Public Commander Keen Forum

So lately, I've been asking myself, if I know how difficult a given level is, and how many points are in a given level, can I estimate how profitable a given level is vs its risk?

Just looking at Keen 1 in the Keen wiki, I did some number crunching. What I found was interesting, and I figured you guys might like to see it too:

https://ibb.co/dQg0jQ

Some explanation:

These calculations were done assuming that you collect every point possible in Keen 1, kill every possible enemy, and never die. This includes shooting the chain for the vorticon commander. "The shots needed total" refers to the number of shots needed to kill everything. "Free Lives" refers to the number of lives you can earn by collecting points, assuming you never die.

Scaled Opportunity is a ranking, 0-10, describing the point and ammo density of each level. 10 is the most dense, 0 is the least dense. The more point and ammo dense a level is, the more rewarding it will be the more you explore it.

Scaled Difficulty is a ranking, 0-10, describing how deadly the level is overall. 10 is the most deadly, 0 is the least deadly. This takes into account the number of kill tiles, gargs, tank bots, and vorticons. These are all weighted by my estimating how often I die to each. I do not include yorps or butler bots, because they just push you into the other things which kill you.

Scaled Difficulty Density refers to the density of the things that can kill you within the level. 10 means that there's a lot of killy things in a compact area, 0 means that the killy stuff is very spread out and/or non-existant.

Best Choice refers to how valuable a level is to the player taking into account the difficulty. 10 is the most valuable, 0 is the least valuable.

Best Choice with Area refers to how valuable a level is to a player when you compare the point and ammo density to the difficulty density.

I want to do this for the other games too. What do you think?

An interesting idea. While I generally distrust quantification (or whatever the right word is) it can bring interesting insights. Something like this could be a crude way of enforcing a difficulty curve for one's levels.

I actually did something similar a while back, but used ammo\points per second as a measure. A level has an *absolute* reward, but also a relative one; you could make a level that was just one long hallways of alternating red\blue doors, forcing Keen to walk back and forth longer and longer distances to get 5'000 points, but it wouldn't be very FUN. A level that can be completed quickly and offer a lot of points is more 'fun-dense'. (To a point, mindlessly collecting points is no fun either.)

This was as part of an attempt to calculate the highest possible score a player could get on all Keen games as well as how long that would take. (In the first three Keen games this pretty much maxes out the counter, you just have to calculate the most time-efficient manner to do so. In Keen Galaxy this requires more strategy.)

Never did finish those calculations.

There's a lot of things these calculations don't catch. I could make a level impossible due to 1 kill tile, fill the level with points, and these calculations would say its the best level ever.

Levellass wrote: ↑Sat Sep 23, 2017 1:24 I actually did something similar a while back, but used ammo\points per second as a measure. A level has an *absolute* reward, but also a relative one; you could make a level that was just one long hallways of alternating red\blue doors, forcing Keen to walk back and forth longer and longer distances to get 5'000 points, but it wouldn't be very FUN. A level that can be completed quickly and offer a lot of points is more 'fun-dense'. (To a point, mindlessly collecting points is no fun either.)

This was as part of an attempt to calculate the highest possible score a player could get on all Keen games as well as how long that would take. (In the first three Keen games this pretty much maxes out the counter, you just have to calculate the most time-efficient manner to do so. In Keen Galaxy this requires more strategy.)

Never did finish those calculations.

How did you calculate ammo/points per second? Did you look at the area of the level, and take into account Keen's speed? That's very clever.

I assume that in your calculations for Keen Galaxy you were ignoring the free instant level ups, like the queen vivas and the lifewaters, right? If you take them into account you can still get an infinite score.

Here's the stats for Keen 2
Here's the stats for Keen 3

You can see the calculations starting to break down here due to weird level design. The Caves of Oblivion should be a very high opportunity level, because you can just grab the points and ignore everything that can kill you, but the calculations don't take that into account. The stats for Keen 3 do not count points that are unobtainable, or enemies that are unreachable, as best as I could figure.

One measure was the theoretical density, the number of not blocking tiles divided by the number of points (I used hex counting in the data to track that.) That was a basic measure of how much ground Keen had to cover which, over all levels, should have required about the same average speed per tile.

However as a more practical measure I simply timed myself doing the levels to get a more 'practical' idea of how long it took to a.) Get through the level with as many points\ammo as possible and b.) Get ALL points. The results were clear; it's best to get the easy points, attempting to get all drastically slows your progress.

Now, if there is any level where you can get more than one life and die you can get an 'infinite' score. For Keen Vorticons this is just any level with more than 20K points. For Keen Galaxy the point limit rises and only levels with 2 1UPs become viable.

Also the score is not infinite, when it reaches its limit it flips to negative points. Therefore the challenge to get maximum points in minimum time is interesting. For Keen Vorticons for example it requires trekking between level 16, the castle and the red maze (For ammo to kill the Gargs at the castle bottom.)

Stealthy71088 wrote: ↑Sat Sep 23, 2017 2:05 There's a lot of things these calculations don't catch. I could make a level impossible due to 1 kill tile, fill the level with points, and these calculations would say its the best level ever.

Every Keen levelling-noob in the 90s


It's a joke but has some truth to it. Anyone remember those old levels with huge areas filled with teddy bears etc? Man, gotta revisit some of those packs one day.

One day we'll need to have a 'worst level ever' competition. I mean, how bad can you make a level? Not just 'Die as soon as you enter' bad but 'I have lost the will to live' bad.

KeenRush has already won it.

Yes I remember well. I indulged in them myself too... on Andy's KeenWright of course. First it was "create new level 200x200" *game crashes*. Then 'F' for fill the whole level with lollipops *game crashes*. Eventually "here's 500 ammo, now fight these 50 Vorticons". A true golden era...

Levellass wrote: ↑Fri Sep 29, 2017 1:08 One day we'll need to have a 'worst level ever' competition.

Ohhh, yes. This needs to happen. The levels should be beatable, but designed to go against player psychology. Inaccessible points, dead ends that force the player to reset, formless platforming with sadistically-placed traps, the works.

I call it "Blind Jumps: The Level".

Levellass wrote: ↑Fri Sep 29, 2017 1:08 One day we'll need to have a 'worst level ever' competition. I mean, how bad can you make a level? Not just 'Die as soon as you enter' bad but 'I have lost the will to live' bad.

KeenRush has already won it.

Actually I really like his levels. Yeah, many are crazy hard but they're oftentimes really clever and creative. Now I have played ancient levelpacks made by other people that were stupid hard, had 80 billion points and were completely unimaginative. I don't really like long hallways filled with teddy bears because it defeats the purpose of having points.

Could we do a levelpack of really horrible keen1 levels? I've thought of something that's guaranteed to enrage someone. Maybe we should set a rule that they have to be completable? Or not so much? We could even vote on who has the most infuriating level?

Stealthy71088 wrote: ↑Tue Oct 03, 2017 13:11 Could we do a levelpack of really horrible keen1 levels? I've thought of something that's guaranteed to enrage someone. Maybe we should set a rule that they have to be completable? Or not so much? We could even vote on who has the most infuriating level?

Make it a rule that whoever sends a level must also send a video him/herself completing it. How far are they willing to go?

This sounds like a plan. I would endorse this.

Stealthy71088 wrote: ↑Tue Oct 03, 2017 13:11 Actually I really like his levels. Yeah, many are crazy hard but they're oftentimes really clever and creative. Now I have played ancient levelpacks made by other people that were stupid hard, had 80 billion points and were completely unimaginative. I don't really like long hallways filled with teddy bears because it defeats the purpose of having points.

You like his levels when he's trying to make them GOOD. Now,take that talent, that ability to make you scream with frustration even as you push forward... and turn it to the dark. How bad could he be? How unfair a level could he build that still leads you on,hour after hour?

The idea really wouldn't be to make a level that people would want to play, or one that would just be challenging. I'm more thinking levels that keeners would play at the seventh level of Hell. Beatable, but super unpleasant, annoying and unfair. Having people submit videos of them playing through their levels would encourage creative evilness that Mortimer would approve of, instead of something that was either overly simple or outright impossible.