AU2003236561B2 - Predictive analysis system and method - Google Patents

Description

WO 2004/003815 PCTIAU2003!000831

I

"PREDICTIVE ANALYSIS SYSTEM AND METHOD" Technical field This invention relates to a predictive analysis system and method.

The invention has particular but not exclusive application to such a system and method when applied to sports coaching and particularly the coaching of a sporting team or individual.

In a preferred aspect of the invention it will be seen that real-time performance and predictive reports, risk assessments and recommendations are produced which relate to the performance of a sporting team in an actual sporting event. These are prepared by a scientific risk management-based analytical, statistical and predictive system which records and analyses data and information about and from sporting events.

Background of Invention It is known to statistically analyse sporting events and performances, whether by a team or by an individual. Quantitative statistical systems are known and are continually being enhanced.

Summary of Invention The present invention aims to provide an alternative to known methods and systems of the above type.

This invention in one aspect resides broadly in a method of coaching a sporting team or individual participants during a sporting event, the method including:generating quantitative data relating to performance in previous sporting events; generating quantitative data relating to performance during the sporting event; generating qualitative data relating to the sporting event and to the participants therein, and WO 2004/003815 PCT/AU20031000831 2 analysing the generated quantitative and qualitative data in a computer controlled system in accordance with a program which includes a risk management methodology to thereby generate an output of predicted strategic significance in the future conduct of the sporting event.

As used herein the expression "quantitative data relating to performance in a sporting event" is to be given a broad meaning and includes data relating to measurable parameters and indicators of individual and team performance.

As used herein the expression "qualitative data relating to performance in a sporting event" is to be given a broad meaning and includes data relating to intangible aspects relating to the sporting event and to participants therein, both prior to and during the event and relating to both the team and individual players. Examples of these intangible aspects include cultural characteristics, mental aspects, environmental factors and psychological impacts and insights.

It is preferred that the analysis of the qualitative data includes regression analysis techniques.

One output of predicted strategic significance can include prediction of a break point in the sporting event.

As used herein the expression "break point in a sporting event" refers to a stage in the sporting event beyond which, if the conduct of play were to continue in the existing pattern, it is highly likely that one team or individual will win the event or a discrete component of the event. The expression "discrete component" of an event refers to components such as a game or a set in a tennis match, a quarter or a half in a football match, an innings in a baseball match etc.

Another output of predicted strategic significance can be an indication of the dominance over the other by one team or individual in the sporting event.

It is preferred that the indication of dominance includes a progressive time based graphical indication above or below a median line, the position above or below the line being indicative of the dominance over the other by one team or individual in the sporting event.

WO 2004/003815 PCT/AU20031000831 3 Another output of predicted strategic significance can include assessment reports of past and future performance of teams and/or individuals in the sporting event.

Another output of predicted strategic significance can include adaptive control indicators. It is preferred that the adaptive control indicators include suggestions for remedial action and/or advice regarding opportunities and/or strategies for offensive and/or defensive moves.

In another aspect this invention resides broadly in a system for coaching a sporting team during a sporting event, the system including:data generating means for generating quantitative data relating to performance in previous sporting events; data generating means for generating quantitative data relating to performance during the sporting event; data generating means for generating qualitative data relating to the sporting event, and a computer including program means for analysing the generated quantitative and qualitative data by a risk management methodology to thereby generate an output of predicted strategic significance in the future conduct of the sporting event.

Description of Drawings In order that this invention may be more easily understood and put into practical effect, reference will now be made to the accompanying drawings which illustrate a preferred embodiment of the invention, wherein:- FIG 1 is a schematic diagram illustrating the method of the present invention; FIG 2 illustrates the graphical output of a so-called Controlometer, and FIG 3 is a schematic block diagram illustrating the system of the present invention.

Description of Preferred Embodiment of Invention As can be seen in FIG 1 which illustrates a method of coaching a sporting team or individual participants during a sporting event, a series of data WO 2004/003815 PCT/AU20031000831 4 inputs 10 is generated. Quantitative data 11 is generated which relates to performance in previous sporting events; quantitative data 12 is generated which relates to performance during the sporting event, and qualitative data 13,14 is generated which relates to the sporting event and to the participants therein. The generated quantitative and qualitative data 11, 12 and 13,14 is analysed in a computer controlled system in accordance with a program which includes a risk management methodology 21 to thereby generate outputs of predicted strategic significance in the future conduct of the sporting event.

The qualitative data may be data 13 relating to past sporting events or to the participants therein, or it may be data 14 relating to the current sporting events or to the participants therein.

The quantitative data 11,12 is subjected to statistical analysis 23 and both it and the qualitative data are subjected to regression analysis 22.

Looked at in another way, both qualitative and quantitative pre-match data can be regarded as historical and relates to past performance. As will be seen with more specific reference to the preferred embodiment described subsequently in more detail, this historical or pre-match data can be regarded as having a relationship value which is established from the history of games between any two clubs who are to compete in a sporting event.

Similarly what can be termed real-time data, both qualitative and quantitative, is current and relates to the immediate lead up to the sporting event and to the actual performance during the sporting event. Again, as will be seen with more specific reference to the preferred embodiment described subsequently in more detail, this real-time data can be regarded as having a "lead-in to the game" value which is established during the period between consecutive games which a team plays during the competition. This "lead-in to the game" value is based on team comparisons and form, taking into account mental, physical and environment factors.

A number of outputs 30 of predicted strategic significance in the future conduct of the sporting event are provided. These include the indication 31 of a break point in the sporting event. A break point is a time or a stage during a sporting event beyond which, if the conduct of play were to continue in the existing pattern, it is highly likely that one team or individual will win the event or WO 2004/003815 PCT/AU20031000831 a discrete component of the event, for example a game or a set in a tennis match, a quarter or a half in a football match, an innings in a baseball match etc.

Another output of predicted strategic significance is an indication 32 of the dominance over the other of one team or individual in the sporting event.

This is seen in more detail in FIG 2 wherein the indication is seen as a progressive time based line which is traced above or below a median line, the position above or below the line being indicative of the progressive relative dominance of one team or participant over the other.

Other outputs of predicted strategic significance include assessment reports 33 and 34 of past and future performance respectively of teams and/or individuals in the sporting event.

Another output of predicted strategic significance is an adaptive control indicator 35 which can include for example, suggestions for remedial action and/or advice regarding opportunities and/or strategies for offensive and/or defensive moves.

Other outputs of predicted strategic significance can be categorised as key performance measures and telling factors.

The intangible and qualitative components and the reports and recommendations based thereon are enhanced and hence are more useful with real-time match statistics being incorporated into operation of the system.

In the following description of general operation of the system, reference will be made by way of illustrative example, to a sporting event in the nature of the game of Australian football.

The intangible pre-match parameters or indicators include psychological lead-up to the game, physical and environmental factors, shock news and events, the M-factor (motivational issues), switch-ons, what is at stake and who has come to play. Together with pre-match statistics these pre-match events overlay and take precedence over past match statistics.

The intangible and qualitative components are converted to mathematical formulae and incorporated into an IT software system. The software is used in line with traditional match statistical information in two forms historical statistical information stored on hard disk before a game starts and secondly, progressive match statistical data fed live into the WO 2004/003815 PCT/AU20031000831 6 system. Input may be by conventional keyboard means for example, by voice activated control systems, or by GPS or other automated positional real-time analysis of the game play. Voice activated input is particularly suitable for inputting real-time intangible parameters.

Reports are triggered when a match reaches a defined break-point which can be at anytime in the match. Assessment reports are also available for a quarter, half or full game and as above are based on a combination of qualitative analysis, quantitative data and a so-called game "controlometer" to be described subsequently.

Such reports are effective to warn a coach that a break-point has been reached and unless a change occurs, the current pattern or trend of play will cause the team to lose the quarter, half, match, etc. Conversely, they warn that the opposition team is going to win the quarter etc unless a change occurs. Reports on break-point for the opposition can also be generated thereby allowing a ready reconciliation with the 'controlometer.' The controlometer reads the game and records who is controlling the agenda of the game. As seen in FIG 2, a pointer or trace moves above and below a line for either team and produces signals when one team reaches certain points eg sustained scoring pressure in the danger zone and vulnerability through certain events such as turnovers, missed scoring opportunities etc.

The controlometer 'reads' the game and has the potential to measure performance responses to changes in tactics, game plan moves, player rotations, etc. The output from this display constitutes a mathematical input to the system as an 'agenda' factor and secondly a visual aid to coaches as the controlometer is used and shows an agenda change from one team to the other.

It will be clear that the system and method of the present invention, by providing the improved analysis which enables these break points to be predicted in advance, complements the intuition of even the best coaches and allows the coaches to send early instructions to the players to enhance the team's fortunes at the predicted break points.

WO 2004/003815 PCT/AU20031000831 7 As well as historical data which provides useful planning information for the coach and a system guide for interpreting patterns in a current game, the best available current match statistics are input in real time into the software system during the game. Such input includes known statistics such as number of kicks, attacks into 50m zone etc. In addition it includes information such as Ratio of scoring opportunities to scores; Individual player performance number of players winning their positions weighted for player ratings; Location and duration of play at points on the field weighted for value; Patterns of play direct and indirect; 4 Changes weighted; Source of scoring turnover, ball-up, throw-in, free-kick etc; 4 Weighting for overall first quarter differential between teams; 4 Performance statistics who and how the ball is won, possession time weighted for distance forward, sequential possessions weighted for distance forward, scoring success/failure ratio, number of possessions per goal (higher the number higher the risk of failure), number of players to goal, flooding indicator, launch of scoring defence or midfield, 50m entry/scoring ratio; and o Quarter score differentials eg third quarter score equals opponents second quarter score.

This statistical quantitative real time match performance data is combined with the intangible qualitative components.

During a game certain events occur which have a psychological impact on the teams and individual players and which can affect the outcome of a quarter or the whole game. These events can even contradict the flow of a game as recorded statistically and therefore their analysis adds a definite discrete value-added factor to statistics.

Such factors include the following which are by no means exhaustive:- 4 Domination of play 4 Scoring against the flow of play Team or ratio of players running hot WO 2004/003815 PCT/AU20031000831 8 Demoralising incident or scoring Tactical change player or game plan Injuries to key players or resulting in team imbalance Pattern of turnovers under pressure resulting in scores o* Reporting of players Players or teams losing control eg 50m penalties and being disciplined Pattern of missing goals for behinds or out of bounds Benching of stars Key players with poor performances *o Handballing consistency over quarters Pattern of top 6 middle 6 bottom 6 performances A primary output from the system (apart from observation of the controlometer and mental intuition from inputting information) is the indication of so-called "break points" for each quarter or for the entire match. A warning is generated at a point when if things remain constant the team is at risk of losing the quarter/game. This indication is accumulative and is determined from combined analysis of tangible and intangible data loaded before and during the game.

These break point warnings add to the coaches' professional intuition in a manner which can be likened to having a "super coach" looking over their shoulder.

One performance measure is the break-point showing for the other coach (which that coach will not see) indicating that the game plan of the coach utilising the present invention is working.

When a warning is generated the system indicates the key aspects of the tangible and intangible indicators which triggered the warning. This will help the coach translate his 'inkling' which may come to him quite intuitively and without rationale into an analysis of the area(s) where the team or game plan is breaking down. The system thus produces advice to the coach where attention or change is needed, which advice or warning the coach is obviously free to accept or reject.

WO 2004/003815 PCT/AU20031000831 9 Such warnings provide anticipatory advice to coaches about areas of performance, or players, or match plays which opposing coaches may be exploiting (or beginning to exploit) or where they are gaining (or beginning to gain) an advantage.

The incorporation of risk management scenario planning techniques into the software facilitates the real time inputting of opposition moves and player changes and facilitates the prediction of impacts.

Looking now in more detail at the present invention in the context of the preferred embodiment of a particular game of Australian football between two teams, the key areas of the system and their interrelationships are as follows:- I Relationship Value. This refers to inputs established from the history of games between any two clubs Quantitative Component Statistical Probability of winning based on analysis of historical match data Qualitative Component Domination Factor measured from head to head history of the two clubs results over last 10 years and last 3 years (weighted) (ii) Special Match Factor measured from history of a particular game or particular game between two clubs eg event/day, location, comparative team structure or game style A generic algorithm representing this aspect of the method and system of the invention is RV f [SP:DF:SMF] WO 2004/003815 PCT/AU20031000831 where RV is Relationship Value SP is Statistical Probability DF is Domination Factor SPM is Special Match Factor In other words, Relationship Value is a function of Statistical Probability, Domination Factor and Special Match Factor.

2 Lead in to Game Value. This refers to inputs established during the week between games based on team comparisons and form, taking into account mental, physical and environment factors Quantitative Component Team Value Factor a weighted combination of: Closeness to preferred or best team available Closeness to team that played over the last two weeks Best six available Second best six available Value of interchange bench Discount for players carrying injuries/return from long absence Predicted win/loss on all match-ups Qualitative Component Mindset Value Factor a weighted combination of: Take-out factor from last game eg positive, increased pressure WO 2004/003815 PCT/AU2003/000831 11 -Take-in factor to game expected relationship at first bounce -Vulnerability factor risk of anti-climax or emotional relief/crash Switch factor closure, when and how effective Team focus factor looking forward or just next game Significant individual focus factor eg goal, motivation Media factor eg last game, switch, special match value, distract Team pressure value eg position, 8 points, top eight berth Individual game pressure retain spot, critical opponent Confidence factor atmosphere, intangibles Coping factor player resilience threshhold Intensity factor absence/presence positive or negative Player feeling factor average and variation from normal Game plan belief factor understanding and achievability Risk assessment of game plan likelihood, consequences Motivation factor extent to which team requires motivation Internal shock, crisis, distraction factor during week Risk of carrying shock, crisis, distraction into match Form rating factor assessment of form to be carried into game Impact of normal pre-match mental 'warm-up' (ii) Physical Value Factor a weighted combination of: WO 2004/003815 PCT/AU2003/000831 12 Critical player presence/absence timing impact relevant Players carrying injuries Match fitness of key players Leadership presence Weather factor rain, wind Venue factor physical and mental Ground conditions style, team suit/adjusted Value of physical warm-up expected (iii) Environment Factor a weighted combination of: Home ground advantage Climatic advantage Time of play advantage Home crowd advantage Travel effect normal coping factor/extent of recent travel Relationships factor critical individual risk carry-in factor Personal crisis factor critical individual risk carry-in factor Public issue or crisis factor team/individual carry-in factor A generic algorithm representing this aspect of the method and system of the invention is LGV f [TVF:MVF:PVF:EF] where LGV is Lead-in Game Value TVF is Team Value Factor MVF is Mindset Value Factor PVF is Physical Value Factor WO 2004/003815 PCT/AU20031000831 13 EF is Environment Factor In other words, Lead-in Game Value is a function of Team Value Factor, Mindset Value Factor, Physical Value Factor and Environment Factor.

3) Game Agenda 'Controlometer'. This output is established by plotting the flow of the agenda of the game and producing warnings and triggers at key points.

Quantitative Component Agenda Control Factor is measured from the following: recording possession, entering 50m and scoring minor possession aberrations are ignored -maximum point reached when team goals key alerts are generated in the following circumstances: when possession leads to scoring opportunity (attack) when possession inside 50m does not result in a goal or score after x seconds (defence) when turnover leads to successful disposal (change) when turnover leads to goal/score (vulnerability) Qualitative Component Validation Factor. This enables coach to reconcile game plan against performance and measure variations by examining for example: WO 2004/003815 PCT/AU2003/000831 14 change of game control comparative possession agenda control and conversion rate domination of midfield, attack and defence cost of turnovers effect on agenda control A generic algorithm representing this aspect of the method and system of the invention is GA(C) f [ACF:VF] where GA(C) is Game Agenda "Controlometer" ACF is Agenda Control Factor VF is Validation Factor In other words, Game Agenda "Controlometer" is a function of Agenda Control Factor and Validation Factor.

4) Key Performance Measures. These are measured by team performance outputs derived from match activities.

Quantitative Components Team Performance Factor which is a function of [Mindset/ Intensity/Fitness and Skill/Game Plan/Winning Possession/Into 50m Zone/Scoring].

It is measured from analyzing match statistics into the following performance criteria: Midfield dominance quotient Effective disposal quotient Winning top six player quotient WO 2004/003815 PCT/AU2003/000831 Winning second top six player quotient Winning third top player quotient Clearances from stoppages Continuing bench value quotient Overall winning position quotient (all 18 positions) Goals from entries into 50m zone Player domination by team top 12 players by effectiveness Patterns of play producing goals First quarter barometer reading on comparative effectiveness Source of goals attack, midfielders, defenders, bench Winning contested possession Comparative time in possession Comparative number of disposals to goal Average time taken to advance Proportion of players without a possession/disposal in last 10 minutes Time in 50m zone/goals scored Qualitative Components Game plan effectiveness. This is measured by assessment of achieving its overall objectives eg shutting down opposition scoring capacity and on particular aspects including: Start of game relationship validation Team switched on to play Anticipation of opposition game plan Effectiveness of changes to game plan Maintenance of team intensity Pressure on opposition WO 2004/003815 PCT/AU2003/000831 16 Adjustment effectiveness stoppages, turnovers, switches, etc Pattern predictability applicable Crowd influence as expected Umpiring factor relevant Spine and centre control Mode reaction time attack, defence or dispute Goalkickers fit pattern of game plan Matchup effectiveness Progressive Mindset, Physical and Environment Factor influence Scoring ease (comparative) ease of entry to scoring zone number of entries to scoring zone goals from marks goals from play disposals per goal A generic algorithm representing this aspect of the method and system of the invention is KPM f [TPF:GPE] where KPM is Key Performance Measure TPF is Team Performance Factor GPE is Game Plan Effectiveness In other words, Key Performance Measure is a function of Team Performance Factor and Game Plan Effectiveness.

Telling Factors. These are significant or critical factors which can influence the outcome.

WO 2004/003815 PCT/AU2003/000831 17 Quantitative Components Projection of progressive Team Performance Factor results (ii) Scoring pattern against the flow of possession (iii) Maintenance of player domination for a defined time period (iv) Repeated turnovers resulting in goals Comparative scoring ratios into (vi) Effective disposal pattern projection (vii) Comparative weighted performance of top six players Qualitative Components Agenda Controlometer predictions (ii) Loss of discipline behavioural change (iii) Demoralising incidents (iv) Loss of intensity Leadership inspiration (vi) Heroic incident (vii) Key player injury loss (viii) Key players benched for poor form (ix) Reports Game plan changes forced by opposition (xi) Match-up changes forced by opposition (xii) Significant take-out mindset value from a quarter 6) Breakpoints. These are points at which if game continues as is, the system predicts a quarter/game win/loss Quantitative Component WO 2004/003815 PCT/AU2003/000831 18 Controlometer prediction for quarter/game (ii) Aggregate Team Performance factor exceeds say (iii) Controlometer warnings from progressive triggers, viz: comparative scoring zone entry and conversion comparative scoring zone entry and not scoring turnovers resulting in scores midfield dominance domination factor exceeds comparative possession scoring pattern against the flow significant increase in uncontested possessions Qualitative Component Proportion of agenda control reaches Refusal to chase opposition Second and third effort quotient 1% quotient Degree of difficulty in scoring Observable kinesics and proxemics 7 Change Indicators The system will thus be seen to produce the following outputs and outcomes: Controlometer patterns, triggers and warnings: validation point for coaching; Aggregate Team Performance Factor determined progressively through the game WO 2004/003815 PCT/AU20031000831 19 o Performance outputs derived from realtime analysis of match statistics Breakpoint areas identified from controlometer and performance analysis Capacity to program for information on demand analysis of play/player Capacity to scenario plan with attached risk assessment Capacity to anticipate possible opposition plans/moves Turning now to FIG 3 which is a block diagram schematically illustrating the system of the present invention for coaching a sporting team during a sporting event, the system is seen to include data generating means 40b, 40c......40n for generating data relating to performance in previous sporting events, and data generating means 50a, 50b, 50c......50n for generating data relating to performance during the sporting event. A computer 60 including program means 70 analyses the generated quantitative and qualitative data by a risk management methodology to thereby generate outputs 80b, 80c......80n of predicted strategic significance in the future conduct of the sporting event.

The data generated by data generating means 40a, 40b, 40c......40n relating to performance in previous sporting and by data generating means 50n relating to performance during the sporting event may be either quantitative data as defined above or qualitative data as defined above.

Accordingly data generating means 40a, 40b, 40c......40n can variously comprise access to statistical databases of historical data, and input devices such as keyboards, palm pilots and the like for accessing further information or for inputting further information into computer Similarly, data generating means 50a, 50b, 50c......50n can variously comprise manual input devices such as keyboards, palm pilots, WO 2004/003815 PCT/AU20031000831 voice/responsive actuators and the like for inputting information regarding progress of the game into computer 60, or remote/automatic input devices such as GPS based devices for inputting positional information regarding the game into computer Program 70 controls computer 60 in the manner described above to generate a series of outputs 80a, 80b, 80c......80n of predicted strategic significance in the future conduct of the sporting event. These outputs variously comprise the outputs described above and can variously be provided by suitable peripheral equipment such as palm pilots, computer screens, print outs, audible alarms or warnings such as artificial voice systems.

It will be appreciated that the present invention has a number of advantages when compared with existing methods and systems for coaching a sporting team or individuals during a sporting event.

Whereas existing systems are based on tangible events and incidents, and measured in quantitative terms based solely on statistics (eg match details), analysis and probability calculations, the present invention will be seen to include intangible aspects of games preparatory to and during a match measured through a combination of qualitative and quantitative assessments and is based on risk management methodologies.

The present invention also is seen in a preferred embodiment to incorporate, inter alia, qualitative regression analysis techniques projected forwards and identifies 'break points' in anticipation. It can be said to replicate the mind, the intellect of a "super-coach"', and derives an absolute match analyses at a level arguably above the mental capacity of the coach, and does this at the time of occurrence rather than post-match.

The outcome reports produced during a game by the present invention attempt to replicate what the hypothetical super-coach could in theory identify, analyse and predict.

Whilst theoretically, and most probably also in practice, a top coach will have an inkling of 'unrest' at break points during a game, it remains the fact that even the very best coaches lose games and can look back and realize that if they had adopted a particular strategy or game play at a particular WO 2004/003815 PCT/AU20031000831 21 point, they might have saved the match. This present system provides coaches with such a regression analysis in realtime.

It will be appreciated that the system can also be used to predict competitions as well as individual games and thus is suitable for use in betting on sporting events or competitions.

It will of course be realised that whilst the above has been given by way of an illustrative example of this invention, all such and other modifications and variations hereto, as would be apparent to persons skilled in the art, are deemed to fall within the broad scope and ambit of this invention as is herein set forth.

Claims (9)

1. A method of coaching a sporting team or individual participants during a sporting event, the method including:- generating quantitative data relating to performance in previous sporting events; generating quantitative data relating to performance during the sporting event; generating qualitative data relating to the sporting event, and analysing the generated quantitative and qualitative data in a computer controlled system in accordance with a program which includes a risk management methodology to thereby generate an output of predicted strategic significance in the future conduct of the sporting event.

2. A method of coaching as claimed in claim 1, wherein the analysis of the qualitative data includes regression analysis techniques.

3. A method of coaching as claimed in claim 1, wherein an output of predicted strategic significance includes prediction of a break point in the sporting event.

4. A method of coaching as claimed in claim 3, wherein an output of predicted strategic significance includes an indication of the dominance over the other by one team or individual in the sporting event.

A method of coaching as claimed in claim 4, wherein the indication of dominance includes a progressive time based graphical indication above or below a median line, the position above or below the line being indicative of the dominance over the other by one team or individual in the sporting event.

6. A method of coaching as claimed in claim 1, wherein an output of predicted strategic significance includes assessment reports of past and future performance of teams and/or individuals in the sporting event. WO 2004/003815 PCT/AU2003/000831 23

7. A method of coaching as claimed in claim 1, wherein an output of predicted strategic significance includes adaptive control indicators.

8. A method of coaching as claimed in claim 7, wherein the adaptive control indicators include suggestions for remedial action and/or advice regarding opportunities and/or strategies for offensive and/or defensive moves.

9. A system for coaching a sporting team or individual participants during a sporting event, the system including:- data generating means for generating quantitative data relating to performance in previous sporting events; data generating means for generating quantitative data relating to performance during the sporting event; data generating means for generating qualitative data relating to the sporting event, and a computer including program means for analysing the generated quantitative and qualitative data by a risk management methodology to thereby generate an output of predicted strategic significance in the future conduct of the sporting event.