Lloyd Danzig A Modern Love Story: Machine Learning Engines & - - PowerPoint PPT Presentation

lloyd danzig
SMART_READER_LITE
LIVE PREVIEW

Lloyd Danzig A Modern Love Story: Machine Learning Engines & - - PowerPoint PPT Presentation

Lloyd Danzig A Modern Love Story: Machine Learning Engines & The Global Sports Betting Industry SHARP ALPHA ADVISORS AGENDA State of the 01 Industry Revenue 02 Models Predictive 03 Analytics Next Gen 04 Statistics Intro to


slide-1
SLIDE 1

A Modern Love Story:

Machine Learning Engines & The Global Sports Betting Industry

Lloyd Danzig

SHARP ALPHA ADVISORS

slide-2
SLIDE 2

01

State of the Industry

AGENDA

02

Revenue Models

Intro to Machine Learning

03

Predictive Analytics

04

Next Gen Statistics

05

Machine Learning Use Cases

06

Questions?

slide-3
SLIDE 3
slide-4
SLIDE 4

State of the Sports Betting Industry

slide-5
SLIDE 5

Green Live, Legal Sports Betting (13 States) Light Green Legal Sports Betting, Not Yet Operational (6 States + DC) Blue Active 2019 Sports Betting Legislation (5 States) Light Blue Dead Sports Betting Legislation in 2019 (19 States) Gray No Sports Betting Bills in 2019 (8 States)

Source: AGA As of: November 7, 2019

U.S. Legalization Map

slide-6
SLIDE 6

► Fans are projected to wager $30 billion on Esports in 2020 ► Sportsbook operators would generate over $2 billion in GGR ► Challenges: lack of reliable data, pricing difficulties, and cheating

Future Trends Betting on Esports

slide-7
SLIDE 7

Future Trends Sports Betting Bots

► Sophisticated forecasting models ► Convert event probabilities into

prices

► Look for differences in model

price and market price

► Seek out arbitrage opportunities

slide-8
SLIDE 8

► “Provably Fair” gaming ► Guaranteed, instantaneous

payouts via smart contracts

► Streamlined, real-time financial

auditing

Future Trends

Blockchain Sportsbooks

slide-9
SLIDE 9

Revenue Models

slide-10
SLIDE 10

1 2 3a 3b

Revenue Model: Sportsbook

Sportsbook operators have to manage risk and set prices/odds proficiently.

Customers view odds set by sportsbook

+190

  • 225

+5.5

(-110)

  • 5.5

(-110)

NEW YORK KNICKS DETROIT PISTONS

slide-11
SLIDE 11

1 2 3a 3b

Alice thinks New York will win Bob thinks New York will lose

Alice risks $100 to win $190 Bob risks $225 to win $100 $225 $100

+$325

Revenue Model: Sportsbook

Sportsbook operators have to manage risk and set prices/odds proficiently.

Customers view odds set by sportsbook

+190

  • 225

+5.5

(-110)

  • 5.5

(-110)

NEW YORK KNICKS DETROIT PISTONS

slide-12
SLIDE 12

1 2 3a 3b

$290

New York wins. Sportsbook returns Alice’s $100 plus $190 winnings

Profit = $325-$290 = $35

Alice thinks New York will win Bob thinks New York will lose

Alice risks $100 to win $190 Bob risks $225 to win $100 $225 $100

+$325

Revenue Model: Sportsbook

Sportsbook operators have to manage risk and set prices/odds proficiently.

Customers view odds set by sportsbook

+190

  • 225

+5.5

(-110)

  • 5.5

(-110)

NEW YORK KNICKS DETROIT PISTONS

slide-13
SLIDE 13

1 2 3a 3b

$290

New York wins. Sportsbook returns Alice’s $100 plus $190 winnings

Profit = $325-$290 = $35

$325

New York loses. Sportsbook returns Bob’s $225 plus $100 winnings

Profit = $325-$325 = $0

Alice thinks New York will win Bob thinks New York will lose

Alice risks $100 to win $190 Bob risks $225 to win $100 $225 $100

+$325

Revenue Model: Sportsbook

Sportsbook operators have to manage risk and set prices/odds proficiently.

Customers view odds set by sportsbook

+190

  • 225

+5.5

(-110)

  • 5.5

(-110)

NEW YORK KNICKS DETROIT PISTONS

slide-14
SLIDE 14

1 2 3a 3b

Alice thinks New York has a 33% chance of winning, represented in fair odds as +203.

Revenue Model: Betting Exchange

NEW YORK KNICKS DETROIT PISTONS +190

  • 225

+5.5

(-110)

  • 5.5

(-110)

Sportsbook Odds:

Exchanges offer a number

  • f dramatic advantages
  • ver sportsbooks, most

notably in the form of drastically improved odds.

slide-15
SLIDE 15

1 2 3a 3b

Alice thinks New York has a 33% chance of winning, represented in fair odds as +203. She offers to accept a wager from anyone interested in Detroit -203 (to win $100).

Revenue Model: Betting Exchange

NEW YORK KNICKS DETROIT PISTONS +190

  • 225

+5.5

(-110)

  • 5.5

(-110)

Sportsbook Odds:

Exchanges offer a number

  • f dramatic advantages
  • ver sportsbooks, most

notably in the form of drastically improved odds.

slide-16
SLIDE 16

1 2 3a 3b

Alice thinks New York has a 33% chance of winning, represented in fair odds as +203. She offers to accept a wager from anyone interested in Detroit -203 (to win $100).

Revenue Model: Betting Exchange

NEW YORK KNICKS DETROIT PISTONS +190

  • 225

+5.5

(-110)

  • 5.5

(-110)

Sportsbook Odds:

The best sportsbook is offering Detroit -225, so Bob accepts the other side of Alice’s wager.

Exchanges offer a number

  • f dramatic advantages
  • ver sportsbooks, most

notably in the form of drastically improved odds.

slide-17
SLIDE 17

1 2 3a 3b

Alice thinks New York has a 33% chance of winning, represented in fair odds as +203. She offers to accept a wager from anyone interested in Detroit -203 (to win $100).

New York wins. Bob pays Alice $203, a small percentage of which goes to the exchange. Operator Profit = $10.15 $203

$193 $10

Exchange

Revenue Model: Betting Exchange

NEW YORK KNICKS DETROIT PISTONS +190

  • 225

+5.5

(-110)

  • 5.5

(-110)

Sportsbook Odds:

The best sportsbook is offering Detroit -225, so Bob accepts the other side of Alice’s wager.

Exchanges offer a number

  • f dramatic advantages
  • ver sportsbooks, most

notably in the form of drastically improved odds.

slide-18
SLIDE 18

1 2 3a 3b

Alice thinks New York has a 33% chance of winning, represented in fair odds as +203. She offers to accept a wager from anyone interested in Detroit -203 (to win $100).

New York wins. Bob pays Alice $203, a small percentage of which goes to the exchange. Operator Profit = $10.15 $203

$193 $10

Exchange

New York loses. Alice pays Bob $100, a small percentage of which goes to the exchange. Operator Profit = $5.00 $100

$95 $5

Exchange

Revenue Model: Betting Exchange

NEW YORK KNICKS DETROIT PISTONS +190

  • 225

+5.5

(-110)

  • 5.5

(-110)

Sportsbook Odds:

The best sportsbook is offering Detroit -225, so Bob accepts the other side of Alice’s wager.

Exchanges offer a number

  • f dramatic advantages
  • ver sportsbooks, most

notably in the form of drastically improved odds.

slide-19
SLIDE 19

Revenue Model: Customer Perspective

Results of Winning $100 Wager

Sportsbook Exchange

$190.00 $193.00 $44.44 $46.80

Alice Bob

slide-20
SLIDE 20

Revenue Model: Customer Perspective

Results of Winning $100 Wager

Sportsbook Exchange

$190.00 $193.00 $44.44 $46.80

Alice Bob

slide-21
SLIDE 21

Revenue Model: Customer Perspective

Results of Winning $100 Wager

Sportsbook Exchange

$190.00 $193.00 $44.44 $46.80

Alice Bob

Ultimately, all parties are better off having used the exchange.

slide-22
SLIDE 22

Revenue Model: Customer Perspective

Results of Winning $100 Wager

Sportsbook Exchange

$190.00 $193.00 $44.44 $46.80

Alice Bob

Amount Risked to win $100

Sportsbook Exchange

$52.63 $51.85 $225.00 $213.68

Alice Bob

Ultimately, all parties are better off having used the exchange.

slide-23
SLIDE 23

Revenue Model: Customer Perspective

Results of Winning $100 Wager

Sportsbook Exchange

$190.00 $193.00 $44.44 $46.80

Alice Bob

Amount Risked to win $100

Sportsbook Exchange

$52.63 $51.85 $225.00 $213.68

Alice Bob

Ultimately, all parties are better off having used the exchange.

slide-24
SLIDE 24

Revenue Model: Customer Perspective

Results of Winning $100 Wager

Sportsbook Exchange

$190.00 $193.00 $44.44 $46.80

Alice Bob

Amount Risked to win $100

Sportsbook Exchange

$52.63 $51.85 $225.00 $213.68

Alice Bob

Ultimately, all parties are better off having used the exchange.

slide-25
SLIDE 25

Predictive Analytics

slide-26
SLIDE 26

Industry Standard Monte Carlo Simulation

Monte Carlo simulation is a method for iteratively evaluating a

deterministic model using sets of nondeterministic (i.e. random) numbers as inputs.

E.g. “What is the probability of rolling a 1 during a single throw of a six-sided die?”

slide-27
SLIDE 27

Industry Standard Monte Carlo Simulation

Monte Carlo simulation is a method for iteratively evaluating a

deterministic model using sets of nondeterministic (i.e. random) numbers as inputs.

E.g. “What is the probability of rolling a 1 during a single throw of a six-sided die?”

x10000

Die # of Outcomes % of Outcomes

16648 16.65% 16521 16.52% 16910 16.91% 16539 16.54% 16843 16.84% 16540 16.54%

slide-28
SLIDE 28

Industry Standard Monte Carlo Simulation

Monte Carlo simulation is a method for iteratively evaluating a

deterministic model using sets of nondeterministic (i.e. random) numbers as inputs.

E.g. “What is the probability of rolling a 1 during a single throw of a six-sided die?”

x10000

Die # of Outcomes % of Outcomes

16648 16.65% 16521 16.52% 16910 16.91% 16539 16.54% 16843 16.84% 16540 16.54%

slide-29
SLIDE 29

Industry Standard Monte Carlo Simulation

Monte Carlo simulation is a method for iteratively evaluating a

deterministic model using sets of nondeterministic (i.e. random) numbers as inputs.

E.g. “What is the probability of rolling a 1 during a single throw of a six-sided die?”

x10000

Die # of Outcomes % of Outcomes

16648 16.65% 16521 16.52% 16910 16.91% 16539 16.54% 16843 16.84% 16540 16.54%

slide-30
SLIDE 30

Industry Standard Monte Carlo Simulation

Monte Carlo simulation is a method for iteratively evaluating a

deterministic model using sets of nondeterministic (i.e. random) numbers as inputs.

E.g. “What is the probability of rolling a 1 during a single throw of a six-sided die?”

x10000

Die # of Outcomes % of Outcomes

16648 16.65% 16521 16.52% 16910 16.91% 16539 16.54% 16843 16.84% 16540 16.54%

=RANDBETWEEN(1,6) =random.randint(1,6)

slide-31
SLIDE 31

Industry Standard Monte Carlo Simulation

slide-32
SLIDE 32

Industry Standard Monte Carlo Simulation

slide-33
SLIDE 33

Industry Standard Monte Carlo Simulation

slide-34
SLIDE 34

Industry Standard Monte Carlo Simulation

slide-35
SLIDE 35

Industry Standard Monte Carlo Simulation

slide-36
SLIDE 36

Industry Standard Monte Carlo Simulation

slide-37
SLIDE 37

Industry Standard Monte Carlo Simulation

slide-38
SLIDE 38

Industry Standard Monte Carlo Simulation

slide-39
SLIDE 39

Industry Standard Monte Carlo Simulation

slide-40
SLIDE 40

Next Gen Statistics

slide-41
SLIDE 41

Computer Vision

Explanatory Augmented Reality Competitor Overlays Viewpoint Synthesis Performance Analysis

slide-42
SLIDE 42

Computer Vision

slide-43
SLIDE 43

Computer Vision

slide-44
SLIDE 44

Computer Vision

slide-45
SLIDE 45
slide-46
SLIDE 46

Wearables

Real-Time Data Analysis Biometric Feedback Unprecedented Insight

slide-47
SLIDE 47

Sports Betting Use Cases

slide-48
SLIDE 48

Use Case Summary

Handicapping Risk Management Bet Recommendations Responsible Gaming Fraud Detection

slide-49
SLIDE 49

Causes for Concern

Backdoor Functionality Black Box Problem Flash Crash Potential Odds Manipulation Fraud Masking

slide-50
SLIDE 50

Causes for Concern

Backdoor Functionality Black Box Problem Flash Crash Potential Odds Manipulation Fraud Masking

The high demand for AI products combined with their complex nature has led many companies to falsely advertise solutions that are far less sophisticated than they purport to be.

slide-51
SLIDE 51

Handicapping (Pre-match)

Machine Learning offers dramatic improvements over industry standards in setting pre-match odds.

Identifying Non-Linear Relationships Analyzing Large Data Sets Improving Without Intervention Enhanced Portability & Transferability

These benefits should all be viewed in the context of reducing human error while freeing up intellectual capital to be deployed elsewhere within an

  • rganization.
slide-52
SLIDE 52

Handicapping (Pre-match)

Machine Learning offers dramatic improvements over industry standards in setting pre-match odds.

Identifying Non-Linear Relationships Analyzing Large Data Sets Improving Without Intervention Enhanced Portability & Transferability

Linear Non-Linear

Many relationships do not fit linear functions. Machine Learning engines

  • ffer dramatic

improvements in extracting such trends.

These benefits should all be viewed in the context of reducing human error while freeing up intellectual capital to be deployed elsewhere within an

  • rganization.
slide-53
SLIDE 53

Handicapping (Pre-match)

Machine Learning offers dramatic improvements over industry standards in setting pre-match odds.

Identifying Non-Linear Relationships Analyzing Large Data Sets Improving Without Intervention Enhanced Portability & Transferability

With vast increases in computational speed and the availability of robust data sets, architectures best equipped to handle large amounts of information will become industry standards.

These benefits should all be viewed in the context of reducing human error while freeing up intellectual capital to be deployed elsewhere within an

  • rganization.
slide-54
SLIDE 54

Handicapping (Pre-match)

Machine Learning offers dramatic improvements over industry standards in setting pre-match odds.

Identifying Non-Linear Relationships Analyzing Large Data Sets Improving Without Intervention Enhanced Portability & Transferability

Beyond their inherent performance advantages, Machine Learning algorithms are able to continuously and iteratively improve themselves.

These benefits should all be viewed in the context of reducing human error while freeing up intellectual capital to be deployed elsewhere within an

  • rganization.
slide-55
SLIDE 55

Handicapping (Pre-match)

Machine Learning offers dramatic improvements over industry standards in setting pre-match odds.

Identifying Non-Linear Relationships Analyzing Large Data Sets Improving Without Intervention Enhanced Portability & Transferability

Transfer Learning is a method by which a model developed for a given task is repurposed for an unrelated one.

Source Tasks

Knowledge

Learning System

These benefits should all be viewed in the context of reducing human error while freeing up intellectual capital to be deployed elsewhere within an

  • rganization.
slide-56
SLIDE 56

Handicapping (In-Play)

Accurate Real-Time Odds Inconsistency Reduction Suspension Minimization Enhanced Risk Management Pre-trained models combined with maximally efficient algorithms allow can be leveraged into competitive advantages.

Not only does Machine Learning increase short-term operator profitability, but it vastly improves the user experience, boosting customer retention.

slide-57
SLIDE 57

Handicapping (In-Play)

Accurate Real-Time Odds Inconsistency Reduction Suspension Minimization Enhanced Risk Management Pre-trained models combined with maximally efficient algorithms allow can be leveraged into competitive advantages.

Pre-trained models and cutting-edge algorithms provide superior speed and accuracy in real time

  • dds generation.

Not only does Machine Learning increase short-term operator profitability, but it vastly improves the user experience, boosting customer retention.

slide-58
SLIDE 58

Handicapping (In-Play)

Accurate Real-Time Odds Inconsistency Reduction Suspension Minimization Enhanced Risk Management Pre-trained models combined with maximally efficient algorithms allow can be leveraged into competitive advantages.

Ideal models will transition seamlessly from pre-match to in-play, minimizing the likelihood

  • f exploitable

inconsistencies in odds.

Not only does Machine Learning increase short-term operator profitability, but it vastly improves the user experience, boosting customer retention.

slide-59
SLIDE 59

Handicapping (In-Play)

Accurate Real-Time Odds Inconsistency Reduction Suspension Minimization Enhanced Risk Management Pre-trained models combined with maximally efficient algorithms allow can be leveraged into competitive advantages.

Operators commonly suspend markets for reasons that can be mitigated or avoided entirely with sufficient algorithmic capabilities.

New Information Suspicious Betting Unbalanced Exposure

Suspension Causes

Not only does Machine Learning increase short-term operator profitability, but it vastly improves the user experience, boosting customer retention.

slide-60
SLIDE 60

Handicapping (In-Play)

Accurate Real-Time Odds Inconsistency Reduction Suspension Minimization Enhanced Risk Management Pre-trained models combined with maximally efficient algorithms allow can be leveraged into competitive advantages.

The ability to predict betting trends and update models in real-time allows for streamlined, automated,

  • ptimal risk management

Not only does Machine Learning increase short-term operator profitability, but it vastly improves the user experience, boosting customer retention.

slide-61
SLIDE 61

Risk Management

Real-Time Book Balancing Efficient Suspension Implementation Increased Turnover Capacity

*Turnover: Total dollar amount of wagers accepted

slide-62
SLIDE 62

Bet Recommendations

NHL Playoffs Insurance

Get up to $50 Back on Your Bet

Site Credit Refund

$25 NBA Parlay Insurance

Cash Back If Bet Doesn’t Hit

Playoffs

Today’s Pick:

slide-63
SLIDE 63

Bet Recommendations

NHL Playoffs Insurance

Get up to $50 Back on Your Bet

Site Credit Refund

$25 NBA Parlay Insurance

Cash Back If Bet Doesn’t Hit

Playoffs

Today’s Pick:

Promotion Type Preference

slide-64
SLIDE 64

Bet Recommendations

NHL Playoffs Insurance

Get up to $50 Back on Your Bet

Site Credit Refund

$25 NBA Parlay Insurance

Cash Back If Bet Doesn’t Hit

Playoffs

Today’s Pick:

Promotion Type Preference Unbalanced Exposure

slide-65
SLIDE 65

Bet Recommendations

NHL Playoffs Insurance

Get up to $50 Back on Your Bet

Site Credit Refund

$25 NBA Parlay Insurance

Cash Back If Bet Doesn’t Hit

Playoffs

Today’s Pick:

Promotion Type Preference Unbalanced Exposure Bet Type Affinity

slide-66
SLIDE 66

Responsible Gaming

Enhanced pattern recognition will revolutionize an operator’s ability to detect deviations from responsible gaming

slide-67
SLIDE 67

Responsible Gaming

Enhanced pattern recognition will revolutionize an operator’s ability to detect deviations from responsible gaming

slide-68
SLIDE 68

Responsible Gaming

Enhanced pattern recognition will revolutionize an operator’s ability to detect deviations from responsible gaming

Sustainable

slide-69
SLIDE 69

Sustainable Gaming

Enhanced pattern recognition will revolutionize an operator’s ability to detect deviations from responsible gaming

  • Avg. Wager

$10.01 Wager StDev $0.41 Bets/Week 4.3 (85% Baseball) % Player Props 17%

  • Max. Bet

$35.00 User_01093 User_26571

  • Avg. Wager

$210.87 Wager StDev $94.36 Bets/Week 29.0 (88% ATS) % Player Props 0%

  • Max. Bet

$1100.00

slide-70
SLIDE 70

Sustainable Gaming

Enhanced pattern recognition will revolutionize an operator’s ability to detect deviations from responsible gaming

Bet 1 NYY -135 Risk: $10 Result: Lose Bet 2 NYM +125 Risk: $10 Result: Lose Bet 3 LAD -195 Risk: $12 Result: Lose Bet 4 BET ACCEPTED NYY -215 Risk: $10 Result: Win

  • Avg. Wager

$10.01 Wager StDev $0.41 Bets/Week 4.3 (85% Baseball) % Player Props 17%

  • Max. Bet

$35.00 User_01093 User_26571

  • Avg. Wager

$210.87 Wager StDev $94.36 Bets/Week 29.0 (88% ATS) % Player Props 0%

  • Max. Bet

$1100.00

>4σ Above Average

slide-71
SLIDE 71

Sustainable Gaming

Enhanced pattern recognition will revolutionize an operator’s ability to detect deviations from responsible gaming

Bet 1 NYY -135 Risk: $10 Result: Lose Bet 2 NYM +125 Risk: $10 Result: Lose Bet 3 LAD -195 Risk: $12 Result: Lose Bet 4 BET ACCEPTED NYY -215 Risk: $10 Result: Win Bet 1 NYY -135 Risk: $10 Result: Lose Bet 2 NYM +125 Risk: $20 Result: Lose Bet 3 LAD -195 Risk: $40 Result: Lose Bet 4 BET REJECTED BKN +3.5 Risk: $80 Result: Blocked

  • Avg. Wager

$10.01 Wager StDev $0.41 Bets/Week 4.3 (85% Baseball) % Player Props 17%

  • Max. Bet

$35.00 User_01093 User_26571

  • Avg. Wager

$210.87 Wager StDev $94.36 Bets/Week 29.0 (88% ATS) % Player Props 0%

  • Max. Bet

$1100.00

Doubling Pattern

>4σ Above Average

slide-72
SLIDE 72

Enhanced pattern recognition will revolutionize an operator’s ability to detect deviations from responsible gaming

Bet 1 NYY -135 Risk: $10 Result: Lose Bet 2 NYM +125 Risk: $10 Result: Lose Bet 3 LAD -195 Risk: $12 Result: Lose Bet 4 BET ACCEPTED NYY -215 Risk: $10 Result: Win Bet 1 NYY -135 Risk: $10 Result: Lose Bet 2 NYM +125 Risk: $20 Result: Lose Bet 3 LAD -195 Risk: $40 Result: Lose Bet 4 BET REJECTED BKN +3.5 Risk: $80 Result: Blocked Bet 1 MIL -4 (-110) Risk: $330 Result: Lose Bet 2 BKN +3.5 (-110) Risk: $220 Result: Lose Bet 3 GS -120 Risk: $465 Result: Lose Bet 4 BET ACCEPTED TOR -3 Risk: $220 Result: Win

  • Avg. Wager

$10.01 Wager StDev $0.41 Bets/Week 4.3 (85% Baseball) % Player Props 17%

  • Max. Bet

$35.00 User_01093 User_26571

  • Avg. Wager

$210.87 Wager StDev $94.36 Bets/Week 29.0 (88% ATS) % Player Props 0%

  • Max. Bet

$1100.00

Doubling Pattern

>4σ Above Average

Sustainable Gaming

slide-73
SLIDE 73

Enhanced pattern recognition will revolutionize an operator’s ability to detect deviations from responsible gaming

Bet 1 NYY -135 Risk: $10 Result: Lose Bet 2 NYM +125 Risk: $10 Result: Lose Bet 3 LAD -195 Risk: $12 Result: Lose Bet 4 BET ACCEPTED NYY -215 Risk: $10 Result: Win Bet 1 NYY -135 Risk: $10 Result: Lose Bet 2 NYM +125 Risk: $20 Result: Lose Bet 3 LAD -195 Risk: $40 Result: Lose Bet 4 BET REJECTED BKN +3.5 Risk: $80 Result: Blocked Bet 1 MIL -4 (-110) Risk: $330 Result: Lose Bet 2 BKN +3.5 (-110) Risk: $220 Result: Lose Bet 3 GS -120 Risk: $465 Result: Lose Bet 4 BET ACCEPTED TOR -3 Risk: $220 Result: Win Bet 1 MIL -4 (-110) Risk: $330 Result: Lose Bet 2 BKN +160 Risk: $393.75 Result: Lose Bet 3

  • S. Curry o3 TOs

Risk: $200 Result: Lose Bet 4 BET REJECTED

  • C. Paul u16 PTS

Risk: $550 Result: Blocked

  • Avg. Wager

$10.01 Wager StDev $0.41 Bets/Week 4.3 (85% Baseball) % Player Props 17%

  • Max. Bet

$35.00 User_01093 User_26571

  • Avg. Wager

$210.87 Wager StDev $94.36 Bets/Week 29.0 (88% ATS) % Player Props 0%

  • Max. Bet

$1100.00

Doubling Pattern Loss-Chasing Pattern New Bet Type

>4σ Above Average

Sustainable Gaming

slide-74
SLIDE 74

Fraud Detection

Syndicate Betting Bonus Exploitation Betting on Behalf

  • f 3rd Parties

Arbitrage

slide-75
SLIDE 75

Use Case Summary

Handicapping Risk Management Bet Recommendations Responsible Gaming Fraud Detection

slide-76
SLIDE 76

Questions?

slide-77
SLIDE 77

Thank You

Office Hours: 1:15pm – 2:00pm

Lloyd Danzig

SHARP ALPHA ADVISORS

slide-78
SLIDE 78

Appendix

slide-79
SLIDE 79

Bookmaking Economics

slide-80
SLIDE 80

1 2 3 4

$225 $100

+$325

Economics: Sportsbook

Customers view odds set by sportsbook

+190

  • 225

+5.5

(-110)

  • 5.5

(-110)

slide-81
SLIDE 81

1 2 3 4

$225 $100

+$325

Economics: Sportsbook

Customers view odds set by sportsbook

+190

  • 225

+5.5

(-110)

  • 5.5

(-110)

slide-82
SLIDE 82

1 2 3 4

$225 $100

+$325

Economics: Sportsbook

Customers view odds set by sportsbook

+190

  • 225

+5.5

(-110)

  • 5.5

(-110)

slide-83
SLIDE 83

1 2 3 4

$225 $100

+$325

Economics: Sportsbook

Customers view odds set by sportsbook

+190

  • 225

+5.5

(-110)

  • 5.5

(-110)

slide-84
SLIDE 84

1 2 3 4

$225 $100

+$325

Economics: Sportsbook

Customers view odds set by sportsbook

+190

  • 225

+5.5

(-110)

  • 5.5

(-110)

Overround: Bookmaker will pay out $100.00 for every $103.71 it collects

slide-85
SLIDE 85

1 2 3 4

$225 $100

+$325

Economics: Sportsbook

Customers view odds set by sportsbook

+190

  • 225

+5.5

(-110)

  • 5.5

(-110)

Overround: Bookmaker will pay out $100.00 for every $103.71 it collects Profit Margin:

slide-86
SLIDE 86

1 2 3 4

$225 $100

+$325

Economics: Sportsbook

Customers view odds set by sportsbook

+190

  • 225

+5.5

(-110)

  • 5.5

(-110)

Overround: Bookmaker will pay out $100.00 for every $103.71 it collects Profit Margin: Expected Profit:

slide-87
SLIDE 87

1 2 3 4

$225 $100

+$325

Economics: Sportsbook

Customers view odds set by sportsbook

+190

  • 225

+5.5

(-110)

  • 5.5

(-110)

Overround: Bookmaker will pay out $100.00 for every $103.71 it collects Profit Margin: Expected Profit:

Simulation:

Generate large set of random numbers between 0 and 1 IF(Value < 0.3325) ฀ Sportsbook Wins $35 ELSE ฀ Sportsbook Wins $0 Average profits across all iterations for estimate of Expected Profit

slide-88
SLIDE 88

1 2 3 4

Alice thinks New York has a 33% chance of winning, represented in fair odds as +203.

The best sportsbook is offering New York +190, so Alice will prefer odds of +203.

Economics: Betting Exchange

slide-89
SLIDE 89

1 2 3 4

Alice thinks New York has a 33% chance of winning, represented in fair odds as +203.

The best sportsbook is offering New York +190, so Alice will prefer odds of +203.

She offers (“lays”) to accept a wager from anyone interested in Detroit -203.

The best sportsbook is offering Detroit -225, so Bob accepts the other side of Alice’s wager.

Economics: Betting Exchange

slide-90
SLIDE 90

1 2 3 4

Alice thinks New York has a 33% chance of winning, represented in fair odds as +203.

The best sportsbook is offering New York +190, so Alice will prefer odds of +203.

She offers (“lays”) to accept a wager from anyone interested in Detroit -203.

The best sportsbook is offering Detroit -225, so Bob accepts the other side of Alice’s wager.

Economics: Betting Exchange

slide-91
SLIDE 91

1 2 3 4

Alice thinks New York has a 33% chance of winning, represented in fair odds as +203.

The best sportsbook is offering New York +190, so Alice will prefer odds of +203.

She offers (“lays”) to accept a wager from anyone interested in Detroit -203.

The best sportsbook is offering Detroit -225, so Bob accepts the other side of Alice’s wager.

Economics: Betting Exchange

Simulation:

Generate large set of random numbers between 0 and 1 IF(Value < 0.3325) ฀ Sportsbook Wins $10.15 ELSE ฀ Sportsbook Wins $5.00 Average profits across all iterations for estimate of Expected Profit

slide-92
SLIDE 92

Platform Comparison

Sportsbook Betting Exchange

Favorable Odds Operator Risk Potential Market Variety Reward/Bonus Programs Bet to Lose Bet Matching Predictive Capacity Max Profit (Operator)

Liquidity remains the largest challenge.

Exchanges offer the benefit of being riskless to operate, since payouts to winners come from deposits by losers.

slide-93
SLIDE 93

Platform Comparison

Sportsbook Betting Exchange

Favorable Odds Operator Risk Potential Market Variety Reward/Bonus Programs Bet to Lose Bet Matching Predictive Capacity Max Profit (Operator)

Liquidity remains the largest challenge.

Exchanges offer the benefit of being riskless to operate, since payouts to winners come from deposits by losers.

slide-94
SLIDE 94

Platform Comparison

Sportsbook Betting Exchange

Favorable Odds Operator Risk Potential Market Variety Reward/Bonus Programs Bet to Lose Bet Matching Predictive Capacity Max Profit (Operator)

Liquidity remains the largest challenge.

Exchanges offer the benefit of being riskless to operate, since payouts to winners come from deposits by losers.

slide-95
SLIDE 95

Platform Comparison

Sportsbook Betting Exchange

Favorable Odds Operator Risk Potential Market Variety Reward/Bonus Programs Bet to Lose Bet Matching Predictive Capacity Max Profit (Operator)

Liquidity remains the largest challenge.

Exchanges offer the benefit of being riskless to operate, since payouts to winners come from deposits by losers.

slide-96
SLIDE 96

Platform Comparison

Sportsbook Betting Exchange

Favorable Odds Operator Risk Potential Market Variety Reward/Bonus Programs Bet to Lose Bet Matching Predictive Capacity Max Profit (Operator)

Liquidity remains the largest challenge.

Exchanges offer the benefit of being riskless to operate, since payouts to winners come from deposits by losers.

slide-97
SLIDE 97

Platform Comparison

Sportsbook Betting Exchange

Favorable Odds Operator Risk Potential Market Variety Reward/Bonus Programs Bet to Lose Bet Matching Predictive Capacity Max Profit (Operator)

Liquidity remains the largest challenge.

Exchanges offer the benefit of being riskless to operate, since payouts to winners come from deposits by losers.

slide-98
SLIDE 98

Platform Comparison

Sportsbook Betting Exchange

Favorable Odds Operator Risk Potential Market Variety Reward/Bonus Programs Bet to Lose Bet Matching Predictive Capacity Max Profit (Operator)

Liquidity remains the largest challenge.

Exchanges offer the benefit of being riskless to operate, since payouts to winners come from deposits by losers.

slide-99
SLIDE 99

Platform Comparison

Sportsbook Betting Exchange

Favorable Odds Operator Risk Potential Market Variety Reward/Bonus Programs Bet to Lose Bet Matching Predictive Capacity Max Profit (Operator)

Liquidity remains the largest challenge.

Exchanges offer the benefit of being riskless to operate, since payouts to winners come from deposits by losers.

slide-100
SLIDE 100

Platform Comparison

Sportsbook Betting Exchange

Favorable Odds Operator Risk Potential Market Variety Reward/Bonus Programs Bet to Lose Bet Matching Predictive Capacity Max Profit (Operator)

Liquidity remains the largest challenge.

Exchanges offer the benefit of being riskless to operate, since payouts to winners come from deposits by losers.