The challenge to create better challenges?

[chengong]

Actually, I just wanted to make a thread about how to make successful challenges, because there are way too many challenges that has potential but ultimately are ruined because the OP makes some easily avoidable mistakes.

Challenges primarily come in two types, and they obviously have different design goals so you will have to keep in mind which on are you going for:

The Challenge challenge

What I mean by this is that the challenge simply has one goal and anyone who complete said goal is considered a winner. The best example of this type of challenge would be:

The Ultimate Jool-5 Challenge:land Kerbals on all moons and return in one big mission

Now let's take a look at what makes this challenge great:

• It has to be epic. Epic means something we see and be like, man I wish I could do that! and not some random task like jumping from one ladder to another, which might be difficult but it's just boring nobody cares if you could do it or not. After all, it has to be something people WANT to be able to do, for them to even have a go at it.
  
• It should be somewhat difficult. Not as strong of a requirement, because there are some really easy challenges that are apparently still very popular, like the K-prize, circumnavigation and machinebird. However when you do a challenge the difficulty usually should come from engineering and not piloting skills, for two reasons. The first one being that this game is primarily an engineering game and not Flight Simulator X. Second being since each person has their own crafts, it's never a fair competition. Unless of course you make a craft for people to use and do not allow any custom crafts. But that kind of defeats the purpose of the game so I have yet to see a successful challenge done that way.
  
• Have some different levels of difficulty. This is mostly because you want an epic, difficult challenge for hard core players. But you also want more entries and not scare everyone off so that the post doesn't just sink to the bottom. You do this by having different levels of difficulty. This also makes the hard core entries feel more epic as an added bonus.
  

If you look at the weekly Reddit challenges they pretty much all do these three important things.

The competition challenge

Basically this is the challenge with a scoring system and a leader board, and a way to win the challenge. I would say this is more difficult to establish since any challenge challenge can be converted into one of these by adding a good scoring system to the mix. Therefore this kind of challenge actually has all the prerequisites of the first type, with added complexity:

• Scoring system. It has to reward what the challenge is ABOUT, and not anything else. Or else you end up in an uncomfortable position where someone submits an entry that earns huge points but isn't really what you wanted your challenge to be. Some of you might know me for doing that.
  

Low cost Launcher

The challenge above an example of a structured scoring system that doesn't focus on what the challenge is about. So you have a challenge for "low cost" but then your scoring system rewards crew capacity instead of you know, low cost? Well that's just asking for trouble isn't it? Now you have no choice but to either change the rules or resort to that "spirit of the challenge" thing, which is ultimately a get out of jail for free card for bad challenges, as good challenges never seemed to need it.

• Variety of solutions. This means that the challenge should (potentially) have a variety of solutions, and not just, who can spam the most number of speratrons onto a probe core. Whenever you have a challenge about spamming parts, you have a big problem. Not only is it unfair to those with no access to super computers. It's just unfun even when you do spam the most. There are certain types of challenge that are especially prone to spamming part. Things that involve speed and jet engines (intakes) for the most part.
  

Challenges that involve lowest mass or part count to some simple goal such as lowest mass to orbit, are especially bad because when you only have three or five parts, and there are only three or five parts to choose from. People will quickly find the absolute optimal solution and kill the challenge.

• Mods. While it may be tempting to either allow all mods or disallow all mods. Neither is a good choice here. Allowing all mods gives you entries with magic engines with infinite fuel and insane thrust. Not allowing any mods, well most people who don't do any modding are probably on Reddit. For convenience, I'll just put up a table here listing the fairness of the most popular mods. Mods will have a power rating from 1 to 5 where 2 is STOCK. Some mods will have 0 which stands for mods that add entirely different functionality to the game and not just variety of parts.
  

[TABLE="class: grid, width: 500"]
[TR]
[TD]STOCK parts[/TD]
[TD]2[/TD]
[/TR]
[TR]
[TD]Infiniglide/Ladder power/Kraken drive[/TD]
[TD]4[/TD]
[/TR]
[TR]
[TD]B9 Aerospace[/TD]
[TD]3[/TD]
[/TR]
[TR]
[TD]KW Rocketry[/TD]
[TD]2+[/TD]
[/TR]
[TR]
[TD]Nova Punch[/TD]
[TD]3[/TD]
[/TR]
[TR]
[TD]FASA[/TD]
[TD]3+[/TD]
[/TR]
[TR]
[TD]FAR[/TD]
[TD]3?[/TD]
[/TR]
[TR]
[TD]NEAR[/TD]
[TD]3+[/TD]
[/TR]
[TR]
[TD]Kethane[/TD]
[TD]4[/TD]
[/TR]
[TR]
[TD]RLA Stocklike[/TD]
[TD]1?[/TD]
[/TR]
[TR]
[TD]MechJeb autopilot functions[/TD]
[TD]3[/TD]
[/TR]
[TR]
[TD]dynamic strut mods[/TD]
[TD]2+[/TD]
[/TR]
[TR]
[TD]Propeller mods[/TD]
[TD]2+[/TD]
[/TR]
[TR]
[TD]Advanced Jet Engine[/TD]
[TD]1[/TD]
[/TR]
[TR]
[TD]realistic fuels and engine configs[/TD]
[TD]4[/TD]
[/TR]
[TR]
[TD]Hot air balloons[/TD]
[TD]4+[/TD]
[/TR]
[TR]
[TD]realistic reentry, life support, part failure[/TD]
[TD]1[/TD]
[/TR]
[TR]
[TD]Near Future packs[/TD]
[TD]3+[/TD]
[/TR]
[TR]
[TD]FTmN and Atomic Age, other NTR[/TD]
[TD]2[/TD]
[/TR]
[TR]
[TD]KAS[/TD]
[TD]0[/TD]
[/TR]
[TR]
[TD]That awesome mini shuttle mod[/TD]
[TD]2[/TD]
[/TR]
[TR]
[TD]Interstellar[/TD]
[TD]4?[/TD]
[/TR]
[TR]
[TD]Impossible Innovations[/TD]
[TD]5[/TD]
[/TR]
[TR]
[TD]HyperEdit[/TD]
[TD]5+[/TD]
[/TR]
[/TABLE]

Please suggest for additions and changes.

Ultimately, allowing Mods is for the convenience and fun of people who want to participate in the challenge, and not about your own convenience. Some people don't want to install mods just to enter the challenge because it's either required or allowed but super OP. Neither do people want to uninstall mods just so they could entry the challenge. So you have to make sure that you allow as many mods as possible, without breaking balance. if that means you have to do research on and try out certain mods, I suggest that you do and not just disallow it because you don't know what it does. Neither should you allow certain mods just because you like it and you don't want to give it up.

I would say for most challenges, tier 3 mod will be the highest you will allow. Some of these things are clearly better than stock like the B9 crew cabins compared to stock crew cabins. So if you have a challenge about crew capacity you certainly want to address that.

Tier 4 mods are questionably powerful and probably should never be allowed unless the challenge is entirely about using that mod, which is ill advised to begin with because not many people will have that mod, whatever it might be.

Anyone allowing tier 5 for anything other than setting up your challenge, is probably insane.

FAR although makes atmospheric flight less tolerant. It also significantly reduces drag, reducing the required Delta V to orbit from 4,500 to 3,500. Which is a huge advantage in of it self. However since the jet engines are nerfed, on top of the more delicate design requirements, spaceplanes are more difficult with FAR. if you have an interplanetary challenge this is probably not a big deal. But make sure you address it for any atmospheric challenges.

And yes, some mods are actually underpowered. Like RLA stock-like, at least its engines. You can beat all of them with the su-74s, which is an incredibly overpowered stock engine. So OP in fact, some challenges are just begging for everyone to spam this engine as much as they can.

On Part clipping

I think the biggest problem with clipping, is the tiny strut, and not really clipping itself. Also this is more of a problem with intakes more than anything else. So if you're not having an intake spam competition, you probably don't need to worry about this.

Edited January 4, 2015 by chengong

[Right]

Good post. I don't see the "spirit of the challenge" defense as a get out of jail free card, but if you have to resort to it...your challenge is in danger of losing your serious competitors.

Also, I don't feel mech jeb is 'powerful' in the same way impossible innovations. It makes the game easier in a less-time-consuming way, not in a TWR, dV, design possibilities way. Afterall, you mentioned yourself that the difficulty of a challenge should be rooted in design, not piloting skills. Not a big deal to me though, I know some people are very passionate about the mechjeb debate

I might recommend simplicity in scoring criteria for people who don't want to spend hours developing a challenge. Keeping the way(s) one can score points to a minimum will reduce the probability of unintended complications.

Good points in all Chen.

[sdj64]

Good points. I'll add - a min/max engineering challenge won't be fun unless the min/maxing is too hard to get perfect, nor rewards CPU power over engineering competence. As you said low part count challenges are very easily perfected. A challenge that rewards most fuel, most tons, or most kerbals simply favors people with better computers. Something like the Low-Mass Jool 5, however, is not either of those and so it keeps people coming and trying to improve a little bit more. A good challenge in the "Competition" category has the simplest scoring system possible, without flourishes or bonuses, lest you make a mistake and not cap your bonuses leading to either of those examples you provided.

I've competed in many challenges and even started a few not very successful ones, so that's what I've found to be the best formula.

In the "Challenge" category for some more examples of very successful ones besides the obvious Jool-5:

Eve Rocks Challenge - The most well put-together Eve return mission challenge, part of its success has to do with the fact that it looks a lot like the Jool-5 challenge.

Constellation Space Program - I wouldn't guess that this could have succeeded because it has so many rules, but it did. Replicating real-world proposed missions is fun for lots of people.

Flying Duna Again - Has its flaws (no maximum score for number of Kerbals?) but it's very difficult which could explain why few people exploited them.

Doing it Apollo Style - Got lots of rules but not too many and does a good job recreating the iconic mission.

Except for Eve Rocks and Constellation (sort of) these are pretty old and not really going anymore but they all did very well when they were popular.

[Voculus]

My biggest beef is going with the "stock only" rule. It neuters the challenge, and with very limited design options, most of the entries are carbon copies of each other.

[dbmorpher]

Not necessarily, stock only forces original ideas also hopefully contestants would look at other attempts and try to do something else.

It also makes it more of a challenge.

[chengong]

The biggest problem with stock only I think is still accessibility. I had to make an entire copy of the game that is stock only so I can do these challenges without having to remove various mods. If you want pristine stock communities you really should head over to Reddit where people do grand tours on pure stock and calculate Dv with Excel.

Also some of the stock only entries are just insane exploit fests, like intake spamming to the point you can run jet engines at 50km.

And keep in mind that design limitations such as stock only, only forces original ideas IF THERE IS ONE. Being a game with limited number of parts, there are a very finite number of possible designs out there, so no, you cannot always have new ideas. usually the best idea has already been tried, you just have to take it further than everyone else to win.

[thereaverofdarkness]

I would rate KW Rocketry as either a 3 or 2.5 at least. Not only do they offer a lot of options, but many of the options are just a lot better than stock. Quick examples:

1.) The Wildcat is not only a more powerful 1.25m rocket but also has a much better TWR than the LV-T45 without sacrificing thrust vectoring or much of the specific impulse.

2.) The Vesta is both more efficient and more powerful than the LV-909, and has a better TWR.

Now I haven't used it in a while and most of these details come fro memory so I might be a bit off, but I remember the Wildcat and Vesta being far more effective that stock engines, and as a result I didn't use much of anything else for 1.25m.

Secondly, I would rate MechJeb at 1, not 3. MechJeb only assists with piloting and maneuvers, and is less competent than a well-trained human at pretty much everything it does. It lacks too much in judgement to make up for it in precision even in maneuvers that require a high level of precision, partly due to the tools available on the navball. I have never used MechJeb and I've yet to see anyone (MechJeb included) that can do any maneuvers I can't also do. It's not much to brag about, it's just a fact that the most difficult maneuvers are still pretty easy once you have enough experience. MechJeb is training wheels; you cannot ride as fast until you take MechJeb off.

[chengong]

I would rate KW Rocketry as either a 3 or 2.5 at least. Not only do they offer a lot of options, but many of the options are just a lot better than stock. Quick examples:

1.) The Wildcat is not only a more powerful 1.25m rocket but also has a much better TWR than the LV-T45 without sacrificing thrust vectoring or much of the specific impulse.

2.) The Vesta is both more efficient and more powerful than the LV-909, and has a better TWR.

Now I haven't used it in a while and most of these details come fro memory so I might be a bit off, but I remember the Wildcat and Vesta being far more effective that stock engines, and as a result I didn't use much of anything else for 1.25m.

Secondly, I would rate MechJeb at 1, not 3. MechJeb only assists with piloting and maneuvers, and is less competent than a well-trained human at pretty much everything it does. It lacks too much in judgement to make up for it in precision even in maneuvers that require a high level of precision, partly due to the tools available on the navball. I have never used MechJeb and I've yet to see anyone (MechJeb included) that can do any maneuvers I can't also do. It's not much to brag about, it's just a fact that the most difficult maneuvers are still pretty easy once you have enough experience. MechJeb is training wheels; you cannot ride as fast until you take MechJeb off.

yea the small engines are not bad but they're also really long. I just had the impression that the larger engines are stupid heavy while Novapunch engines are lighter with equal performances.

Having mechjeb does not prevent you from also using the stock navball tools so it can't possibly be less than 2 in power rating. Plus mechjeb is a tool and you have to use it intelligently, so you're not going to let it fly a spaceplane into orbit for you because you know it sucks at it. however you can let it do a landing and it will start firing engines at the last possible second, and save more fuel than you doing it manually. For me Mechjeb is most useful at planning burns, which you can then tweak manually which is a lot faster than doing it manually from start. Short burns can also be carried out by Mechjeb with great precision, although it won't do longer ones for your ion missions. Again it's a tool and you have to use it right instead of expecting it to do everything for you.

I tend to be lazy and use auto dock for example. It may not be as efficient as manual but you always have the option to have it hold the ship in the proper direction while you do the docking translations, so helpful when you don't have a ship with balanced RCS, which is kind of annoying to have to do without mods like RCS build aid.

Edited January 3, 2015 by Claw

[Right]

however you can let it do a landing and it will start firing engines at the last possible second, and save more fuel than you doing it manually

Landing is one of my favorite mechjeb features, but its not even the most efficient at this. You can do it with less dV that it does with practice.\

Having mechjeb does not prevent you from also using the stock navball tools so it can't possibly be less than 2 in power rating.

Also, if this were true, most every mod would be at least a 2 since you can always use stock parts.

[HafCoJoe]

I would say that to make a successful challenge, it must have simple rules, and be entertaining. Plain as that. Difficulty usually isn't a part of the question in challenges as people usually know how to make something hard.

This one is crazy hard to do, but it still stands as an example of a good challenge imho:

I made this jewel a while back based off WhackJob's challenge in the "Kerbin Cup" that Rowsdower had going a few months ago, but it was so difficult that no one took it on.

Assemble Arkingthaad ULTRA-lite.

The rules are plain and simple. Build this monstrosity - you may use any technique (ALT-F12 excluded), any mod (Besides Hyperedit), and any means necessary to accomplish this.

Currently it looks like this:

Download the persistence HERE. The save file should still work.

Winners are calculated are counted by amount of in-game time used. Stock players have their own category, and Hyperedit users have their own too.

Edited January 3, 2015 by Avera9eJoe

[chengong]

Also, if this were true, most every mod would be at least a 2 since you can always use stock parts.

Because mechjeb you could attach to your ship and gain additional benefits. Where as if you used RLA engines you no longer have stock engines with you.

if you had a craft made with RLA engines instead of su-74s then you craft would have less performance. But if you have a craft and attach a mechjeb box to it, it doesn't loose any power.

I would say that to make a successful challenge, it must have simple rules, and be entertaining. Plain as that. Difficulty usually isn't a part of the question in challenges as people usually know how to make something hard.

This one is crazy hard to do, but it still stands as an example of a good challenge imho:

I'm sorry but it just doesn't turn me on, When I look at it I'm like, who cares? why should I be able to do that? It lacks epicness.

You can't just say "make it entertaining", that's like saying, "make it good!". This is a game so obviously good is entertaining and vice versa. The problem is HOW you make it entertaining.

Edited January 3, 2015 by chengong

[Claw]

I really like the idea of this thread. This idea pops up from time to time, attempting to inject a little more rigor into the challenge creation philosophy. There was a thread around here attempting to create some standardized "what's allowed" type of rules, but I think that one faded away.

I would say that to make a successful challenge, it must have simple rules, and be entertaining.

I think I tend to agree with this. Simple, straight forward rules are easier to understand and stick to. And frankly, if the challenge isn't entertaining, then it is likely to get skipped.

...When I look at it I'm like, who cares? why should I be able to do that? It lacks epicness.

...The problem is HOW you make it entertaining.

But this also gets to the crux of the matter. Entertainment for one person isn't entertainment for another. Some people love exploring planets. Some people love building massive space stations, or miro space planes. That's why it's nice to see a variety of challenge types, because everyone has different interests. The epic challenge of exploring the entire Jool system simply doesn't appeal to some people, as an example.

I didn't try to make my recent challenge an epic challenge, but left the rules flexible so that people can do what they want. Some people took the chance to do their own epic adventure. I'm sure leaving flexibility also left loopholes in the rules, but it's also nice to keep it simple. Perhaps you can pull some examples of good and bad from that as well.

---

In either case, keep up the brainstorming. Just please remember to keep the examples positive and the critiques about philosophy, since everyone has their own ideas of what is epic, entertaining, and lame.

Cheers,

~Claw

[Pds314]

Can we make this, or something like it, stickied? It might help the challenge quality a lot.

[chengong]

shameless bump

[redsh]

I think this deserves to be sticked. Really good and useful guide.