ETWKBC Code v1

import urllib.request
import json
import time
import math

BASE_URL = "https://manifold.markets/api"
HEADERS = {
    # "Authorization": f"Key {API_KEY}",
    "Content-Type": "application/json"
}

KBC_SLUG = "exponential-time-weighted-keynesian"

# Fetch market info
def get_market(slug):
    req = urllib.request.Request(f"{BASE_URL}/v0/slug/{slug}", headers=HEADERS)
    with urllib.request.urlopen(req) as response:
        return json.load(response)

def get_current_weighted_sum(mkt):
    weighted_sum = 0
    for a in mkt['answers']:
        weighted_sum += a['probability'] * int(a['text'])

    return weighted_sum

# Fetch all bets for market, paginating through recursively until nothing comes back
def get_market_bets(slug, before_id=None):
    url = f"{BASE_URL}/v0/bets?contractSlug={slug}"
    if before_id is not None:
        url += f"&before={before_id}"

    try:
        req = urllib.request.Request(url, headers=HEADERS)
    except e:
        print(f"HTTP Error code: {e.code}")
        print(f"Reason: {e.reason}")
        print(f"Headers: {e.hdrs}")
        print(f"Error content: {e.fp.read()}")
        raise e

    with urllib.request.urlopen(req) as response:
        bets = json.load(response)

    # base case, just return
    if len(bets) == 0:
        return []

    return bets + get_market_bets(slug, bets[-1]["id"])


bets = get_market_bets(KBC_SLUG)

print(f"Total Bets: {len(bets)}")
market_info = get_market(KBC_SLUG)

current_weighted_sum = get_current_weighted_sum(market_info)
print(f"Current Weighted Sum: {current_weighted_sum}")

market_creation_time = market_info["createdTime"]
current_or_close_time = min(market_info["closeTime"], round(time.time() * 1000))
total_time = current_or_close_time - market_creation_time

for a in market_info["answers"]:
    # Bets for that answer (including implicit ones from bets on other answers)
    answer_bets = sorted(
        [b for b in bets if b["answerId"] == a["id"]], key=lambda a: a["createdTime"]
    )
    a['num_bets'] = len(answer_bets)
    # Numerical value of the answer
    answer_val = int(a["text"])

    # val * prob * duration for each bet
    weighted_sum = (
        answer_val
        * answer_bets[0]["probBefore"]
        * (answer_bets[0]["createdTime"] - market_creation_time)
    )

    for i, b in enumerate(answer_bets):
        bet_duration = (
            answer_bets[i + 1]["createdTime"] - b["createdTime"]
            if i + 1 < len(answer_bets)
            else current_or_close_time - b["createdTime"]
        )
        weighted_sum += answer_val * b["probAfter"] * bet_duration

    a["weighted_sum"] = weighted_sum

# Sum all the answer's weighted sums, divide by total market time
final_sum = sum(a["weighted_sum"] for a in market_info["answers"]) / total_time

# Divide by 2
final_answer = final_sum / 2

print("")
print(f"Time Weighted Sum: {final_sum}")
print(f"Final Answer: {final_answer}")

# Find linear weighting for resolution

# Find the closest options
lower = 0
upper = 99999999999999999999
exact = None

for a in market_info["answers"]:
    val = int(a["text"])
    if val < final_answer and val > lower:
        lower = val
    if val > final_answer and val < upper:
        upper = val
    if val == final_answer:
        exact = val

if exact is not None:
    print(f"Resolution: 100% {exact}")
else:
    lower_percent = int(round((upper - final_answer) / (upper - lower) * 100))
    upper_percent = 100 - lower_percent
    print(f"Resolution: {lower_percent}% {lower}, {upper_percent}% {upper}")