I attempted to make new medals but I don't care if anyone uses them


Fixed:

TheMonsterOfTheDeep wrote:

I attempted to make new medals but I don't care if anyone uses them


Fixed:

I like 'em

@MATU
Your chat is probably in a separate Window. Lost mine too before.

MartinBraendli2 wrote:

bobbybee wrote:

Wait, why did we skip ATC#0?

It's Scratch, lists start at 1 here.
Before I used Scratch, I always wondered, why arrays and stuff start at 0. Being forced to use lists with first index = 1, really shows you, why in grown up languages indices usually start at 0. It makes so many formulas shorter/simpler.

Actually I was always told that Engineers count from 1, Scientists count from 0 :-) (Mind you it was my Computer Science lecturer who said this…)

Hence how you know Fortran was for and by engineers.

Spamela wins

1st place: MartinBraendli's Spamela & HugoTheHedgehog
He tells me he has a challenge idea already!

2nd place: CodeLegend's STINGerla

3rd place: TheMonsterOfTheDeep's Snipar

(Thanks TheMonsterOfTheDeep for the medals)

Full results here: https://docs.google.com/spreadsheets/d/1v6Ju2EOpOMio7FpGXbxEPvtHGt3_d9izopxMFg_2VIs/edit?usp=sharing

Regardless of which bots won, I hope everybody had some fun making these, despite all the controller bugs. Congrats on learning a new language and doing some “l33t hax”

We'll be waiting for Martin's ATC#4!

^ Yes this was very fun

@CodeLegend & TheMonsterOfTheDeep: Don't forget to add a link to this thread to the medal.

@MATU: Thanks for this great challenge. Could you please edit your first post to show rankings/medals?

MartinBraendli2 wrote:

@CodeLegend & TheMonsterOfTheDeep: Don't forget to add a link to this thread to the medal.

@MATU: Thanks for this great challenge. Could you please edit your first post to show rankings/medals?

Wait, now you're creating multiple ATCs?

Yay good job everyone!

bobbybee wrote:

MartinBraendli2 wrote:

@CodeLegend & TheMonsterOfTheDeep: Don't forget to add a link to this thread to the medal.

@MATU: Thanks for this great challenge. Could you please edit your first post to show rankings/medals?

Wait, now you're creating multiple ATCs?

What do you mean by that? Yes, there are multiple challenges, but always only one open challenge, created by the winner of the last one.

Number 4 Number 4 Number 4!

Repost (just to make sure, that followers of this thread don't miss ATC#4)

ATC#4: Tic Tac Toe will be released in 24h (~11:00 PM in forum-time). It will be a coding puzzle where the first correct submission determines the winner. While the rules might look similar to previous challenges, I can guarantee you, that the solution looks nothing like them (so don't worry, no recursion involved).

In my opinion, ATC#4 will be harder than ATC#1 and #2. It will involve lateral thinking, knowledge of Scratch and other programming languages and deep knowledge of Tic Tac Toe (Yes, you can still learn something about Tic Tac Toe, at least I did, while trying to solve the puzzle myself). I also found solving this puzzle to be quite rewarding and I hope, you will feel the same way.

Should you manage to crack ATC#4 within 4h, I will add “(yourName) is probably the best Scratcher I know” to my signature.

ATC#4 is out
from now on, the ATC Overview thread will be used for such announcements, so make sure you follow it.

I know it's closed, but I dug up this topic and tried to make a new bot. This bot does pretty well in 1v1 against InfiniteLoop, but gets destroyed by CopyBot:
XMorph
“I can't lie to you about your chances, but you have my sympathies.”
Revision 5:

Be interesting to see a genetic algorithm used here.
Seems that some evolving bots just might be feasible.

ilikelegos wrote:

Be interesting to see a genetic algorithm used here.
Seems that some evolving bots just might be feasible.

Perhaps we could take away the front-end, remove setIntervals and run it in node.js to make the engine go faster?

iamunknown2 wrote:

ilikelegos wrote:

Be interesting to see a genetic algorithm used here.
Seems that some evolving bots just might be feasible.

Perhaps we could take away the front-end, remove setIntervals and run it in node.js to make the engine go faster?

Hmm… I don't know much about node.js
Was just about to get into it for multiplayer stuff.

ilikelegos wrote:

iamunknown2 wrote:

ilikelegos wrote:

Be interesting to see a genetic algorithm used here.
Seems that some evolving bots just might be feasible.

Perhaps we could take away the front-end, remove setIntervals and run it in node.js to make the engine go faster?

Hmm… I don't know much about node.js
Was just about to get into it for multiplayer stuff.

When Martin Brändli did his golf puzzle, I cloned his testbed in C so we could experiment with different algorithms at speed…

http://gtoal.com/scratch/constraint_solver.c.html

Maybe that approach would work for you if you don't want to use javascript?

G

gtoal wrote:

ilikelegos wrote:

iamunknown2 wrote:

ilikelegos wrote:

Be interesting to see a genetic algorithm used here.
Seems that some evolving bots just might be feasible.

Perhaps we could take away the front-end, remove setIntervals and run it in node.js to make the engine go faster?

Hmm… I don't know much about node.js
Was just about to get into it for multiplayer stuff.

When Martin Brändli did his golf puzzle, I cloned his testbed in C so we could experiment with different algorithms at speed…

http://gtoal.com/scratch/constraint_solver.c.html

Maybe that approach would work for you if you don't want to use javascript?

G

Oooh… neat!

5000 rounds:

Snipar 495EX
STINGerla 118EX
SPAMerla 379EX
HugoTheHedgehog 846EX
RunRunProtect 54EX
WonkThePlank 136EX
IcyFireBot 43EX
BotyMcBotFace EX
Stinger 424EX
TTAIB 13EX
RunRunRun 33EX
TestBot 25EX
InfiniteLoop 153EX
RSH EX
PORK 10EX
CopyBot 63EX
Iamunknown 345EX