Determining odds of a possible outcome of an event which occurs during a contest
A method of determining odds of a possible outcome of an event which occurs during a contest is disclosed. The contest may be a sporting contest such as a football game. The method includes accepting a set of one or more current parameter values (such as score or field position) which influence the likelihood of the possible event outcome (such as the likelihood of a running play). The set of current parameter values is used to retrieve situation-specific likelihood data from a data store. Odds associated with the possible outcome are set based at least in part on the likelihood data. The odds may be adjusted using one or more non-situation-specific factors indicating an average likelihood of the event outcome (such as an average likelihood of a running play for a team regardless of situation). The method may expedite odds determination. An associated system and machine readable medium are also disclosed.
The invention relates to the determination of odds of a possible outcome of an event which occurs during a contest, for possible use in electronic wagering systems which permit bets to be placed during live contests, such as sporting contests.
BACKGROUND OF THE INVENTIONElectronic wagering systems which permit bets to be placed during live sporting contests are known. In one arrangement, the entity that accepts the wager (the house) provides a betting interface to a user (a bettor) by way of a web site. During the contest (e.g. a football game), which the user may watch via a televised broadcast, the user accesses the web site via an Internet-connected computer. The house presents a set of propositions to the user via the web site in respect of which bets may be placed. A “proposition” is a query regarding an event which has multiple potential outcomes. The event is typically a discrete event which is expected occur in the relatively near future (e.g., a next play, kick, or throw) in the context of a sporting contest (or simply “sport”). The set of propositions changes dynamically as play progresses. The propositions displayed at any given moment depend upon the current status of the game, as well as the type of sporting contest being played. For example, propositions may include, in the case of a football game in which a play is about to commence: “Will the next play be a run or a pass?”. In the case of a soccer match where a penalty kick is imminent, a proposition may be “Will the penalty kick be successful?”. And in the case of a baseball game where a pitch is about to be thrown, a proposition may be “Will the next pitch be a ball or a strike?”.
Each proposition is presented along with the odds associated with each potential outcome. The term “odds” refers to the amount that a bet on an event outcome pays, which is based on the actual likelihood of the event outcome but may include other factors, such as a house commission (referred to as “juice” in betting parlance). The proposition and associated odds are cumulatively referred to as a “line”. The act of presenting lines to a user so as to permit the user to place bets is referred to as the posting of lines “on the board” (or simply “on board”).
The user may place a bet in respect of any line that is currently on board, e.g., by using his computer keyboard or mouse to select an outcome and specify a betting amount. When the relevant event is about to commence, lines pertaining to that event are typically taken “off the board” (or simply “off board”) by the house to prevent users from betting once the actual outcome is suggested or apparent. After the event has occurred, the house is responsible for assessing the outcome of the event, determining whether each bet placed in respect of that event has won or lost, informing the user of the outcome, and crediting or debiting a user's account appropriately based on the amount of the wager and the odds associated with the selected outcome.
The determination of odds is an important aspect of known electronic wagering systems because the odds will ultimately determine whether or not the house will profit. As is known in the art, the house does not usually place any bets, but merely serves as a broker, matching users who bet on opposing outcomes. From the house's perspective, it is desirable for the bets placed on opposing outcomes of a proposition to be equal (balanced), so that the winners' profits can be paid from the losers' losses, while the house keeps the commissions which it charges for brokering the transactions. The odds for opposing outcomes should thus ideally be set to encourage equal betting between outcomes across multiple users. As well, the odds should not be such that a few lucky bets placed on unlikely outcomes will result in a depletion of the house's funds.
The odds that are presented to a user are usually set by a human administrator acting on behalf of the house. The administrator may also be in charge of posting lines onto the board, removing lines from the board, and determining bet outcomes and payouts. Accordingly, the administrator may be burdened with many tasks.
Contributing to the burden on a human administrator is the limited amount of time that may be available for performing the above-noted tasks. During the course of a single National Football League (NFL) or National Collegiate Athletic Association (NCAA) football game, for example, 100-150 plays are typically executed. When no timeouts, commercial breaks or injury stoppages have been made, an NFL team is normally allowed a maximum of 40 seconds after the completion of a play to commence the next one to avoid a penalty (a maximum of only 25 seconds is allowed in the NCAA). Teams typically begin a play with 5 to 15 seconds left on the play clock, leaving at most a 25 to 35 second window of opportunity for a human administrator to assess the situation, determine appropriate odds for each of the propositions to be presented and post lines on the board so that users may review the odds and place bets. If a football team adopts what is known as a “no-huddle” or “hurry-up” offense, this window of opportunity may be reduced to as short as 5 to 10 seconds.
During this short period of time, odds must be determined and lines placed on the board as quickly as possible, so that users have sufficient time read the odds and place bets. If a user does not have enough time to place a bet, potential income for the house may be lost. Creating odds with speed is therefore important for purposes of promoting betting activity and for increasing the potential income for the house. The degree of speed that is desirable may of course depend on the type of contest that is under way (e.g. slower response times may be acceptable for sports having a slower pace, such as golf or baseball, than for other sports such as football). Nevertheless, faster odds determination is generally better from the perspective of the house.
Rapid odds determination may also be desirable for contests other than sporting contests (e.g. elections).
A manner of quickly determining odds associated with an event outcome would therefore be desirable.
SUMMARY OF THE INVENTIONA method of determining odds of a possible outcome of an event which occurs during a contest is disclosed. The contest may be a sporting contest such as a football game. The method includes accepting a set of one or more current parameter values (such as score or field position) which influence the likelihood of the possible event outcome (such as the likelihood of a running play). The set of current parameter values is used to retrieve situation-specific likelihood data from a data store. Odds associated with the possible outcome are set based at least in part on the likelihood data. The odds may be adjusted using one or more non-situation-specific factors indicating an average likelihood of the event outcome (such as an average likelihood of a running play for a team regardless of situation). The method may expedite odds determination. An associated system and machine readable medium are also disclosed.
In accordance with an aspect of the present invention there is provided a method of determining odds of a possible outcome of an event which occurs during a contest, comprising: accepting a set of one or more current parameter values, a likelihood of the possible outcome of the event being dependent upon each parameter; utilizing the set of current parameter values to retrieve likelihood data from a data store; and setting odds associated with the possible outcome of the event based at least in part on the likelihood data.
In accordance with another aspect of the present invention there is provided a computer system for determining odds of a possible outcome of an event which occurs during a contest, comprising: a user interface; a display; a data store; and a processor for: receiving a set of one or more current parameter values from the user interface, a likelihood of the possible outcome of the event being dependent upon each parameter; utilizing the set of current parameter values to retrieve likelihood data from the data store; setting odds associated with the possible outcome of the event based at least in part on the likelihood data; and displaying an indication of the odds on the display.
In accordance with still another aspect of the present invention there is provided a machine-readable medium encoded for determining odds of a possible outcome of an event which occurs during a contest, comprising: machine executable code for accepting a set of one or more current parameter values, a likelihood of the possible outcome of the event being dependent upon each parameter; machine executable code for utilizing the set of current parameter values to retrieve likelihood data from a data store; and machine executable code for setting odds associated with the possible outcome of the event based at least in part on the likelihood data.
Other aspects and features of the present invention will become apparent to those ordinarily skilled in the art upon review of the following description of specific embodiments of the invention in conjunction with the accompanying figures.
BRIEF DESCRIPTION OF THE DRAWINGSExemplary embodiments of the invention will now be described with reference to the attached drawings in which:
Electronic wagering system 10 is a distributed system having four components, namely, a bet server 14, an administrator computer 18, a broadcast server 26 and a bettor computer 22. The bet server 14, administrator computer 18, broadcast server 26 and bettor computer 22 may be located in different geographic locations. In the illustrated embodiment, the bet server 14 and broadcast server 26 are co-located. The four devices 14, 18, 26 and 22 intercommunicate by way of a network 20, which in the present embodiment comprises the public Internet, but which may comprise other types of networks in alternative embodiments.
The bet server 14 is a computing device, such as an INTEL® processor-based server platform for example, which executes bet server software. The bet server 14 has a number of responsibilities which generally pertain to the betting process. First, the bet server 14 is responsible for establishing betting sessions with one or more bettor computers 22. This may entail authenticating bettor usernames and passwords entered at betting computers 22, to ensure that any subsequent betting transactions which may occur are genuine. Second, the bet server 14 is responsible for receiving and validating bet requests from one or more bettor computers 22. Validation may entail checking a bettor account for sufficiency of funds as well as checking bet timestamps to ensure that bets are not placed after a betting time window has closed. Third, the bet server 14 receives grading notifications from the administrator computer 18 indicating which of a number of potential event outcomes has actually occurred. This information is used for the purpose of assessing whether the bets that have been placed are winners or losers. The bet server 14 may have additional responsibilities. The bet server software may be implemented as a .NET Windows service written in the C# programming language, and may be loaded from a machine readable medium, such as a removable magnetic or optical disk 15. In the present embodiment, the bet server 14 is located at a sportsbook location.
The bet server 14 has an associated database 16. Database 16 is a SQL server database which serves as a central data store for various types of data associated with the electronic wagering system 10. For example, database 16 includes data regarding bettor accounts, ongoing sporting contests (“games”), current propositions and lines regarding upcoming game events, actual event outcomes, placed bets, and bettor payouts. As well, the database 16 stores historical data 17 consisting of historical league and team averages for various game event outcomes. The historical data 17 is used for purposes of quickly determining odds associated with a particular event outcome for a team of interest in a particular set of game circumstances, as will be described. The database 16 is typically co-located with the bet server 14, although this is not required.
Administrator computer 18 is a computing device executing software which permits a human administrator 19 acting on behalf of the house to control the electronic wagering system 10. In particular, the software executing at administrator computer 18 presents a control panel graphical user interface (GUI) 50 for controlling, for a chosen sporting contest, which bets (i.e. which lines) are available to bettors 23 at any given time as well as the odds associated with potential event outcomes. As lines and odds are determined by the administrator 19, they are communicated to broadcast server 26 (described below) which in turn broadcasts this information to one or more bettor computers 22. The GUI 50 also includes an editable scoreboard which is updated by the administrator 19 throughout the course of a contest for presentation to bettors 23, also by way of broadcast server 26. As well, the GUI 50 of administrator computer 18 allows the administrator 19 to “grade” lines, i.e., to formally indicate which of the multiple potential outcomes of a recent event actually occurred. When lines are graded, appropriate grading notifications are sent from administrator computer 18 to the bet server 14, so that bet outcomes may be determined.
The control panel GUI 50 employs a number of mechanisms which permit the administrator 19 to quickly determine and adjust odds associated with event outcomes and post lines “on the board” (i.e. cause lines in respect of which bets may be placed to be displayed at bettor computers 22), so that bettors 23 are afforded as much time as possible to review odds and place bets. These mechanisms include automatic determination of default odds associated with certain event outcomes, as well as the optional automatic adjustment of odds for certain event outcomes, as will be described. The software executing on administrator computer 18 may be implemented as a C#.Net Windows® application, and may loaded from a machine readable medium, such as a removable magnetic or optical disk 21 for example. Administrator computer 18 is located at an administrator location, which may conveniently be a location having a broadband Internet connection.
Broadcast server 26 is a server which broadcasts scoreboard, line and odds information received from administrator computer 18 to all connected bettor computers 22. The broadcast server 26 essentially alleviates the administrator computer 18 from the burden of communicating with multiple bettor computers. Broadcast server 26 receives notifications indicative of the creation and modification of lines, as well as indications that lines are to be posted on board or taken off board and scoreboard updates, from administrator computer 18. This information may be combined with other data retrieved from database 16 and broadcast to all connected bettor computers 22. In the present embodiment, the broadcast information takes the form of XML messages which are created from .Net DataSets generated at broadcast server 26. The server 26 uses a TCP/IP multicasting system to efficiently transmit the XML messages to all connected bettor computers 22 (the same messages are sent to each computer 22). To achieve these objectives, the broadcast server 26 executes software 27, which may be loaded from a machine readable medium, such as a removable magnetic or optical disk 27 for example.
Bettor computer 22 is a computing device, such as an Internet connected personal computer (PC), providing a client interface 24 which allows a bettor 23 to interact with the electronic wagering system 10. The client interface 24 dynamically displays data based on scoreboard, line and odds information received periodically from broadcast server 26 and upon bet information received from bet server 14 to a bettor 23. The client interface 24 allows a bettor 23 to perform such actions as: view current game information including game status and score; view currently posted lines; place bets; and view bet outcome information. The client interface 24 may for example be implemented as a Java™ applet 25 which is launched from a main web page. The main web page may be a simple ASP .Net application that may be hosted on the broadcast server 26, the bet server 14, or on a separate machine, whose purpose may solely be to provide the Java™ applet to bettors via their web browsers. The user may access the main web page by way of a commercially available web browser, such as Internet Explorer™ or Firefox™ for example, executing at bettor computer 22. Conveniently, most Internet-connected computing devices equipped with suitable web browsers may become bettor computers 22. It will be appreciated that the number of bettor computers 22 connected to the system 10 at any given time may be greater than one, however only one computer 22 is illustrated in
The electronic wagering system 10 may include other components in addition to bet server 14, administrator computer 18, broadcast server 26 and bettor computer 22, which are not illustrated in
Referring first to
Scoreboard panel 32 displays current game status and score information to the bettor 23. The information displayed within scoreboard panel 32 is the same sort of information that is displayed on a scoreboard at a stadium where NFL football games are played. The various fields appearing in scoreboard panel 32 are referred to generically as dynamic contest parameters, since they change dynamically as the contest progresses. The fields that are displayed in panel 32 may differ for different sports. In the present NFL football example, the fields include quarter, score, time left to play in the current quarter (the “clock”), field position, down, yards to go to next first down, and possession information. The information displayed in scoreboard panel 32 is received from broadcast server 26 and originates from administrator computer 18.
Betting panel 34 indicates the status of the bettor's financial account and allows the bettor 23 to select a default betting amount for “one-click” betting (described below). The balance information displayed in betting panel 34 originates from bet server 14.
The lines panel 36 presents lines that are currently on the board, i.e., visible by the bettor 23 and available for betting. The set of displayed lines changes dynamically as the football game progresses. In the example illustrated in
The odds for each outcome are displayed in American terms within the button associated with that outcome. In American terms, a positive value +X means that 100 dollars must be wagered to win X dollars, whereas a negative value −Y means that Y dollars must be wagered to win 100 dollars. The manner in which these odds are determined will be described below.
The user may alternatively choose to have odds displayed in European terms through a configuration option of client interface 24. Positive American odds are converted to European terms by dividing the odds by 100 and adding 1 (e.g. +250 would convert to 3.50). Negative odds are converted by inverting the odds (i.e. dividing 1 by the odds), multiplying by −100 and adding 1 (e.g. −150 would convert to 1.67). The lines information displayed in lines panel 36 is received from broadcast server 26.
Bet list panel 39 displays a list of the bets that have been placed by the bettor 23 during the current game, along with bet outcomes and payouts. After a bet is placed, the bet outcome will indicate “pending” until grading is completed by the administrator 19.
To place a bet, a bettor 23 who has specified a default betting amount in the betting panel 34 of client interface 24 merely clicks on (i.e. selects) a button associated with the desired event outcome in the current lines panel 36. The amount of the bet will be automatically be set to equal the default betting amount specified in betting panel 34. This is referred to as “one-click” betting. Advantageously, bets may be placed quickly in this manner, which increases the likelihood of a greater number of bets. Alternatively, the user may configure the client interface 24 to present a confirmation prompt before any bet is accepted, so that the likelihood of unintentionally placed bets is reduced.
Turning to
Scoreboard panel 52 provides various controls for updating dynamic contest parameters, such as current game status and score information, that are ultimately presented to bettor 23. In the present embodiment, the administrator 19 updates scoreboard content via edit boxes 54 and radio buttons 55, as game play progresses. The administrator 19 may merely mimic scoreboard updates which occur on the scoreboard of a televised broadcast of the game. After each scoreboard update or set of updates is made in panel 52, the administrator 19 selects the “Lines Onboard” button 68 (below lines panel 56) in order to communicate the updated scoreboard, along with any new lines, to bettor(s) 23 via broadcast server 26 (the Lines Onboard button 68 is described in more detail below). In situations when no new lines need to be posted for an extended period (e.g. during a basketball game), such that the administrator 9 wishes to refresh bettors' client interface 24 without creating lines, the “Send Scoreboard” button 58 (also below lines panel 56) may be selected in order to simply update bettors' scoreboards. The “Send Scoreboard” button 58 has an associated hotkey <F12> on the keyboard of administrator computer 18, which may be selected instead of button 58. The hotkey frees the administrator 19 from having to use a computer mouse or other pointing device, which may be cumbersome and time-consuming, in order to send scoreboard updates to bettors.
Lines panel 56 is used by the administrator 19 for the purpose of creating lines, posting lines on the board, and removing lines from the board as game play progresses. Four exemplary lines 60, 62, 64 and 66 are illustrated in
Active field 70 is a checkbox which indicates whether the line is currently “active”, i.e., editable by the administrator 19 and eligible for posting on the board. When a line is active, the line will be posted on board when the administrator selects the “Lines Onboard” button 68 and will be taken off board when the administrator 19 selects the “Lines Offboard” button 69, unless the line is locked (see below description of lock field 72). “Lines Onboard” button 68 and “Lines Offboard” button 69, as well as their associated keyboard hotkeys <Enter> and <Esc> respectively, are intended to permit the administrator 19 to quickly post lines onto the board and quickly remove lines from the board, respectively. Selection of button 68 may be referred to generically as the generation of a first user prompt while selection of button 69 may be referred to as the generation of a second user prompt (in this context, the term “user” refers to the administrator 19). When a line is inactive, the line is not editable and is not posted on board or taken off board when the administrator 19 selects button 68 or 69 respectively. In the present embodiment, odds are not displayed for inactive lines. An administrator 19 may maintain inactive lines within lines panel 56 for possible later activation when game conditions are suitable.
Lock field 72 is a checkbox which indicates whether the line is currently “locked”. When a line is locked, the line will remain on board even when the “Lines Offboard” button 69 has been selected. Typically, only lines whose outcomes are not expected to be known for some time, (i.e., lines with “long term outcomes”) are locked. For example, if a line is based on the proposition “Which team will score more points this quarter?”, the event outcome will not be apparent until the quarter is completed. It may be desirable to display this line during most of the multiple plays/downs that comprise a football quarter, in order to permit bets to be placed as the quarter progresses. During the quarter, it may also be desirable to post and remove other lines based on propositions whose outcomes will be known more quickly, such as lines based on the proposition “What will the next play be?”. By locking the line with the long-term outcome, the administrator 19 may post and remove the other lines using buttons 68 and 69 without affecting the former line's “on board” status. If it is necessary to remove all lines (regardless of locked status), the administrator 19 may select the “Lines Offboard ALL” button 73, or associated hotkey <Shift+Esc>.
The status field 74 of lines panel 56 indicates whether the line is currently on board or off board. In order to provide a clear visual indication of status, the “on board” indicator may be one color while the “off board” indicator is another color. As well, both indicators may employ appropriate text in large, bold fonts. These characteristics may reduce the likelihood of the administrator 19 inadvertently leaving on board lines when they should be off board, and vice versa.
The proposition field 76 of a line indicates the proposition on which the line is based. In the present embodiment, which pertains to an NFL football game, any one of the ten propositions listed in Table 1 below may appear field 76 (potential event outcomes are indicated in the last column of Table 1):
Values in square brackets in the propositions of Table 1 are placeholder (i.e. variables) that are replaced with actual values in the lines that are displayed in lines panel 52 of control panel interface 50 and lines panel 36 of client interface 24 (
The number and types of propositions for other leagues/contests or in alternative embodiments may differ. For example, propositions for basketball may include the following:
1) What will the next point be?
2) What team will get the next block?
3) What team will get the next steal?
4) What team will get the next 3-point field goal?
5) Which player will get a steal first?
6) Which player will get a 3-point field goal first?
7) Which player will get a block first?
8) Which team will beat the spread?
9) Will the total score be over or under?
10) Which team will reach [pts] points first?
11) Which player will reach [pts] points first?
Propositions for baseball may include the following:
1) What will the outcome of this plate appearance be?
2) Will [player] score a run as a result of this plate appearance?
3) How many RBI's will [player] get during this plate appearance?
4) Will [player] get an RBI during this plate appearance?
5) Will [player] ground into a double play?
6) How many runs will the [team] score this inning?
7) How many hits will the [team] have this inning?
8) Will the [team] get a stolen base this inning?
The outcomes field 78 of a line indicates potential event outcomes for that line. Each line will have at least two mutually exclusive outcomes in field 78, with associated odds for each outcome. When a line is activated, the “default odds” for each potential outcome will be displayed within an edit box. The term “default odds” refers to a likelihood of an event outcome, versus “odds” in the sense of a bettor payout as define above. There are two types of default odds. The first type, which may be referred to as “precompiled default odds”, is based on historical data 17 (
The odds are displayed in field 78 as percentages due to the fact that most administrators are able to best appreciate the odds in this form. The percentage odds for all of the potential event outcomes in a single line will normally total 100% (e.g. the percentages in edit boxes 80, 82 and 84 of line 60 total 100%). To maintain this 100% total in the event that the administrator 19 manually adjusts one or more of the odds, which may occur as described below, one of the edit boxes may be configured to automatically set its value to “100%-T %”, where T % is the total of the percentages in all other edit boxes. The edit box having this capability may be visually distinct from the other edit boxes in order to alert the administrator 19 to this feature. For example, in line 60 of
Depending upon the type of proposition in respect of which odds are displayed, the odds appearing in the edit boxes of outcomes field 78 may be precompiled default odds or standard default odds. Precompiled default odds are only used for propositions for which the precompiled default odds have been previously generated and stored in historical data 17 of database 16. The set of propositions for which precompiled default odds are available is typically hard-coded in the system 10. In the present embodiment, precompiled default odds are available for propositions 1 to 6 of Table 1 above. Standard default odds are used for propositions 7 to 10 of Table 1.
Where the default odds displayed in the edit boxes of field 78 are precompiled default odds, they may either be unadjusted odds or adjusted odds. Unadjusted odds are based solely on game situation-specific league average statistics, and may be referred to as “straight precompiled default odds”. Adjusted default odds are also based on game situation-specific league average statistics, but further incorporate team-specific statistics which are intended to reflect any deviation of the team from league averages. In order for the odds to be displayed in the edit boxes in adjusted form, an odds adjustment formula should exist for the relevant outcome (odds adjustment formulae do not necessarily exist for all outcomes). As well, the administrator 19 should have activated odds adjustment, i.e., configured the control panel interface 50 to automatically apply the relevant formula to cause the percentages displayed in the edit boxes of field 78 to incorporate team-specific statistics.
To activate odds adjustment, the administrator 19 should have entered various statistics specific to the current team. This is typically done prior to game time (although the statistics may be entered or updated during game time). The entered statistics may be referred to as “factors” since they constitutes variable factors in otherwise fixed equations that are used for odds adjustment. The administrator 19 should also toggle a setting which activates automatic application of the formulae. This is typically done on demand at game time. Both of these steps are performed by way of a separate odds adjustment window 100 (
The exact formulae for performing odds adjustment are proposition and outcome dependent, however generally they are all quadratic equations defining a curve with the three points (0,0), (A,B) and (1,1), as follows:
Where:
-
- X=league-average probability of an event outcome occurring for a given situation (i.e. precompiled default odds to be adjusted)
- Y=adjusted outcome probability (i.e. adjusted odds)
- A=league-average probability of an event outcome occurring regardless of game situation
- B=team-specific probability of an event outcome occurring regardless of game situation
Although equation (1) may occasionally create unrealistic adjustments when B varies greatly from A, for most scenarios equation (1) results in reasonable adjustments. The specific formulae used for various propositions and event outcomes of an (American) football game will be set forth later in this description.
The odds adjustment window 100 that is launched when button 98 is selected is illustrated in
League selector radio buttons 102 allow the administrator 19 to select the relevant league for the current sporting contest. The selection of league via buttons 102 determines which league's statistics from historical data 17 of
In the present embodiment, odds adjustment is only performed for the NFL and NCAA football leagues, thus only these leagues may be selected via buttons 102. Alternative embodiments may offer odds adjustment for other leagues or sports.
The team-specific factors panel 104 of
The “Use Factors” checkbox 118 is an on-off toggle for controlling whether or not the odds adjustment formulae are to be automatically applied. When checkbox 118 contains a check mark, odds adjustment is automatically performed using whatever data has been entered in team-specific factors panel 104, as well as league-average data read from the database 16. The values in the various fields of panel 104 may be entered well in advance of a sporting contest, based on statistics obtained from a source of sports statistics such as web site www.statistics.com for example, so that the administrator 19 may merely click checkbox 118 during the contest to activate automatic odds adjustment or deactivate it.
Regardless of whether odds adjustment is performed, the odds in the edit boxes of field 78 of
In addition to displaying default odds in percentage form in the edit boxes of outcomes field 78, the actual odds that will be seen by bettors 23 are displayed nearby in superscript, in American or European terms. These odds are generated in two steps. First, the percentage odds P (i.e. decimal probability P) are converted to American terms with an applied “juice” factor by way of equation (1) below:
AmericanOdds=if((1+JuiceFactor)×(−100×P)/(1−P)<−100, (2)
(1+JuiceFactor)×(−100×P)/(1−P),
[100×(1−P)]/[(1+JuiceFactor)×P)])
In equation (2), if the term (1+JuiceFactor)×(−100×P)/(1−P) is less than −100, the American odds are computed using the term (1+JuiceFactor)×(−100×P)/(1−P); otherwise they are computed using the term [100×(1−P)]/[(1+JuiceFactor)×P)].
The juice factor constitutes the house's commission for acting as a broker. In the present embodiment, a base juice factor of 15% (i.e. 0.15) is applied, i.e. the odds are reduced by 15%. For unlikely events, the odds may be reduced by a higher percentage juice factor, with the percentage increasing progressively for event outcome likelihoods that are below a threshold. For example, a threshold may be set at 16%, such that for likelihoods that are less than 16%, an additional juice factor is added to the base juice factor. The additional juice factor may be determined by multiplying the difference between the threshold and the likelihood by a multiplier, such as 3 (e.g. for a likelihood of 13%, the additional juice factor will be (16%-13%)×3 or 9%, for a total juice factor of 15%+9% or 24%). The rationale for increasing the juice factor for unlikely or extremely unlikely event outcomes is to prevent a lucky bettor from depleting the funds of the house by placing a substantial bet on an unlikely or extremely unlikely outcome.
Second, the resultant odds are rounded down to the nearest increment of 5.
To generate European odds, the conversion from the American odds described earlier would be applied to the rounded odds. Conversion is performed after rounding so that any payout will be the same regardless of the user's choice of American or European style odds display.
The above two steps are performed automatically by the software executing at administrator computer 18. Thus, the three percentages 22%, 18.8%, and 59.2% displayed in edit boxes 80, 82 and 84 are automatically converted to +305, +375 and −170 shown at 90, 92 and 94 respectively. These are the odds that will ultimately be displayed at bettor computers 22.
Referring again to
Referring to
Situation ID table 210 is illustrated in
The third column 236 of table 220 contains outcome probabilities (i.e. precompiled default odds). The probabilities of table 220 are actually averages which have been generated from historical records of the past N NFL seasons, where N is an integer (e.g. 3). To generate a probability for a given game situation (e.g. field position: opponent's 15 yard line; down: 3rd; yards to go: 2), historical records of NFL teams faced with the identical situation are examined to determine what play the teams executed in that situation. The fact that historical averages are deemed to be probabilities of future occurrence of the event outcomes reflects an underlying assumption that event outcome probabilities are generally static over time.
An astute reader familiar with the rules of NFL football may note that the set of outcomes represented by the outcome IDs enumerated above is an incomplete set of “next plays” that may be executed by a football team. It is also possible for a fourth outcome, namely, an incomplete pass, to occur. The reason that no outcome ID is included in table 220 for this outcome is that the likelihood of an incomplete pass may be determined indirectly by subtracting the combined probabilities of a RUSH, PASS and TURNOVER from 1.0 (i.e. 100%). In general, omission of one of a finite set of potential event outcomes may be used as a technique for limiting the size of any of tables making up statistics 218, since the likelihood of the outcome may be computed indirectly as described. Conversely, it may be desirable to include all potential event outcomes in each of tables of statistics 218 to avoid the need for such “100%-X” computations.
So, by using tables 210 and 220, the league-average likelihood of a running play, complete pass, or turnover (or, indirectly, of an incomplete pass) may be determined for a particular combination of the field position, down, and yards to go parameters, all by way of simple table lookups.
It should be appreciated that not all events in respect of which bets may be placed using system 10 will be influenced by the same set of dynamic contest parameters. For example, the likelihood of success of a field goal may be influenced only by field position, and not by down or yards to go to next down. For this reason, situation ID table 210 of
Referring back to
Initially, a situation ID representative of the current contest status is retrieved (S1002) based on the current values of the field position, down, and yards to go dynamic contest parameters. The ID is retrieved by way of a table lookup in situation ID table 210 (
Next, the precompiled default odds for the outcome of interest (i.e. the likelihood of a running play) are retrieved based on the determined situation ID (S1004 of
Subsequently, an assessment is made as to whether automatic odds adjustment is to be performed (S1006 of
Based on the proposition in respect of which odds are being determined (i.e. “What will the next play be?”), the odds adjustment formula to be applied is identified (S1008). In the present example, the odds adjustment formula to be applied is equation (3) below:
Where:
-
- Run=probability of a running play (i.e. precompiled default odds) in the current game situation, before odds adjustment
- AdjRun=probability of a running play after odds adjustment
- RunFreq=run frequency for the current team regardless of game situation
- LgAvgRun=league-average run frequency regardless of game situation
It will be appreciated that equation (3) includes a team-specific team factor RunFreq and a league-average factor LgAvgRun. These are specific examples of the factors B and A, respectively, of equation (1) above. The team-specific factor RunFreq is retrieved from the appropriate field of odds adjustment window 100, which will have been filled in by the administrator 19 prior to game time. In the present example, the RunFreq value will be retrieved from either the home run frequency field 112 or visitor run frequency field 116 (
Using the retrieved factors, equation (3) is then applied to effect odds adjustment (S1014 of
This percentage reflects the belief that the current team is more likely the than the average NFL team to execute a running play in the current game situation.
At this stage, the administrator 19 may consider whether manual adjustment of the odds is warranted (S1016). This could occur, for example, in poor weather conditions or toward the end of a game when the result is apparent. In the present example, the administrator 19 has observed that the offensive team is passing more than usual, thus he reduces the likelihood of a running play to 88% (S1018).
Thereafter, the 88% percentage is converted to American style odds, resulting in odds of −733.3 (S1020). Subsequently the juice factor is applied (15% in this case) and rounding is performed, as described previously (S1022). The resulting odds, i.e. −845, are then displayed in the lines panel 56 of control panel interface 50 of administrator computer 18 (
Depending upon the proposition and event outcome for which odds determination is being performed, the formula that is applied when automatic odds adjustment has been elected at S1006 to S1014 of
1. “What will the next play be?”/Running Play
(see equation (3) above)
2. “What will the next play be?”/Completed Pass
Where:
-
- AdjPass=probability of a completed pass after odds adjustment
- PassAcc=pass accuracy adjustment factor determined at the discretion of the administrator using team-specific historical pass completion percentages as a guide line (Pass Completion Percentage=Number of Complete Passes÷Number of Pass Attempts)
- LgAvgPCPct=league-average pass completion “percentage” (actually expressed as a probability)
- Pass=probability of a completed pass (i.e. precompiled default odds) in the current game situation, before odds adjustment
- PassInc=probability of an incomplete pass (i.e. precompiled default odds) in the current game situation
- AdjRun=probability of a running play after odds adjustment
- TO=probability of a turnover (i.e. precompiled default odds) in the current game situation
Per equation (4), the ratio between “Pass Complete” and “Pass Incomplete” is adjusted using a quadratic equation having the same structure as equation (1) above.
3. “What will the next play be?”/Pass Incomplete
AdjPassInc=1−AdjRun−TO−AdjPass; (5)
Where:
-
- AdjPassInc=probability of a incomplete pass after odds adjustment
- AdjRun=probability of a running play after odds adjustment
- TO=probability of a turnover (i.e. precompiled default odds) in the current game situation
- AdjPass=probability of a completed pass after odds adjustment
Equation (5) requires adjusted odds for the probability of running play and the probability of a completed pass to be computed in order for the adjusted odds for an incomplete pass to be computed. The precompiled default odds for a turnover are used in unadjusted form because no odds adjustment is available for the “turnover” outcome in the present embodiment.
4. “Will the [offensive team] get a first down on this series of downs?”/Yes
Where:
-
- AdjFdY=probability of a first down in the current game situation after odds adjustment
- Fdy=probability of no first down (i.e. precompiled default odds) in the current game situation
- Score=projected score for team of interest
- LgAvgNTdPts=league-average non-offensive touchdown points per game
- LgAvgDr=league-average drives per game
- LgAvgTdPts=league-average points per touchdown
- LgAvgFdPct=league-average first down percentage (expressed as a probability)
Equation (6) accounts for a team's projected first down percentage relative to the league average. The “B” factor in equation (6) above is the team-specific first down percentage, which is derived from the team's projected score. A team's projected first down percentage is determined by curve fitting data of all of the league's teams' historical first down percentages for the past N seasons and offensive touchdowns per game (where N is an integer) to a quadratic equation by determining the lowest sum of squared absolute errors. The numeric coefficients of equation (6) were determined by curve fitting to the sum of the least squares between data points. The data points are derived from historical team-specific statistics for first down percentage and offensive touchdowns per game. The projected score may be computed by subtracting a team's spread from the points expected to be scored by both teams in total (over/under) and then dividing the result by two. The latter two factors may for example be based on the spread and total points being offered by sportbooks on the game.
5. “Will the [offensive team] get a first down on this series of downs?”/No
AdjFdN=1−AdjFdY (7)
Where:
-
- AdjFdN=probability of no first down in the current game situation after odds adjustment
- AdjFdY=probability of a first down in the current game situation after odds adjustment
6. “What will the [offensive team] do on this drive?”/Touchdown
Where: - AdjTd=probability of a touchdown for this drive in the current game situation after odds adjustment
- Score=projected score for team of interest
- LgAvgNTdPts=league-average non-offensive touchdown points per game
- LgAvgDr=league-average drives per game
- LgAvgTdPts=league-average points per touchdown
- LgAvgTd=league-average offensive touchdowns per drive
- Td=probability of a touchdown for this drive in the current game situation (i.e. precompiled default odds)
Equation (8) accounts for a team's projected offensive touchdowns per drive relative to the league average.
7. “What will the [offensive team] do on this drive?”/No Score
AdjNS=1−AdjTd−FG (9)
Where:
-
- AdjNS=probability of no score for this drive in the current game situation after odds adjustment
- AdjTd=probability of a touchdown for this drive in the current game situation after odds adjustment
- FG=probability of field goal for this drive in the in the current game situation
For the remaining propositions/outcomes that are set forth above in Table 1, no automatic odds adjustment is available.
Odds adjustment formulae may be used for adjusting odds associated with event outcomes for other sports. For example, in basketball the following equation (10) may be used for the proposition “Which team will get the next 3-point field goal?”:
Where:
-
- Adnx3=the probability that a team will get the next 3-point field goal, after odds adjustment
- 3pm=a team's 3-point field goals made per game
- 3pa=a team's 3-point field goals allowed per game
- opp3pm=the opposing team's 3-point field goals made per game
- opp3pa=the opposing team's 3-point field goals allowed per game
It is noted that, in this case, the odds adjustment equation (10) does not conform to equation (1) above. The term “3-point field goals allowed” refers to the amount of 3-point field goals given up against all opposing teams per game.
As should now be appreciated, the manner of determining odds described above allows odds determination to be performed quickly, due to the used of precompiled default odds which are quickly retrievable based on current values of dynamic contest parameters. As well, when it is desired to adjust the odds to incorporate team-specific deviations from league averages, odds determination may be expedited due to the automatic application of odds adjustment formulae whose team-specific factors have been previously entered.
As will be appreciated by those skilled in the art, modifications to the above-described embodiment can be made without departing from the essence of the invention. For example, although the above-described embodiment pertained to odds determination in the context of an American football game, the same odds determination technique could be used for event outcomes for other types of sport, such as individual participant sports (e.g. golf or tennis), or for contests other than sporting contests, such as elections for example.
The above described embodiment refers to the use of “dynamic contest parameters” to perform table lookups of situation-specific likelihood data. It should be appreciated that the term “dynamic contest parameters” is not necessarily limited to parameters which are traditionally displayed on scoreboards. For example, parameters such as weather conditions or degrees of player injuries could form the basis of situation-specific likelihood data lookups.
The above-described embodiment also employs the term “league-average”. This term should not be understood to necessarily connote an averaging calculation. Some embodiments may for example employ a median determination.
For the sake of simplicity, odds adjustment calculations may take the form of a straight multiplier rather than a quadratic equation as previously described. For example, if a specific football team is known to be 30% more likely to run than the league average of 40%, the league average may be adjusted by multiplying it by 130%, to arrive at a team-specific likelihood of 52%. This type of odds adjustment may be advantageous in its simplicity; however it may not always produce acceptable results. For example, when the likelihood of an event outcome for the league overall is high, multiplication by a scalar may cause the resultant team-specific likelihood to exceed 100%—a statistical impossibility.
It is not necessary to employ situation IDs for uniquely identifying game situations. In alternative embodiments, current parameter values of dynamic contest parameters may be used directly to access tables in which situation-specific league statistics are stored. The statistics may take the form of multi-dimensional tables.
Precompiled default odds may be generated for a sport and league by examining historical records of any number of past seasons. For contests which do not have annually occurring seasons (e.g. Olympic competitions or political elections), a “season” may refer to an interval between contests (or between logical contest groupings). Intervals could be larger than a year or smaller than a year.
It will also be appreciated that database 16 need not necessarily be a SQL server database or even a relational database. Other forms of data stores for the various types of data associated with the electronic wagering system 10 described above could be employed. The alternative data stores need not necessarily be in same location as bet server 14.
The situation-specific league statistics 218 (
Bet server 14, broadcast server 26, bettor computer 22 and administrator computer 18 are not necessarily servers and computers in the traditional sense. Various forms of computing devices having processors in communication with memory capable of storing machine executable code and displays capable of displaying a user interface, such as laptop computers, tablet computers, or personal digital assistants for example, could be employed, provided that they are capable of operation in the manner described above.
Numerous further modifications and variations of the present invention are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims, the invention may be practiced otherwise than as specifically described herein.
Claims
1. A method of determining odds of a possible outcome of an event which occurs during a contest, comprising:
- accepting a set of one or more current parameter values, a likelihood of said possible outcome of said event being dependent upon each parameter;
- utilizing said set of current parameter values to retrieve likelihood data from a data store; and
- setting odds associated with said possible outcome of said event based at least in part on said likelihood data.
2. The method of claim 1 wherein said setting odds further comprises adjusting said likelihood data, X, to obtain adjusted likelihood data Y, utilizing at least one factor indicating a mean or median likelihood of said event outcome for a plurality of different sets of parameter values.
3. The method of claim 2 wherein said at least one factor comprises a first factor, A, obtained by determining a mean or median likelihood of said event outcome for a plurality of different sets of parameter values across a group of entities which have performed said event.
4. The method of claim 3 wherein said at least one factor further comprises a second factor, B, obtained by determining a mean or median likelihood of said event outcome for a plurality of different sets of parameter values for an entity of said group of entities which is performing said event.
5. The method of claim 4 wherein said adjusted likelihood data, Y, is related to said unadjusted likelihood data, X, by the formula Y=((B−A2)/(A−A2))×X+(1−((B−A2)/(A−A2)))×X2.
6. The method of claim 5 wherein said contest is a sport and each entity of said group of entities is a participant in said sport.
7. The method of claim 6 wherein each said participant is an individual participant where said sport is an individual sport and a team where said sport is a team sport.
8. The method of claim 1 wherein said method is for determining odds of all possible outcomes of said event and wherein said setting odds is for setting odds associated with each possible outcome of said event based at least in part on said likelihood data.
9. The method of claim 8 further comprising, responsive to said setting odds, populating a display with an odds indication for each possible outcome of said event.
10. The method of claim 9 further comprising:
- establishing a given odds indication for a given possible outcome as an automatically adjusted odds indication;
- accepting manual adjustment of any said odds indication; and
- on a manual adjustment of an odds indication other than said automatically adjusted odds indication, commensurately adjusting said automatically adjusted odds indication to maintain a certain cumulative probability of all possible outcomes of said event.
11. The method of claim 10 further comprising juicing each said odds indication.
12. The method of claim 11 further comprising, on receiving a user prompt, exporting an indication of said event and each said odds indication.
13. The method of claim 11 further comprising:
- on receiving a first user prompt, providing an indication of said event and each said odds indication;
- receiving bets on at least one possible outcome of said event; and
- on receiving a second user prompt, ceasing to receive bets on any possible outcome of said event.
14. The method of claim 7 wherein said sport is American football and wherein said parameter values are values for field position, down, and yards to go to the next first down.
15. A computer system for determining odds of a possible outcome of an event which occurs during a contest, comprising:
- a user interface;
- a display;
- a data store; and
- a processor for: receiving a set of one or more current parameter values from said user interface, a likelihood of said possible outcome of said event being dependent upon each parameter; utilizing said set of current parameter values to retrieve likelihood data from said data store; setting odds associated with said possible outcome of said event based at least in part on said likelihood data; and displaying an indication of said odds on said display.
16. A machine-readable medium encoded for determining odds of a possible outcome of an event which occurs during a contest, comprising:
- machine executable code for accepting a set of one or more current parameter values, a likelihood of said possible outcome of said event being dependent upon each parameter;
- machine executable code for utilizing said set of current parameter values to retrieve likelihood data from a data store; and
- machine executable code for setting odds associated with said possible outcome of said event based at least in part on said likelihood data.
17. The machine-readable medium of claim 16 wherein said machine executable code for setting odds further comprises machine executable code for adjusting said likelihood data, X, to obtain adjusted likelihood data Y, utilizing at least one factor indicating a mean or median likelihood of said event outcome for a plurality of different sets of parameter values.
18. The machine-readable medium of claim 17 wherein said at least one factor comprises a first factor, A, obtained by determining a mean or median likelihood of said event outcome for a plurality of different sets of parameter values across a group of entities which have performed said event.
19. The machine-readable medium of claim 18 wherein said at least one factor further comprises a second factor, B, obtained by determining a mean or median likelihood of said event outcome for a plurality of different sets of parameter values for an entity of said group of entities which is performing said event.
20. The machine-readable medium of claim 19 wherein said adjusted likelihood data, Y, is related to said unadjusted likelihood data, X, by the formula Y=((B−A2)/(A−A2))×X+(1−((B−A2)/(A−A2)))×X2.
Type: Application
Filed: May 23, 2005
Publication Date: May 4, 2006
Inventors: Ronnie Tarter (Jolly Harbour), Greg Oliver (Toronto), Gabriel Heskin (Jolly Harbour), David Ross (Toronto)
Application Number: 11/135,731
International Classification: A63F 9/24 (20060101);