MagiMaster's Simple Puzzles

Rubicon Forum

Welcome, Guest. Please login or register.
January 08, 2009, 11:19:51 PM

Forum users can use the crate icon (or [level] and [/level] tags) when writing a post, to make a direct link to a level.

3984 Posts in 202 Topics by 2408 Members
Latest Member: Aryn

Here is my solution to DYTADAM: kebuneb. (This solution also works for HYBYXEX.)

Here are my solutions:
Tiny 2 & Tiny 2b: tirekox.
Tiny 3: cykicos.
Tiny 3b: bumybyz

Here's two new Passing Train puzzles. They're a little different that the others. FYSOXIH and DOMEDYG.

Here's my solution to Passing Trains 5: vylobad.

my solution to fysoxih:

rugoxyt

You're gonna hate me Magimaster: fovohub

(you should force two more spots blank)
Solutions to tinies: lonafad, pubonyh

Try this tiny: bovuzat
and this trains: nicityz

Here's a different delayed trains: hidudud I wanted to greatly increase the difference time between the different gates opening, so I slowed down the RNG.

my solution to passing trains 5: giteden

Yeah. There should be two more forced blanks in that level then. Cheesy

(And BOVUZAT and NICITYZ are pretty cool.)

Unfortunately, HIDUDUD still causes the trains to leave at about the same time fairly often. It's due to the distribution used. Both trains leave at the first 0, which is a geometric distribution. Looking up the mean and variance, the trains will, on average, leave on the 16th barrel and something like 2/3 of the trains will leave withing the first 32 barrels. To make them less likely to leave near each other, a two digit counter with random initial digits would work better. (That'd be a uniform distribution. On average they'd leave after 128 barrels and 2/3 within +/-73 of that. [Not exactly, but close enough.])

Here is my solution to Passing Trains 4: hanuzax. I keep the trains intact with their original boxcars, in order, for the duration of the trips, and track switching is controlled by train position, completely independent of departure time. Enjoy!

the solution for nicityz kubybix

I solved the trains in the middle here: dagovys I even made all four train cross paths at the exact same spot.

Quote from: MagiMaster on October 30, 2006, 06:52:47 AM

Yeah. There should be two more forced blanks in that level then. Cheesy

(And BOVUZAT and NICITYZ are pretty cool.)

Unfortunately, HIDUHUD still causes the trains to leave at about the same time fairly often. It's due to the distribution used. Both trains leave at the first 0, which is a geometric distribution. Looking up the mean and variance, the trains will, on average, leave on the 16th barrel and something like 2/3 of the trains will leave withing the first 32 barrels. To make them less likely to leave near each other, a two digit counter with random initial digits would work better. (That'd be a uniform distribution. On average they'd leave after 128 barrels and 2/3 within +/-73 of that. [Not exactly, but close enough.])

Yeah, of course the randomness of the barrels are isn't changing. The time between each was changed. If you are generating one every turn, you're average gate opening is going to be at the 16th turn, with a majority being between the 8th and the 24th turn. Your average difference in openings will be around 16 (24 - 8)turns.

But slow down the rate of barrels to say, one every five turns, then your average opening wil be on the 80th turn (still 16 barrels) with a majority being between the 40th turn and the 120th turn. Your average difference in openings will now be 80 turns (120 - 40). That's 64 more turns of action generally possible (on average) than the other way.

If that doesn't convince you try the opposite direction. speed up the RNG to 16 barrels a turn. Most of the time, the gates will open the exact same turn. you won't have any difference at all. The speed of the RNG matters.

It should be obvious from just trying it, too. With the slowed down RNG with a flat track between trains (like here: nakenib), most of the time only one train even makes it out the gate, because the other train has already come the whole distance across. With the speedy RNG, that hardly ever happens. The difference in time between openings is much smaller.

No, I haven't changed how often the number zero is generated. I just stretched out the time between each zero (the the number of barrels between hasn't changed, and doesn't need too).

A two digit RNG isn't necessary. You adjust the tractor path length , the number of barrels dropped per tractor circuit, and the base barrel that you are measuring against. Watch the difference gate openings in this example: vagunib Each train must race to cross the finish line (four screen lengths away) before the other train even starts. And one train usually does win. Try this marathon with a speedy RNG, and the trains will meet in the middle most of the time, losing before getting half a screen length.

Yeah. I know that slowing down the barrels does make the average time between trains greater. All I'm saying is that for the number of barrels generated, a uniform distribution has a better average difference. (And you usually don't have to wait so long for the first train to leave.) Here's my example: BEGUSYT

Also, here's a new Tiny puzzle. RITINEZ

kyfodih is my take on Tiny 4.