GB2421812A - Predicting outcomes of staged processes, eg sporting events - Google Patents

Abstract

A system for generating data indicative of a predicted outcome relating to a staged process such as a boxing or football match or a microprocessor calculation, the staged process involving one or more resources, each having parameters associated therewith, and comprising one or more stages, each stage having a plurality of possible occurrences associated therewith, the system comprising: ```a storage system arranged to store first data indicative of generalised probabilities associated with an occurrence and second data indicative of said parameters; ```a processing system arranged to retrieve said first and second stored data so as to modify the generalised data whereby to generate data indicative of probabilities particular to the or each resource for at least one occurrence of each stage, ```wherein the processing system is arranged to combine the probabilities associated with the stages so as to generate data indicative of a probability of at least one outcome for the process.

Description

242-1812 ytem and Method for Predicting Outcomes

Field of the Invention

The present invention relates to a method and system for generating data indicative of a predicted outcome relating to a staged process, the staged process involving two or more resources and comprising one or more stages, each stage and/or staged process having a plurality of possible occurrences associated therewith. The system is particularly, but not exclusively, suitable for predicting outcomes of random processes, such as sporting events. In addition, the present invention relates to a system for processing transactions relating to outcomes that have been selected from at least some such possible outcomes.

Background of the Invention

There are many methods and systems for predicting outcomes of events that inherently comprise a degree of randomness. Typically outcomes are computed by making use of empirical data, in which actual event instances have been observed and outcomes measured; or by means of simulations, in which the event domain is modelled and events occurring within the domain are simulated. Examples of these sorts of events include sports matches (e.g. football, boxing, horse racing and the like) and predictive pipeline branching processes executed by microprocessors.

A problem with adopting the simulation approach is that performing the simulations can be extremely processor-intensive and time consuming, whilst a problem with the empirical approach is that it is not always possible to perform the required measurements. In the case of simulations, because they are so time consuming, the computations are often forcibly halted before they have reached the required degree of accuracy. As a result the prediction data can have a significant amount of error associated therewith.

There is therefore a need for a more flexible method of predicting outcomes of events.

Summary of the Invention

In accordance with a first aspect of the present invention, there is provided a method of generating data indicative of a predicted outcome relating to a staged process, the process involving one or more resources and comprising one or more stages, each of which having a plurality of possible occurrences associated therewith, the method comprising: performing a process comprising: retrieving data indicative of generalised probabilities associated with an occurrence; retrieving data identifying parameters associated with the or each resource; and modifying the generalised data in accordance with the parameters so as to generate data indicative of probabilities particular to the or both resources for the occurrence.

Advantageously, the generalised probability data are stored in a repository (essentially being expert data) for use in generating prediction data, the data in the repository being customised for individual resources, or participants of, the staged process on the basis of parameters associated with the resources. Preferably the process is repeated for two or more occurrences in a stage and for a plurality of stages of the staged process. An overall outcome for the staged process can be evaluated by combining the probabilities associated with the stages so as to additionally generate data indicative of a probability of an outcome for the staged process.

Interactive staged processes will typically involve two or more resources: for example staged computer processes (such as a calculation) will involve use of CPU and memory resources; and sporting processes such as boxing will involve use of two participants. In the case of a boxing match, a staged process is a match and a stage is a round of the match, so that the probability data generated in respect of individual rounds are used to determine the probability of certain outcomes of the match. In the case of a computer process, a staged process might be a specified calculation routine and a stage might be a step of the overall routine.

According to a second aspect of the present invention there is provided a transaction system for processing transactions relating to one or more outcomes of a staged process, wherein the staged process comprises one or more stages, the transaction system comprising: a storage system arranged to store a plurality of records, each record comprising first probability data indicative of probabilities of one or more outcomes of said stage and an outcome of the staged process, each said outcome being indexed by one or more selectable parameters, a processing system arranged to retrieve a record in response to data indicative of selected parameters, the processing system being arranged to generate second probability data indicative of odds associated with the outcome of a stage and of the staged process on the basis of the first probability data held by said retrieved record and to transmit, to a terminal remote from said processing system, data indicative of said second probability data.

Thus when an estimate of one or more possible outcomes of, say, a boxing match, is required, characteristics of the participants in the match are identified, outcome data corresponding to these participants are retrieved from the storage system, and the processing system generates odds data corresponding to these outcome data for one or more rounds of the match in addition to outcome data for the overall match. The probability data associated with an outcome of the match are derived from the probability of outcomes of the rounds of the match, which means that the outcome data for the rounds and the match are mathematically connected, resulting in a mathematically correct evaluation of outcome data for the match. Since generating the odds data only involves retrieval and simple manipulation of these data - which generally have been generated off-line - lengthy and complex simulations in real time can be avoided.

Further features and advantages of the invention will become apparent from the following description of preferred embodiments of the invention, given by way of example only, which is made with reference to the accompanying drawings.

Brief Description of the Drawings

Figure 1 is a schematic diagram of an operating environment for embodiments of the invention; Figure 2 is a schematic diagram showing components of first and second server computers shown in Figure 1 configured in accordance with embodiments of the invention; Figure 3 is a flow diagram showing steps involved in identifying probabilities of occurrences associated with staged processes; Figure 4a is a table showing data indicative of probabilities of occurrences associated with an staged process; Figure 4b is a mapping showing the relationship between occurrences and bettable staged processes; Figure 4c is a table showing data indicative of probabilities of bettable staged processes; Figure 5 is a schematic diagram showing components of the events engine shown in Figure 2; and Figure 6 is a flow diagram showing steps involved in processing a staged process in respect of which bets have been placed according to a second aspect of the invention.

Detailed Description of the Invention

Embodiments of the invention are concerned with methods of, and systems for, predicting outcomes of staged processes that inherently comprise a degree of randomness. Examples of these sorts of staged processes include sports matches (e.g. football, boxing, horse racing and the like) and predictive pipeline branching processes executed by microprocessors, among others.

Broadly speaking, in embodiments of the invention a repository of "expert knowledge" is maintained, which has been populated from off-line measurements and observations of actual staged processes. When an estimate of one or more possible outcomes of a specific staged process is required, data are retrieved from the repository, characteristics of the specific process are analysed, and the retrieved data are modified by means of an algorithm in accordance with the characteristics particular to this process. This means that lengthy and complex simulations can be avoided.

Embodiments of the invention are also concerned with a distributed transaction system in which odds relating to specific outcomes of a staged process are calculated, bets are offered and placed in respect of the outcomes, the process is performed and bets can be settled on the basis of the outcome of the performed process and bets that have been placed. The distributed transaction system generates odds data on the basis, preferably, of the outcome data generated in the manner described above. Distributed transaction systems arranged according to the invention are capable of calculating odds in relation to outcomes of one or more stages of a staged process in addition to an outcome in relation to an overall result of the process. Such a distributed transaction system is particularly well suited to staged processes such as sports matches, which comprise a plurality of stages (e.g. in boxing, each boxing match comprises a plurality of rounds (each round being a stage); in football, each game comprises a plurality of "halves" (each half being a stage)), and to staged events such as games comprising a plurality of levels that the participant has to work through in order to complete the game (each level being a stage).

Embodiments of a first aspect of the invention will now be described with reference to Figures 1 to 3, which present components of a distributed transaction system 101 according to the invention. Figure 1 is a schematic diagram showing two server terminals Si, S2, a first of which Si is arranged to compute outcome data in relation to a staged process and in relation to stages of such a process, and to store these data in a data storage system DB I such as a database. The second server S2 is arranged to generate odds data on the basis of the data stored in the storage system DB1 in relation to specific processes and characteristics thereof. In addition the server S2 is arranged to execute processes and stages of processes (under the control of a computer). The server terminal S2 is responsive to requests from client terminals Cl, C2, C3, received via network Ni so as to output data indicative of generated odds, of executed processes and stages of such processes, and to receive data indicative of bets placed in respect of a process.

As can be seen from Figure 2, server terminal Si comprises processing unit (CPU), memory 215, disc storage 213 and 110 adapter device, which facilitates interconnection of the terminal SI with other such devices. Operating system programs are stored in the disc storage 213, and when loaded into memory control, in a known manner, low level operation of the terminal SI.

The various components are interconnected in a known fashion. Embodiments of the invention comprise outcome engine 201 which comprises probability evaluating software component 203 and database software 205, together with a repository (here DBI shown in Figure 1) arranged to store expert knowledge in relation to possible outcomes of staged processes; the repository DB1 is accessible by probability evaluating software 203 by means of the database software component 205 in accordance with conventional query and retrieval procedures. The repository DB1 will typically be provisioned and updated by means of a third party and the contents of this storage system DB 1 is referred to as "expert knowledge" because it comprises data indicative of the average, normally distributed, outcome of a stage of a process. The probability evaluating software component 203 additionally has access to data indicative of parameters that are characteristic of particular processes; for example, in the case where the process is a sports match such as boxing, the parameters include characteristics of boxers in terms of their capacities or performance levels (which can be an absolute value or a relative value) and optionally how their performance varies over time (this being particularly well suited to evaluating the outcomes of individual stages of the process (in this example rounds of a boxing match)). These data are referred to hereinafter as parameter data. The interrelationship between the expert knowledge and parameter data will now be described with reference to Figure 3 for an example of a staged process being a boxing match.

A boxing match is taken to comprise a plurality of rounds, and within each round there is a plurality of possible occurrences such as Boxer 1 winning a round by 10/9; Boxer 1 winning by 10/8; Boxer 1 winning by 10/7 with knockdown; etc. (and similarly for Boxer 2). For each of these possible occurrences, database DB 1 stores a probability interval (e.g. 25% 30%), which defines a range within which the probability of the occurrence lies. Each value within the range relates to a probability of the occurrence and corresponds to a particular combination of parties participating in the event; in order to associate actual combinations of parties with actual (rather than range) probability values, the outcome engine 201 uses characteristics of individual participants - the parameter data - to identify an actual value within that range. These parameter data are referred to as al, a2 in the following description and in Figure 3.

Accordingly, at step 301, the outcome engine 201 selects an outcome for the first round (e.g. red boxer to win round 1 by 10/9) and retrieves (step 303) a range corresponding to this outcome from the repository of expert knowledge (e.g. 65% - 70%). The outcome engine 201 then selects a first set of participants, represented by a corresponding set of parameters (step 305) and, at step 307, modifies the range in accordance with the selected parameters so as to identify (step 309) a value within the range indicative of the probability of the red boxer winning round I by 10/9. In this example, values of parameter al associated with the respective participants are used to modify the selected range, and the modification is based on the relative weightings between the individual values of al associated with the selected participants. Once a probability value has been generated, the probability is stored (step 311) in a probability file 221a in association with whichever parameter value gave rise to the probability. This process is repeated for a different set of participants (i.e. different values of al (selected cyclically at step 313)), each resulting in different values in the range of 65% and 70% being generated (at step 309). This method therefore creates a link between value of al and likely outcomes. Once all possible values of al have been used, the method is repeated for different outcomes (selected cyclically at step 315) and indeed different rounds (selected cyclically at step 317).

The parameter discussed thus far, al, is a static capacity parameter, which is to say that it applies independently of the stage in the staged process. In order to factor in changes in capacity over time, a further parameter, a2, can also be used; this might be important if, for example, one of the participants has greater stamina than the other participant(s) and if the process proceeds into several or many stages (in which case stamina might be expected to become significant). Accordingly, after each stage, i.e. at step 317, a time- dependent function can be applied in order to determine a value for a2, which, in turn, is used to select a value within the range. This time- dependent function preferably applies an offset to a parameter a! in dependence on the number of stages that have been completed, and each participant, or value of al, might have a different such function (a2) associated therewith. In one arrangement the stage-dependent function takes the following form: Round(i) 1 2 3 4 5 6 7 8 9 10 11 12 FUflCtOflMm 0 0 0 0 0 0 1 1 -2 -2 -2 -2 FUflCtJOflMax 0 0 0 0 0 0 0 0 0 1 2 2 Function 1 where the values indicate an amount by which parameter al is offset (within the range associated with an occurrence).

This method therefore provides a means of indexing the likelihood of an outcome by parameter values al, a2, and means that retrieval of probability values for an outcome in respect of particular participants (having particular capability characteristics al, a2) - at a later date becomes a simple matter of retrieving a probability file from storage 221 on the basis of the parameter data al, a2 used to index the probability file 221 at step 311.

An example of a first probability file 221a that has been generated for a staged process and for one set of parameters is shown in Figure 4a; the notation used to denote different types of occurrences is defined in the accompanying key. Each column 4011 relates to a stage (round) of the process (boxing match) and each row 4O3 relates to a possible outcome for (or occurrence in) a round.

The outcome engine 201 is additionally arranged to process the first probability file so as to generate probabilities relating to combinations of outcomes, which can subsequently be offered as bet-able events. Examples of bet-able occurrences include a knockdown, a draw, a score (irrespective of winner), and then winner-specific outcomes such as Red boxer to win by 10/8; Blue boxer to by knocked out etc. The probability of each of these bet-able outcomes can be computed for each round based on data in the first probability file 221a and in the maimer schematically indicated in Figure 4b. For example, a bet can be placed for Red to win a round by 10/8 (RX8), and the probability of this outcome is computed from a combination of PLrx8) and PLrx8d), both of which are stored in the first probability file 221a.

So far we have considered bet-able occurrences in relation to individual stages (rounds) of a process (match). Embodiments of the invention can also be used to generate the probability of an overall match result. Such overall match results include, for example, red winning by a knockout; blue winning by a knockout in Round 3 and so on. The probabilities associated with an outcome of the overall process can either be calculated using AND Boolean logic (based on data in the probability file 221 a) or by simulating an entire event (match) a specified number of times and evaluating the probabilities of the various match outcomes from the simulation data.

The probability of the match ending with a Red knockout can be determined from the probability data indicative of red winning and the occurrence of a knockout; however, the probability of a match ending with a particular score cannot be evaluated from a Boolean combination of occurrences, and instead the entire simulation approach is used. The entire simulation method will be described later, in relation to the second aspect of the invention (processing staged events upon which bets have been placed), so it will simply be assumed at this point that the probabilities of overall match occurrences have been calculated. It will be appreciated from the foregoing that the probabilities associated with an overall outcome of the staged process (i.e. a match result) is directly linked to, and thus dependent on, the probabilities of outcomes for individual stages of the process (i.e. results of a round).

Once probabilities corresponding to the bet-able staged events have been computed, the outcome engine 201 stores the probability data in second probability file 221b, which, as for the first probability file 221a, is indexed in accordance with the associated parameters. An example of a second probability file 221b is shown in Figure 4c.

A second aspect of the present invention will now be described, which is concerned with use of the first probability files 221a stored at step 311 and second probability files 221b derived therefrom. It will be appreciated from the foregoing that each of the probability files 221a, 221b stores data representative of the likelihood of a particular outcome for a particular set of parameters. This information can be used to generate predictions for a variety of applications, one of which relates to betting, in particular betting in relation to outcomes of staged sporting events.

Referring back to Figures 1 and 2 client terminals Cl, C2, C3 represent computing devices configured to communicate with server S2 in order to place bets on one or more staged processes andlor individual stages of the process.

Each client terminal Cl... C3 can be any one of an internet-enabled PC, a mobile phone, a PDA or the like, and preferably the server S2 comprises a web server 303 arranged to accept requests for data from the client terminals Cl C3 (e.g. input via a browser running on the client terminal in the manner well luiown in the art). The server S2 also includes an application server 305, which receives data from the client terminals via the web server 303, passing the data to and from the events engine 307. The events engine 307 is arranged to generate and transmit odds data in response to requests from a client terminal Cl, to receive data indicative of a bet having been placed in respect of a particular stage of a process andlor the overall process, to invoke a sporting event, to evaluate winnings in relation to the invoked event and to transmit an appropriate winning notification to terminal Cl.

Operation of the events engine 307 will now be described in more detail with reference to Figures 5 and 6 showing, respectively, components of, and steps performed by, the events engine 307 when co-operating with a client terminal Cl. Turning firstly to Figure 5, in one arrangement the events engine 307 comprises a communications module 500 arranged to control input and output of data between the engine 307 and other components of the server S2 as well as coordinating communication between components of the events engine 307 itself. The events engine 307 also includes a plurality of software modules, each designed to perform a specific function, and thus conveniently embodied as an object. These objects include staged process object 501, stage object 503, occurrence object 505 and odds object 506; the staged process object 501 is arranged to select participants of the staged process together with characteristics of the participants, i.e. al and a2, (from the store 321), and to select probability data (from files 221a, 221b) on the basis of the values of al and a2. The participant data 321 thus relate, directly, to the parameters used to create and index probability data 221a, 22lb according to the first aspect of the invention.

The staged process object 501 is additionally arranged to pass the probability data to the odds object 506, which is arranged to evaluate odds corresponding to the probability data (described in detail below). The communications module 500 is responsible for communicating the odds data to client terminals Cl... C3 and for forwarding data, indicative of bets that have been placed, to the occurrence object 505. It is to be noted that odds are calculated in relation to stages of the process (in the case of boxing, a stage is a round) in addition to an overall outcome of the process per se (i.e. of a match).

The staged process object 501 is also operable to execute a process in respect of which bets have been placed and accordingly is operable to invoke a plurality of stage objects 503, each corresponding to a stage of the process (e.g. as described above, in boxing each match (staged process) has multiple rounds (stages)). The status of any bet that has been placed is stored and updated by object bet:Occurrence 509. During progress of a staged process, the status of the overall process and a stage thereof can be managed either by the associated objects 501, 503 respectively or, more preferably, by objects state:Process object 511 and state:Stage object 513. In the present embodiment status information is maintained and updated by objects state:Process object 511 and state:Stage object 513.

Turning now to Figure 6, operation of the event engine 307 will be described by means of an example in which the process is a boxing match comprising a maximum of 12 stages in the form of rounds. At step 601 the events object 501 selects two participants, either at random or according to a specified selection procedure, and then selects probability data, from store 221, corresponding to the selected participants (step 603). These steps involve firstly querying store 321 to retrieve parameters corresponding to the selected participants and then querying store 221 in respect of the retrieved parameters.

Referring back to Figure 4c, data indicative of the probabilities associated with each of the bettable events listed in column 411i, for both the match and individual rounds Ri... R12, are retrieved. These data are then passed to odds object 506, which calculates odds associated with each occurrence based on the probability data listed in probability file 221b (step 605) and a specified take (e.g. 5%, 10%), in a manner known in the art.

Having calculated the odds for each stage of the process and for the process itself, those for a first round Ri and for a match result are transmitted to client devices Cl, C2, C3 (via communication module 500 and web server 303) for display (step 607). Users then indicate whether or not they wish to place a bet in respect of the round and indeed the match, and assuming that a bet is placed, data indicative of the magnitude and outcome of the bet, together with the user (human or software agent) placing the bet are then transmitted to the web server 303 and passed to the events engine 307 via the application server 305 (step 609). The data indicative of the bet are stored in bet:Occurrcnce object 509, for use in settling bets, as will be described in more detail below.

The events engine 307 then processes the next (in this case first) stage of the process (step 611). There are many ways in which a stage can be executed, and in a preferred embodiment this involves selecting an occurrence listed in probability file 221a at random. This process can best be described by means of an analogy in terms of lotto balls. Referring to Figure 4a, each occurrence is assigned an integer number of balls of a specified colour, and in a proportion dictated by the probabilities specified in the table. Thus, referring to column 4011 the number of balls assigned to occurrence _rxx will be 2; the number assigned to occurrence _rx9 20; the number assigned to occurrence _rx8 15; the number assigned to _rx8d 1 and so on, for each occurrence. Having made the assignments, all of the balls are placed into a pool and one is selected at random from the pool. More specifically, the staged process object 501 will trigger stage object 503 corresponding to the first round Ri to select a ball at random, and once a ball has been selected, the corresponding occurrence colour, and thence occurrence 403k, will be identified.

The identified occurrence is then used to derive which actual outcome(s) correspond(s) to the occurrence (step 612) by means of the Boolean relationships shown in Figure 4b: - general bets - Bettable name= Draw ("XX") events= _rxx OR _bxx Bettable name= Score 10/9 ("X9") events= _rx9 OR bx9 Bettable name= Score 10/8 ("X8") events= _rx8 OR _rx8d OR bx8 ORbx8d Bettable name= Score 10/7 ("X7") events= _rx7 OR rx7d OR bx7 ORbx7d Bettable name= Knockdown ("D") events= _rx8d OR rx7d OR _bx8d OR bx7d Bettable name= Knockout ("0") events=rx7o OR _bx7o - red bets - Bettable name= Red ("R") events= _rx9 OR _rx8 OR _rx8d OR _rx7 OR _rx7d OR rx7o Bettable name= Score 10/9 ("RX9") events= rx9 Bettable namc= Score 10/8 ("RX8") events= _rx8 OR rx8d Bettable name= Score 10/7 ("RX7") events= rx7 OR rx7d Bettable name= Red Knockdown ("RD") events= _rx8d OR _rx7d Bettable name= Red Knockout ("RO") events= _rx7o - blue bets - Bettable name Blue ("B") events= _bx9 OR _bx8 OR _bx8d OR _bx7 OR _bx7d OR bx7o Bettable name= Score 10/9 ("BX9") events= _bx9 Bettable name= Score 10/8 ("BX8") events= bx8 OR _bx8d Bettable name= Score 10/7 ("BX7") events= _bx7 OR _bx7d Bettable name= Blue Knockdown ("BD") events= _bx8d OR bx7d Bettable name= Blue Knockout ("BO") events= _bx7o Once the actual outcome has been identified, the staged process object 501 compares the outcome against the data stored in the bet:Occurrencc object 509, in order toidentify those users who placed bets corresponding to the actual outcome (step 613). In one arrangement the bet:Occurrence object 509 might maintain a record of previous winnings and the like in respect of each registered user, and, for each user who placed a bet for the current round, the bet:Occurrence object 509 adds or subtracts from the record in dependence on whether or not the user won or lost the bet.

Having settled the bets the event engine 307 then determines whether or not there are any further rounds to be executed: this is dependent on the result of the just-completed round (since, if this were round 12 then this would be the end of the match; and if the result this round were a knockout the match would have terminated before all of the stages have been completed). If further stages are to be processed the staged process object 501 returns to step 607 and repeats these steps in respect of the next and subsequent stages of the process.

If, on the other hand, the completed stage corresponds to the end of the staged process, the event engine 307 collates all of the occurrences selected from the probability file 22la at successive instances of step 611 (i.e. for each stage of the event thus far) in order to calculate an overall match outcome (step 615). The overall match outcome calculated at step 617 is then compared with the bets received at step 609 (those corresponding to an overall staged process outcome) and the bet:Occurrence object 509 updated accordingly.

It will be appreciated that the history of previous staged processes should be taken into account when characterising individual participants, since it is the relative capabilities of participants that determines the values in probability files selected at step 603. Accordingly, once a staged event has ended, the parameters corresponding to the participants selected at step 601 can be updated in accordance with the match outcome (step 619). In general, the parameters reflect the participant's experience such that their parameters are dependent on the number of events a participant competes in and wins. In one arrangement parameter al can be updated according to the following formula: al MIN(al + ((total won/N)*w1), 1.0) (1) where: N is the maximum number of events in which the participant can be involved before he is removed from the system; and wi is a weight having a value selected so as to assign more importance to the number of events in which the participant has been involved. A similar, event-related, expression can be derived for the other parameter, a2, the expression essentially modifying the values of Function 1 set out above in dependence on the number of events in which the participant has been involved.

Turning back to Figure 5, it will be appreciated that, having updated the parameters (al andlor a2), the database 321 will be updated to include data identifying the new performance parameters, together with data indicative of the total number of matches participated in and the number of matches won.

As an alternative to updating and storing the value of al and/or a2 after upon completion of a staged process (at step 619), and because data identifying the number of matches won and competed in are, in any event, stored in the database at step 619, the parameter value(s) could alternatively be updated after participants have been selected at step 601. In this approach the probability files 221a selected at step 603 will be dependent on the updated performance parameters rather than those stored in the database 321.

Additional Details In the description above it is stated that the probability of an overall result of a staged process is calculated by simulating an entire staged process. In a preferred embodiment and referring to the description relating to step 611, a simulation involves selecting an occurrence at random (e.g. by the lotto method) in relation to each stage of the process, repeating this process a specified number of times (e.g. 20,000) and recording the outcome in respect of each stage of each event. The probabilities associated with various staged event outcomes are then estimated on the basis of such simulated outcomes.

Whilst in the above description server SI and S2 are shown as being logically and physically separate, they could instead be integrated within one terminal; alternatively the various functional components could be distributed over a plurality of different terminals. In one arrangement the probability data 221a, 221b are stored as XML files.

The above embodiments are to be understood as illustrative examples of the invention. Further embodiments of the invention are envisaged. In particular, embodiments of the invention can be used to evaluate outcomes and probabilities of outcomes for processes comprising only a single stage, such as horse racing, and for processes that comprise a plurality of stages but for which each stage relates to a particular incident of the process. An example of the latter would be football, where a football match might comprise a plurality of corners, free kicks, penalties etc., each of which represents a type of incident occurring within any given match, and in respect of which a bet could be placed.

It is to be understood that any feature described in relation to any one embodiment may be used alone, or in combination with other features described, and may also be used in combination with one or more features of any other of the embodiments, or any combination of any other of the embodiments.

Furthermore, equivalents and modifications not described above may also be employed without departing from the scope of the invention, which is defined in the accompanying claims.

Claims (15)

1. Claims 1. A system for generating data indicative of a predicted outcome

   relating to a staged process, the staged process involving one or more resources, each having parameters associated therewith, and comprising one or more stages, each stage having a plurality of possible occurrences associated therewith, the system comprising: a storage system arranged to store first data indicative of generalised probabilities associated with an occurrence and second data indicative of said parameters; a processing system arranged to retrieve said first and second stored data so as to modify the generalised data whereby to generate data indicative of probabilities particular to the or each resource for at least one occurrence of each stage, wherein the processing system is arranged to combine the probabilities associated with the stages so as to generate data indicative of a probability of at least one outcome for the process.
2. 2. A method of generating data indicative of a predicted outcome relating to a staged process, the process involving one or more resources and comprising one or more stages, each of which having a plurality of possible occurrences associated therewith, the method comprising: performing a process comprising: retrieving data indicative of generalised probabilities associated with an occurrence; retrieving data identifying parameters associated with the or each resource; and modifying the generalised data in accordance with the parameters so as to generate data indicative of probabilities particular to the or both resources for the occurrence.
3. 3. A method according to claim 2, including repeating the process for two or more occurrences in a stage.
4. 4. A method according to claim 3, including repeating the process for occurrences relating to two or more stages of the staged process.
5. 5. A method according to any one of claim 2 to claim 4, including storing the generated data.
6. 6. A method according to claim 5 dependent on claim 4, including: for each stage performing a further process comprising: randomly selecting an occurrence; and retrieving the probability associated therewith, combining the probabilities associated with the stages so as to generate data indicative of a probability of an outcome for the staged process; and repeating the further process until a predetermined terminating condition has been met.
7. 7. A method according to claim 6, including storing the probability data.
8. 8. A method according to any one of claim 2 to claim 7, in which data are retrieved in respect of two resources, a first resource having a first set of parameters associated therewith and a second resource having a second set of parameters associated therewith, and the step of modifying the generalised data comprises applying an offset to the generalised data by an amount dependent on the relationship between said first set of parameters and said second set of parameters.
9. 9. A method according to claim 8, including evaluating the difference between respective parameters in said first and second sets in order to derive said relationship.
10. 10. A transaction system for processing transactions relating to one or more outcomes of a staged process, wherein the staged process comprises one or more stages, the transaction system comprising: a storage system arranged to store a plurality of records, each record comprising first probability data indicative of probabilities of one or more outcomes of said stage and an outcome of the staged process, each said outcome being indexed by one or more selectable parameters, a processing system arranged to retrieve a record in response to data indicative of selected parameters, the processing system being arranged to generate second probability data indicative of odds associated with the outcome of a stage and of the staged process on the basis of the first probability data held by said retrieved record and to transmit, to a terminal remote from said processing system, data indicative of said second probability data.
11. 11. A transaction system according to claim 10, wherein the transmitted data are indicative of second probability data relating to a plurality of stages of the staged process.
12. 12. A transaction system according to claim 10 or claim 11, wherein the processing system is arranged to receive data indicative of a selected outcome after transmitting said data indicative of the second probability data to the terminal.
13. 13. A transaction system according to claim 12, wherein the processing system is arranged to process the, or each, stage of the staged process in response to receipt of selected outcome data so as to generate data indicative of an actual outcome.
14. 14. A transaction system according to claim 13, wherein the storage system is arranged to update at least some of said parameters associated with a record on the basis of a predetermined function and the actual outcome data.
15. 15. A transaction system according to any one of claim 10 to claim 14, wherein the storage system is operable to process the method according to any one of claim 5 to claim 9, so as to generate a said record.