AI Talks New Boxing Game

[Cap]

I asked Google Gemini if it could code a boxing game:

"...Well, Dave a text-based boxing simulator, like Title Bout Championship Boxing 2.5, uses stat-driven algorithms and randomized probability instead of real-time physics. You can build a similar engine quickly using Python, Object-Oriented Programming (OOP), and random number generation.

1. The Core Architecture (Python)You can build a lightweight simulator that defines fighter traits (like punching power, defense, and stamina), simulates a 12-round bout, and resolves damage mathematically. Here is an example to get you started: python import random
class Boxer:
def __init__(self, name, power, defense, stamina):
self.name = name
self.power = power # Out of 100
self.defense = defense # Out of 100
self.stamina = stamina # Out of 100
self.health = 100

def simulate_punch(attacker, defender):
# Calculate success probability
hit_chance = 50 + (attacker.power - defender.defense)
hit_chance = max(10, min(90, hit_chance)) # Cap between 10-90%

if random.randint(1, 100) <= hit_chance:
damage = random.randint(5, 15) * (attacker.power / 50)
defender.health -= damage
print(f"{attacker.name} lands a punch! {defender.name} takes {damage:.1f} damage.")
else:
print(f"{attacker.name} misses!")

def fight_round(boxer1, boxer2):
print("\n--- New Round ---")
# Simulate 20 punch exchanges per round
for _ in range(20):
if boxer1.health <= 0 or boxer2.health <= 0:
break

# Alternate turns
if random.random() > 0.5:
simulate_punch(boxer1, boxer2)
else:
simulate_punch(boxer2, boxer1)

# Initialize Boxers
fighter1 = Boxer("Fighter A", power=75, defense=60, stamina=90)
fighter2 = Boxer("Fighter B", power=65, defense=80, stamina=85)

# Simulate 3 rounds
for r in range(3):
fight_round(fighter1, fighter2)
Use code with caution.

2. Mechanics to Expand Your Game To match the depth of Title Bout, you will want to expand your simulation logic:

Fighter Cards: Build a database (like JSON or SQLite) storing individual boxer traits (e.g., inside fighter, outside fighter, KO percentage).

Round Control: Implement a 'control' value that fluctuates based on who is landing more hits or pushing the pace.

Action Tables: Instead of simple math, build a 100-number table where different number ranges (e.g., 1-15 = jab, 16-30 = hook) represent different in-ring actions pulled from a boxer's specific rating card.

Status Checks: At the end of every round, roll for cuts, knockdowns, and referee stoppages based on current health levels.

Cap

[hamed2]

Interesting, my brother worked with a guy trying to learn Python for some business applications - he said it wasn't easy to learn.

[Rob Morris]

I have noticed that with the rise of AI there are a lot more sports management games, either being released or in development. If you are on Reddit then you can follow the sub gm games and you will see what I mean. Some of them have been termed "AI Slop" (not by me I might add).

There are a few boxing games (Corner King, The Comeback King and Heavyweight Boxing Champion) which are all available on Steam but are all either demo's or early access at the moment.

[Antonin]

Originally Posted by Rob Morris

I have noticed that with the rise of AI there are a lot more sports management games, either being released or in development. If you are on Reddit then you can follow the sub gm games and you will see what I mean. Some of them have been termed "AI Slop" (not by me I might add)...

I keep saying that AI has made people lazy and stupid, at least where literature and the visual arts are concerned. I know many, if not most, people these days disagree. But that's ok. I'm old.

I made a decent living for 40+ years doing writing and editing type jobs in the corporate sector. The skills I worked so hard to develop are obsolete now, I know.

[zyberianwarrior]

i just spent 4 days redoing a spreadhseet could of bene sonner if AI had an actualk BRAIN. wehn you do stuf fliek thsi you ahev to be VERY SPECIFIC with yoru questiins and screen shot images works as well

[WeiLo]

When it comes to large language models, "AI" does not stand for "Artificial Intelligence"

It stands for "Autocomplete Intelligence"

[zyberianwarrior]

agree WeiLo.

[zyberianwarrior]

you know with a little bit tweaking on the post fight scoring rules this spreadheet i worked on it CAN be used for TBCB (right now its OVR 12 to 1 while TBCB has 15-0 on its OVR if you folks are interested let me know it CAN work. And yes for those that think its sounds familiar it is the Rey-Hady spreadsheet used for Title Bout 2.

its done in Libre Office not 100% sure if its excel friendly but it can work i just looked at TBCB again it CAN and IT cant work. What can work is the BASIC function of it but what cant work is the special macro and pivot table that is heavile relied upon in this spreadsheet so it will be problematci for it to work in excel and i dont have excel period.

Goiod news is Libre office is FREE. Again if you are intersted let me know i'm still a little bit recovering form that 4 day marathon of getting all of the special stuff to work. But tweaking the rules to TBCB CAN work.

[prismark]

Hi - I took this further over the weekend and coded something up - send me a DM with e-mail if want a working prototype to provide feedback. I'm about 10 hrs in and I think its pretty good.

[zyberianwarrior]

not bad really just ran 26 bouts in about 10 minutes 26-0 (26 ko's) easy to pick opponents this is more career mode but not a bad start