Re: An attempt to revive Robocode

------=_Part_565232_1611661196.1777004862699
Content-Type: multipart/alternative; 
	boundary="----=_Part_565233_1993817199.1777004862699"

------=_Part_565233_1993817199.1777004862699
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hi Billy,
I've spent a few days recently building=20
this https://github.com/SergiiRedchuk/roboclod then when I finished, I=20
found this group and your post :-)
My AI experiment is different from yours though. Instead of chatting with=
=20
AI using prompts I created agentic optimization loop where Claude Code=20
reads the API and battle evaluation results, figures out code improvements=
=20
itself, runs evaluators in headless mode again, compares new metrics with=
=20
previous, promotes or reverts the changes, and suggests its next steps.=20
Then rinse and repeat.
Here's a few slides showing what it=20
does https://github.com/SergiiRedchuk/roboclod/blob/main/doc/RoboclodSlides=
.pdf
Enjoy :-)

On Monday, January 5, 2026 at 6:38:15=E2=80=AFAM UTC-5 Billy Jones wrote:

> Hi everyone,
>
> I am a longtime competitive programmer but have just recently discovered=
=20
> Robocode which is honestly way more fun than the generic competitions I a=
m=20
> used to. However, I was surprised by the lack of active online competitio=
ns=20
> (besides https://literumble.appspot.com which I think is a great and=20
> generous contribution to the community).
>
> After a bit of thinking and tinkering with the Robocode "CLI", I decided=
=20
> to build an Arena <https://robocode-arena.com/> with a continously=20
> running ladder of robots. I have been experimenting with it for a month o=
r=20
> so now and thought I'd share it with everyone here in case anybody is sti=
ll=20
> Robocoding or wants to pick it up again.
>
> I also figured it would be fun to see what the current LLMs are capable o=
f=20
> when it comes to Robocode. So, I ran the following experiment=20
> https://robocode-arena.com/blog/ai-experiment which some of you may find=
=20
> interesting.
>
> Thanks,
> Billy

--=20
You received this message because you are subscribed to the Google Groups "=
Robocode" group.
To unsubscribe from this group and stop receiving emails from it, send an e=
mail to [email protected].
To view this discussion visit https://groups.google.com/d/msgid/robocode/1f=
4c0d84-629f-4df8-864f-62561c1ca160n%40googlegroups.com.

------=_Part_565233_1993817199.1777004862699
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hi Billy,<div>I've spent a few days recently building this=C2=A0https://git=
hub.com/SergiiRedchuk/roboclod then when I finished, I found this group and=
 your post :-)</div><div>My AI experiment is different from yours though. I=
nstead of chatting with AI using prompts I created agentic optimization loo=
p where Claude Code reads the API and battle evaluation results, figures ou=
t code improvements itself, runs evaluators in headless mode again, compare=
s new metrics with previous, promotes or reverts the changes, and suggests =
its next steps. Then rinse and repeat.</div><div>Here's a few slides showin=
g what it does=C2=A0https://github.com/SergiiRedchuk/roboclod/blob/main/doc=
/RoboclodSlides.pdf</div><div>Enjoy :-)<br /><br /></div><div class=3D"gmai=
l_quote"><div dir=3D"auto" class=3D"gmail_attr">On Monday, January 5, 2026 =
at 6:38:15=E2=80=AFAM UTC-5 Billy Jones wrote:<br/></div><blockquote class=
=3D"gmail_quote" style=3D"margin: 0 0 0 0.8ex; border-left: 1px solid rgb(2=
04, 204, 204); padding-left: 1ex;">Hi everyone,<br><br>I am a longtime comp=
etitive programmer but have just recently discovered Robocode which is hone=
stly way more fun than the generic competitions I am used to. However, I wa=
s surprised by the lack of active online competitions (besides=C2=A0<a href=
=3D"https://literumble.appspot.com" target=3D"_blank" rel=3D"nofollow" data=
-saferedirecturl=3D"https://www.google.com/url?hl=3Den&amp;q=3Dhttps://lite=
rumble.appspot.com&amp;source=3Dgmail&amp;ust=3D1777090458727000&amp;usg=3D=
AOvVaw2SeGrloZMTogS2jwIYAeOW">https://literumble.appspot.com</a> which I th=
ink is a great and generous contribution to the community).<br><br>After a =
bit of thinking and tinkering with the Robocode &quot;CLI&quot;, I decided =
to build an <a href=3D"https://robocode-arena.com/" target=3D"_blank" rel=
=3D"nofollow" data-saferedirecturl=3D"https://www.google.com/url?hl=3Den&am=
p;q=3Dhttps://robocode-arena.com/&amp;source=3Dgmail&amp;ust=3D177709045872=
7000&amp;usg=3DAOvVaw1abxZqln6s5p_5JKO1f5uy">Arena</a> with a continously r=
unning ladder of robots. I have been experimenting with it for a month or s=
o now and thought I&#39;d share it with everyone here in case anybody is st=
ill Robocoding or wants to pick it up again.<br><br>I also figured it would=
 be fun to see what the current LLMs are capable of when it comes to Roboco=
de. So, I ran the following experiment=C2=A0<a href=3D"https://robocode-are=
na.com/blog/ai-experiment" target=3D"_blank" rel=3D"nofollow" data-saferedi=
recturl=3D"https://www.google.com/url?hl=3Den&amp;q=3Dhttps://robocode-aren=
a.com/blog/ai-experiment&amp;source=3Dgmail&amp;ust=3D1777090458727000&amp;=
usg=3DAOvVaw0JY4Ug2FZ94MQGkyN8x-Oi">https://robocode-arena.com/blog/ai-expe=
riment</a>=C2=A0which some of you may find interesting.<br><br>Thanks,<br>B=
illy</blockquote></div>

<p></p>

-- <br />
You received this message because you are subscribed to the Google Groups &=
quot;Robocode&quot; group.<br />
To unsubscribe from this group and stop receiving emails from it, send an e=
mail to <a href=3D"mailto:[email protected]">robocode+u=
[email protected]</a>.<br />
To view this discussion visit <a href=3D"https://groups.google.com/d/msgid/=
robocode/1f4c0d84-629f-4df8-864f-62561c1ca160n%40googlegroups.com?utm_mediu=
m=3Demail&utm_source=3Dfooter">https://groups.google.com/d/msgid/robocode/1=
f4c0d84-629f-4df8-864f-62561c1ca160n%40googlegroups.com</a>.<br />

------=_Part_565233_1993817199.1777004862699--

------=_Part_565232_1611661196.1777004862699--