stakepd - pandas for betting and prediction markets

#

stakepd is a Python package that extends pandas with specialized functions for betting/prediction market analytics. It provides custom DataFrame and Series methods as well as utilities for modeling, analyzing, and optimizing sports bets (or betting of any kind).

Features

#

• Betting-specific DataFrame and Series extensions
• Odds conversion and normalization
• Betting optimization functionality
• Integration with pandas workflows

Contents

#

1. Installation
2. Usage Examples
3. Documentation
   • pandas.Series.implied_probability
   • pandas.DataFrame.implied_probability
   • pandas.Series.odds
   • pandas.DataFrame.odds
   • pandas.Series.convert
   • pandas.DataFrame.convert
   • pandas.Series.as_fraction
   • pandas.Series.as_float
   • pandas.Series.overround
   • pandas.DataFrame.overround
   • pandas.Series.vig
   • pandas.DataFrame.vig
   • pandas.Series.devig
   • pandas.DataFrame.devig
   • stakepd.payout
   • stakepd.profit
   • stakepd.kelly_criterion
4. Contact

Installation

#

Terminal:

pip install stakepd

Jupyter:

!pip install stakepd

Usage Examples

#

Example 1:

import pandas as pd
import stakepd as spd

# Example: Calculate implied probabilities from odds
odds = pd.Series([2.0, 3.5, 1.8]) # A pandas Series of decimal odds
implied_probability = odds.implied_probability() # Some stakepd methods work directly on pandas objects
print(implied_probability)

Output:
0    0.500000
1    0.285714
2    0.555556
dtype: float64

Example 2:

import pandas as pd
import stakepd as spd

# Example: Calculate the potential profit from a Series of bets
stakes = pd.Series([100, 50, 10]) # Series of betting stakes
odds = pd.Series([100, -200, 300]) # Series of American betting odds corresponding to stakes
payout = spd.profit(stakes, odds, 'american') # Some stakepd methods are called from spd module
print(payout)

Output:
0    100.0
1     25.0
2     30.0
dtype: float64

Documentation

#

pandas.Series.implied_probability

#

pandas.Series.implied_probability(odds_type = 'decimal')

Calculate the implied probability from a series of odds.

Parameters:

• odds_type (str, optional): Type of odds provided. Supported values are 'decimal', 'fractional', and 'american'. Default is 'decimal'.

Returns:

• pandas.Series: Series of implied probabilities as decimals (float).

Example:

odds = pd.Series([2.0, 3.5, 1.8])
implied_probabilities = odds.implied_probability()
print(implied_probabilities)

Output:
0    0.500000
1    0.285714
2    0.555556
dtype: float64

Example (fractional odds):

odds = pd.Series([1/4, 2/1, 10/2])
implied_probabilities = odds.implied_probability(odds_type = 'fractional')
print(implied_probabilities)

Output:
0    0.800000
1    0.333333
2    0.166667
dtype: float64

Example (american odds):

odds = pd.Series([100, -150, 200])
implied_probabilities = odds.implied_probability(odds_type = 'american')
print(implied_probabilities)

Output:
0    0.500000
1    0.600000
2    0.333333
dtype: float64

pandas.DataFrame.implied_probability

#

pandas.DataFrame.implied_probability(odds_type = 'decimal')

Calculate the implied probability from a DataFrame of odds.

Parameters:

• odds_type (str, optional): Type of odds provided. Supported values are 'decimal', 'fractional', and 'american'. Default is 'decimal'.

Returns:

• pandas.DataFrame: DataFrame of implied probabilities as decimals (float).

Example:

odds = pd.DataFrame({"A": [2.0, 1.5, 5.0], "B": [50, 8.0, 25.0]})
implied_probabilities = odds.implied_probability()
print(implied_probabilities)

Output:
   A_implied_probability  B_implied_probability
0               0.500000                  0.020
1               0.666667                  0.125
2               0.200000                  0.040

pandas.Series.odds

#

pandas.Series.odds(odds_type = 'decimal')

Convert implied probability to the specified odds type.

Parameters:

• odds_type (str, optional): Type of odds provided. Supported values are 'decimal', 'fractional', and 'american'. Default is 'decimal'.

Returns:

• pandas.Series: Series of float odds in odds_type format.

Example:

prob = pd.Series([0.5, 0.4])
odds_vals = prob.odds(odds_type = 'decimal')
print(odds_vals)

Output:
0    2.000000
1    2.500000
dtype: float64

pandas.DataFrame.odds

#

pandas.DataFrame.odds(odds_type : str = 'decimal')

Convert implied probability to the specified odds type.

Parameters:

• odds_type (str, optional): Type of odds provided. Supported values are 'decimal', 'fractional', and 'american'. Default is 'decimal'.

Returns:

• pandas.DataFrame: DataFrame of odds in odds_type format.

Example:

implied_probabilities = pd.DataFrame({"A": [0.5, 0.4], "B": [0.3, 0.2]})
odds_df = implied_probabilities.odds(odds_type = "decimal")
print(odds_df)

Output: 
   A_decimal_odds  B_decimal_odds
0             2.0        3.333333
1             2.5        5.000000

pandas.Series.convert

#

pandas.Series.convert(from_type, to_type)

Convert between types of odds or implied_probability

Parameters:

• from_type (str): input type. Supported values are 'decimal', 'fractional', 'american' and 'implied_probability'.
• to_type (str): output type. Supported values are 'decimal', 'fractional', 'american' and 'implied_probability'.

Returns:

• pandas.Series: Series of floating point odds/implied probabilities in to_type format.

Example:

decimal_odds = pd.Series([2.0, 1.5, 3.0])
american_odds = decimal_odds.convert(from_type = 'decimal', to_type = 'american')
print(american_odds)

Output:
0    100.0000
1    -200.0000
2    200.0000
dtype: float64

pandas.DataFrame.convert

#

pandas.DataFrame.convert(from_type, to_type)

Convert between types of odds or implied_probability

Parameters:

• from_type (str): input type. Supported values are 'decimal', 'fractional', 'american' and 'implied_probability'.
• to_type (str): output type. Supported values are 'decimal', 'fractional', 'american' and 'implied_probability'.

Returns:

• pandas.DataFrame: DataFrame of floating point odds/implied probabilities in to_type format.

Example:

decimal_odds = pd.DataFrame({"A": [2.0, 1.5, 5.0], "B": [50, 8.0, 25.0]})
american_odds = decimal_odds.convert(from_type="decimal", to_type="american")
print(american_odds)

Output:
   A_implied_probability_american_odds  B_implied_probability_american_odds
0                                100.0                               4900.0
1                               -200.0                                700.0
2                                400.0                               2400.0

pandas.Series.as_fraction

#

pandas.Series.as_fraction()

Convert fractional odds from floating point to fraction in string form

Returns:

• pandas.Series: Series of fractional odds in string form.

Example:

fractional_odds = pd.Series([0.25, 0.5, 2.0, 1.0])
fraction_strings = fractional_odds.as_fraction()
print(fraction_strings)

Output:
0    1/4
1    1/2
2    2/1
3    1/1
dtype: object

pandas.Series.as_float

#

pandas.Series.as_float()

Convert fractional odds from fraction strings to floating point. Reverse of pandas.Series.as_fraction().

Returns:

• pandas.Series: Series of fractional odds as floating point.

Example:

fractional_strings = pd.Series(["1/4", "1/2", "2/1"])
fraction_odds = fractional_strings.as_float()
print(fraction_odds)

Output:
0    0.25
1    0.50
2    2.00
dtype: float64

pandas.Series.overround

#

pandas.Series.overround(odds_type = 'implied_probability')

Calculate the overround of a Series of odds

Parameters:

• odds_type (str, optional): Type of odds provided. Supported values are 'decimal', 'fractional', 'american', and 'implied_probability'. Default is 'implied_probability'.

Returns:

• float: floating point overround as decimal

Example:

implied_odds = pd.Series([0.8, 0.5, 0.3])
overround_value = implied_odds.overround(odds_type = 'implied_probability')
print(overround_value)

Output:
np.float64(0.6000000000000000)

pandas.DataFrame.overround

#

pandas.DataFrame.overround(odds_type = 'implied_probability')

Calculate the overrounds of a DataFrame of odds

Parameters:

• odds_type (str, optional): Type of odds provided. Supported values are 'decimal', 'fractional', 'american', and 'implied_probability'. Default is 'implied_probability'.

Returns:

• pandas.Series: Series of overrounds as decimal corresponding to the indexes of the input DataFrame

Example:

implied_odds = pd.DataFrame({"A": [0.8, 0.5], "B": [0.3, 0.7]})
overround_value = implied_odds.overround(odds_type = 'implied_probability')
print(overround_value)

Output:
0    0.1
1    0.2

pandas.Series.vig

#

pandas.Series.vig(odds_type = 'implied_probability')

Calculate the vigourish or juice of a Series of odds

Parameters:

• odds_type (str, optional): Type of odds provided. Supported values are 'decimal', 'fractional', 'american', and 'implied_probability'. Default is 'implied_probability'.

Returns:

• float: floating point vigourish as decimal

Example:

decimal_odds = pd.Series([2.0, 1.5, 3.0])
vig_value = decimal_odds.vig(odds_type = 'decimal')
print(vig_value)

Output:
np.float64(0.3333333333)

pandas.DataFrame.vig

#

pandas.DataFrame.vig(odds_type = 'implied_probability')

Calculate the vigourish values or juice values of a DataFrame of odds

Parameters:

• odds_type (str, optional): Type of odds provided. Supported values are 'decimal', 'fractional', 'american', and 'implied_probability'. Default is 'implied_probability'.

Returns:

• pd.Series: Series of vigourish values corresponding to the indexes of the input Dataframe

Example:

implied_probs = pd.DataFrame({"A": [0.8, 0.5], "B": [0.3, 0.7]})
vig_values = decimal_probs.vig()
print(vig_values)

Output:
0    0.090909
1    0.166667
dtype: float64

pandas.Series.devig

#

pandas.Series.devig()

Calculate devigged implied probabilities from a Series of implied probabilities with vigrourish

Returns:

• pandas.Series: Series of devigged decimal implied probabilities as floating point values

Example:

implied_probabilities = pd.Series([0.8, 0.5, 0.3])
devigged_probs = implied_probabilities.devig()
print(devigged_probs)

Output:
0    0.5000
1    0.3125
2    0.1875
dtype: float64

pandas.DataFrame.devig

#

pandas.DataFrame.devig()

Calculate devigged implied probabilities from a DataFrame of implied probabilities with vigrourish. Input df has each row as a set of vigged implied probabilities

Returns:

• pandas.DataFrame: DataFrame of devigged decimal implied probabilities as floating point values. Each row corresponds to a row in the input DataFrame.

Example:

implied_probabilities = pd.DataFrame({
            "A": [0.8, 0.5],
            "B": [0.3, 0.7]
        })
devigged_probs = implied_probabilities.devig()
print(devigged_probs)

Output:
          A         B
0  0.727273  0.272727
1  0.416667  0.583333

stakepd.payout(stakes, odds, odds_type)

#

spd.payout(stakes, odds, odds_type)

Calculates payout from a bet or payouts from a Series of bets

Parameters:

• stakes (float, int or pandas.Series): scalar or Series of stake/bet values.
• odds (float, int or pandas.Series): scalar or Series of odds in odds_type format. Must be scalar if stakes is scalar, or must be Series if stakes is Series (must be same length)
• odds_type (str): Type of odds provided. Supported values are 'decimal', 'fractional', 'american', and 'implied_probability'

Returns:

• float: payout of bet as a floating point number OR
• pands.Series: Series of payouts with same length as stakes and odds

Example (Scalar):

stakes = 100
odds = 2
payouts = spd.payout(stakes, odds, 'decimal')
print(payouts)

Output: 
200

Example (Series):

stakes = pd.Series([100, 50, 10])
odds = pd.Series([2.0, 3.0, 1.5])
payouts = spd.payout(stakes, odds, 'decimal')
print(payouts)

Output: 
0    200.0
1    150.0
2     15.0
dtype: float64

stakepd.profit(stakes, odds, odds_type)

#

spd.profit(stakes, odds, odds_type)

Calculates the profits from a bet or a Series of bets. Profit = Payout - Stake.

Parameters:

• stakes (float, int or pandas.Series): scalar or Series of stake/bet values.
• odds (float, int or pandas.Series): scalar or Series of odds in odds_type format. Must be scalar if stakes is scalar, or must be Series if stakes is Series (must be same length)
• odds_type (str): Type of odds provided. Supported values are 'decimal', 'fractional', 'american', and 'implied_probability'

Returns:

• float: profit of bet as a floating point number OR
• pands.Series: Series of profit values with same length as stakes and odds

Example (Scalar):

stakes = 100
odds = 2
profits = spd.profit(stakes, odds, 'decimal')
print(profits)

Output: 
100

Example (Series):

stakes = pd.Series([100, 50, 10])
odds = pd.Series([2.0, 3.0, 1.5])
profits = spd.profit(stakes, odds, 'decimal')
print(profits)

Output: 
0    100.0
1    100.0
2      5.0
dtype: float64

stakepd.kelly_criterion(odds, true_probability, odds_type, bank_roll, kelly_percentage)

#

Calculate the amount to bet on an EV bet according to the kelly criterion formula.

Parameters:

• odds (float, int or pandas.Series): scalar or Series of odds in odds_type format. Must be scalar if true_probability is scalar, or must be Series if true_probability is Series (must be same length)
• true_probability (float or pandas.Series): scalar or Series of estimated ’true’ probabilities of events.
• odds_type (str): Type of odds provided. Supported values are 'decimal', 'fractional', or 'american'
• bank_roll (float, optional): The amount of bankroll available for making bets with. Default of 100.0.
• kelly_percentage (float, optional): Kelly percentage as a decimal which can be adjusted for fractional kelly bets. Default of 1.0.

Returns:

• float: amount that should be wagered according to Kelly Criterion, OR
• pands.Series: Series of kelly wager values with same length as true_probability and odds

Example:

bankroll = 100
odds = pd.Series([2.0, 3.0])
probabilities = pd.Series([0.4, 0.3])
wagers = spd.kelly_criterion(odds, probabilities, 'decimal', bankroll)
print(wagers)

Output:
result
0    0.0
1    0.0
dtype: float64

0.0 means that you shouldn’t make the bet.

Example:

bankroll = 1000
odds = pd.Series([2.0, 3.0, 1.5])
probabilities = pd.Series([0.6, 0.4, 0.7])
wagers = spd.kelly_criterion(odds, probabilities, 'decimal', bankroll)
print(wagers)

Output:
0    200.0
1    100.0
2    100.0
dtype: float64

Contact

#

For any enquiries related to ads, software bugs, or anything else, please email bhatwadekarenterprises@gmail.com.