videos which are also on the wiki for the question about rates. Towards the end of the second video he gives theoretical rates for different styles of farms - all backed with very detailed maths. Your farm would average 325/h if your village centre had a full 17x17 spawning platform around it.
In your video you started by placing the beds on one end of the bed platform - that is your initial village centre and it's far from the centre of the platform and doesn't have 8(+1) blocks in all directions from there (it has 5). Even if you'd placed the centre bed first, on a game upgrade, the village can reset and the centre can become any of the beds (NW lower corner of the pillow), so the platform should ideally be bigger to account for that. It's not a huge deal - and it's only recently that the big YouTubers have cottoned on to this fact.
ref detection of Golems and cats: Your video suggested that if they drop down 6 blocks they're out of the detection range (that's true of spawners by the way) but for cats and golems they have to leave the village perimeter AND EITHER their timer expires and they stop being a village dweller, OR they join another village and stop being a dweller of the original village. The timer takes far too long for an efficient farm. Default village size is 24 blocks vertically around the village centre, so they'd need to drop at least 13 blocks below the beds, and in reality 41-80 so they can join a new, unconnected village. Your farm would actually be marginally more efficient if you moved the kill chamber up as it would kill the golems very slightly quicker (no-one would notice!)
Reliability is related to what happens after some game crashes and game upgrades. Again, the village can be reset, causing villagers to link to workstations they can't reach. It doesn't happen every time, but every time there is an update I see a rush of posts on Reddit from people saying, "My iron farm stopped working". Fundamentally, if there are workstations that a villager can't reach to work at, but could potentially connect to, then that can damage reliability. I suspect this can also happen more easily with servers/realms as well, but that's anectodal.
Eggfur (talk) 18:09, 30 March 2021 (UTC)
My view is that redstone components are unnecessary for teaching readers how to construct an iron farm. They are also impractical for an early-game farm. They are an enhancement over and above what is needed to have a regular source of iron.
Cast as an enhancement, a video can be made to focus on that feature, but there is no need to go into detail about the construction of the rest of the farm. In some cases an exotic mechanism like a trident killer may be needed for a farm to work, but such farms (like conversions of pillager outposts) are understood to be advanced builds. Iron is a basic necessary resource (I sure have to be careful with that resource in my island survival world with no villagers within thousands of blocks, no hope of making a farm), so for this tutorial, videos should emphasize simplicity and practicality, not thought experiments or examples that require something exotic.
I can see that it would be appropriate to include your video if given the proper context, but at the moment it has no context other than being in a list of other videos, and it violates the desired quality stated in MCW:VIDEO: "If a video is used in a tutorial, it is preferred that the content in the video also be represented internally, such as in prose, in screenshots, or in schematics". Except for the two videos specifically about mechanics, the rest of the videos in this tutorial lack that necessary context. Without that, there's no need for them. Amatulic (talk) 19:15, 30 March 2021 (UTC)

Video priorities[edit]

In my view, for any tutorial (not just this one), if it has videos at all (and in most cases videos aren't needed), the videos should be curated by the community based on clear guidelines. Inclusion criteria could have this order of preference:

  1. Mechanics. Absolutely #1 priority, but this sort of video is rare in our tutorials. Videos explaining mechanics without focusing on a specific design are most appropriate for a tutorial, to give readers an understanding of how to create their own farms without merely copying a design.
  2. Example designs that could reasonably built in Survival mode without exotic materials, and reasonably early in game. When I've written tutorials, I always write from that perspective, although I don't make videos. RAH's video is almost there, but the design suffers from needing magma blocks (easily corrected with an alternate killing chamber design) and putting villagers in the sky. And Eggfur's design requires redstone components. For that reason I'd prefer they be omitted from this article. Before Bedrock 1.11, when the spawn region was defined by the geometric center of the doors in the village, my favorite survival activity was to find a village and convert it to an iron farm, without needing to move villagers around or disturb their work (example). Unfortunately that doesn't seem practical anymore, but it would be great if someone came up with a village-conversion design using new mechanics.
  3. Lowest priority: "Thought experiment" videos about maximizing yield showing designs that are often extravagant and impractical except in Creative mode, but may be useful for educational purposes. This type of video is the kind that proliferates in tutorials, but it's also the kind that we need the least.

One thing I really don't like seeing is a video that starts out by begging the viewer to click on "like". Don't insult me by asking me before I've seen it, ask me at the end when I can make an informed decision. Amatulic (talk) 17:42, 26 March 2021 (UTC)

I agree with most of what you're saying. Bedrock Iron farms are a bit strange though. They're actually simple to optimise, but hard to make reliable. That's why my (Eggfur's) tutorial and the "Mjolnir" one by Old Guy are actually special, since they're the only two designs I know of that are reliable. You take your choice - you either trade with every villager before placing them in the iron farm, or you use redstone (it's a few pistons and a hopper clock - OK you have to have been to the Nether, but it's hardly intricate!). Anything else I know of creates a design that breaks. I'm less worried about Magma blocks in RAH's video, which can be obtained in day 2 of a world quite easily, but more about the incorrect mechanics described and the fact it's not reliable - surely simplicity can't be more important than function? As an aside, I did think about showing a village conversion - it's simple to do, but it adds quite a bit of time to an already lengthy build video. You just need to remove the beds, bells and workstations from the village after you've built your farm, but it makes getting the villagers in the first place a lot easier.
I also agree with you on the whole "likes/subscribe" thing. It's not ideal for a wiki tutorial. However, for creators, it can be an important motivator - which drives them to create more so in the end it's arguably good for the community. I'd be perfectly happy to do a redo of my tutorial in a more wiki friendly way, if we can agree to get over redstone being an issue. If that's what it needs to make the farm work, it shouldn't be excluded in my view! I've also been planning a proper mechanics tutorial, but the people you want to target with simple builds are the ones least likely to watch, understand or be able to apply a mechanics video.
Could you also take a look at the cat cap comment I posted, please. I can provide proof that it doesn't affect iron golem spawning if you like, but I'd rather not have to, since there can't be any evidence it does...
Eggfur (talk) 11:19, 30 March 2021 (UTC)
Let me answer those points.
What do you mean by a design that breaks? What would make RAH's farm completely cease to function? Being sub-optimal isn't the same as being broken.
If production rate is not an issue (you just want reliable production) you can still make 100% reliable farms without venturing into the Nether or resorting to exotic mechanisms that aren't practical early on in survival mode. I've made reliable iron farms with common materials (the most exotic being lava), without building in the sky or needing impractical contrivances to position villagers.
If the villagers relink to a workstation they can't reach and less than 75% of them are able to work as a result, then the farm would completely break. That shouldn't happen under normal circumstances and RAH reports that he's had no issues, which is great. There was also a recent bugfix to "prevent villagers stealing workstations", but I don't know for sure if that's a full cure. But anything that causes the village to be recreated would cause the village centre to shift and the workstations to relink - breaking the farm. If it's a big enough issue to trouble what should go in the wiki? I'd suggest probably not, which is why I support keeping RAH's video. Eggfur (talk) 10:38, 31 March 2021 (UTC)
I would say simplicity is way more important than function for the purposes of a tutorial. A tutorial is a teaching article. It should teach the reader concepts in the context of building a farm. The most minimal basic farm to illustrate a concept should be the main focus of a tutorial. Improvements to make things more reliable could be described in later sections, and videos included only if necessary to illustrate a point being made in the prose. My objection to the videos here is that only two of them (the two about mechanics) actually address points made in the article text. The others are simply there for no particular reason.
About likes/subscribe - I don't object to including that at the end of a video, but I find it insulting and blatantly promotional if it's put in my face up front, and I would advocate immediate removal of any video that does this, as a violation of the video policy here. It makes it seem as the whole point is just to attract likes. Furthermore, I see no need to "motivate" anyone to produce YouTube videos. In most cases, videos aren't even needed in tutorial articles and could easily be substituted by some descriptive text accompanied by a couple of images. If someone wants to make a video to share information, that's fine. If the video is made as a means to grow an audience or supplement an income, then there's a conflict of interest and the producer of the video has no business putting it into this wiki. The Minecraft Wiki isn't anyone's personal publicity tool.
So this isn't about my video - happy not to include it, as that was the point of me asking. I also agree with you on the "likes" issue. My point was really that creators probably don't create videos for the wiki, but once made they may be useful, but have the whole like/subscribe thing. Where I think we disagree is on what makes a good tutorial. The simplest iron farm is 20 beds (5 blocks above the ground), 10 villagers with 8 workstations, a block with a campfire on it surrounded by some other blocks (one block away from the campfire) to stop the golem escaping. It's a terrible iron farm of course, but if we're saying simplicity over efficiency, that's the answer. It's also a terrible way of teaching the mechanics - because it ignores spawning range and village centres. That's where I disagree with you: you can't teach mechanics by showing how to build a farm that doesn't account for village centre or spawn platform size. So are the videos supposed to teach you how to get iron or teach the mechanics of iron farming. You seem to be suggesting either, but not both. Eggfur (talk) 07:54, 31 March 2021 (UTC)
About cat caps, I'd like to see where anything about this was resolved in the bug tracker. One of the mechanics videos in this article shows solid evidence that the cat cap is (or was) a limitation in Bedrock Edition. But things may have changed since then. Amatulic (talk) 16:43, 30 March 2021 (UTC)
I've just rewatched foxynotail's video and there isn't any evidence of a cat cap affecting iron Golem spawning. He says it a few times, but doesn't show anything. It was a pioneering video, but with what we know now there are so many things that have turned out to be incorrect in that video - even if we ignore the change in mechanics in 1.16. What evidence would work for you? I've already spawned 12 cats at my iron farm and it makes no difference to iron Golem spawns, and of course they can't spawn that many naturally Eggfur (talk) 07:54, 31 March 2021 (UTC)
Village converted to iron farm
I'm convinced. I just finished building a farm and there are cats everywhere, but the iron golems keep coming. I think that section on cat caps could be removed. Whatever was discovered as a cat cap a few revisions ago clearly no longer applies.
Give it a look (on the right). This is what I mean by a basic farm that requires nothing exotic and no need to transport villagers. It produces only 200 ingots/hour, but that's still about 10X what a converted-village iron farm produced back before 1.11 when the spawn center was determined by doors, and generally if I get three stacks of iron early on in the game, I'm pretty much set and don't need more for a very long time.
It seems reliable enough if you take steps to protect the village (if you don't, then if just one zombie manages to slip into the sleeping bunker, you end up with nothing but zombie villagers). The golem spawn area accounts for center-shifting, but because the villagers all return to their own beds every night, and the village is fenced in so they don't get lost, there's very little chance of claims on beds switching around. The beds and job site blocks are close enough together that the villagers can pathfind between them. Every morning they pour out of the cobblestone bunker and go to work, and every evening they go back there to sleep. The village center here happened to be such that iron golems spawn directly under the nearer bridge most of the time.
I also offer this as an example of a practical survival-mode build that doesn't need a video. A few screenshots and a simple description of the construction is sufficient. There are some nuances to the bridges and water trap, but otherwise it's all basic concepts for guiding the iron golem into a lava blade. Amatulic (talk) 07:23, 1 April 2021 (UTC)
That's a really cool design. You should probably mention that you need to remove any beds outside of the communal sleeping area. Personally I'd never do a video tutorial for it because I reckon over 60% of people who tried it would get something wrong enough to make the farm bad or non-functional. Probably more than that from just a picture and description. Eggfur (talk) 21:48, 1 April 2021 (UTC)
Thanks. I'm quite pleased with how it's working. I lucked out finding a source of lava a short walk from the village (the ruined portal in the picture). Because the farm itself is outside the village fence, it also catches hostile mobs that wander into the water. I wasn't sure when I started it if the villagers could pathfind across bridges and down stairs between their beds and job sites, or if iron golems would spawn on top of the bridges, but it all works. Fencing in the village first helps, because some of the villagers get confused and try to wander away from the village when they lose their old beds, so you have to nudge them close enough to the bunker until they detect a bed. The experience made me wish I could attach a lead to a villager.
Video tutorials tend to get bogged down in construction details. With this farm the construction doesn't matter much. You have to account for the size of an iron golem when making the bridges and underground water path to the killing area, and make sure the beds are all positioned correctly, but everything else is just mechanics: all the prep work and planning, making the water trap big enough, managing water flow, fencing and lighting the village, making a kill room, and so on — all covered adequately by other tutorials, no need to re-hash them. I may include this as an example of a feasible survival build in the article. The only real way to mess up this farm is to neglect prepping the village properly before you even begin building the farm, lest you end up losing all your villagers.
Back to the videos: Thinking about it more, I think what I object to the most is the potential for an indiscriminate list of videos at the end of a tutorial. Just a title "So-and-so's build" with a brief 1-line description and a video doesn't make the grade, in my view. If your video presents a unique or innovative concept for making a farm more efficient, or a completely new thinking-out-of-the-box type of farm (for example the "River Biome Trident Farm" in Tutorials/Drowned farming, added by the video author but it was such an innovative idea I thought it should stay), then I prefer to see the concept described in text and diagrams, and include the video as an example of a working implementation. The two mechanics videos, for instance, could be moved up into the Bedrock section of the article, as they are a fine accompaniment to the text already there.
So after some more thought, I wouldn't object to your video if it was presented in some useful context. Being part of an indiscriminate list, I have no way of knowing why I should be interested or why it should be useful to me. Bottom line, I really don't like to see lists of videos. Explain why it's interesting. This is a tutorial article, so use the opportunity to teach why your farm matters in the article (or better still, in a proposal on the talk page). A video should supplement the text, not substitute for it. Amatulic (talk) 23:10, 1 April 2021 (UTC)