Tag Archives: artificial intelligence

Predicting Soft-Tissue Injuries In Rugby / Rugby League – Free Software

Clubs have always wanted a way to maximise a teams preparation, while at the same time reducing the risk of Soft-Tissue injuries during the coming week.

Rugby players can monitor the risk of soft-tissue injury that they may be exposing themselves to during the coming week, either as a result of the squads training, or as a result of their own individualised training program.

Although this software is based on our very effective artificial intelligence software, it has been greatly simplified compared to our bespoke systems, and makes no assumptions about the availability of historical data, technical or support staff.  This is an entry level solution, for those clubs or individual athletes wishing to explore this technology, but who would not have the willingness or resources to consider a more customised approach.  And it is absolutely FREE of charge.

The software (called the Predictor), has been designed to analyse the predicted risk of soft-tissue injury in the coming week, for one or more athletes.

The system, which utilises an Excel spreadsheet for input and output purposes, requires a minimal amount of information such as:

  • Your planned training session durations and associated RPEs (Relative Perceived Exertion – 0 to 10)
  • Last weeks Load result, which was calculated by the spreadsheet

  • Several Muscle Soreness and Well-being ratings at the start of the week

Once this weeks planned session times and anticipated RPE’s, last weeks total Load, and current Muscle Soreness and Well-being ratings have been entered, the Injury Predictor based on an artificial neural network, will produce a predicted risk of injury, for each individual during the coming week.

This Entry Level Injury Predictor for Rugby / Rugby League requires a Microsoft Windows environment, together with Microsoft Excel.  In the event that you are an Apple Mac owner (as I am), then you will need to run your Microsoft Windows & Excel environment on your Mac using Apple’s Bootcamp utility, or a third party  virtual machine environment such as Parallels.

Subscribe to our mailing list and receive your FREE software.

Once you have subscribed –

  • A Confirmation Email will be sent to the email address supplied below.  If you don’t see shortly, then you may want to check your Junk folder.
  • When you have confirmed your subscription using the link provided in the email, you’ll be directed to the Software Download web page.
* indicates required




Injury Prediction & Management In A “Box” For Amateur Rugby Players

Amateur Rugby clubs have always wanted a way to maximise a teams preparation, while at the same time reducing the risk of Soft-Tissue injuries during the coming week.

What may be a world first, individual Rugby players of all ages and experience, who as enthusiastic amateurs, can monitor the risk of soft-tissue injury that they may be exposing themselves to during the coming week, either as a result of the squads training, or as a result of their own individualised training program.

Although this product is based on our very effective artificial intelligence software, it has been greatly simplified compared to our bespoke systems, and makes no assumptions about the availability of historical data, technical or support staff.  This is a low cost of entry solution, for those amateur clubs or enthusiastic individual athletes wishing to explore this technology, but who would not have the willingness or resources to consider a more customised approach.

The software (called the Predictor), has been designed to analyse the predicted risk of soft-tissue injury one week in advance, for one or more athletes.

The system, which utilises an Excel spreadsheet for input and output purposes, requires a minimal amount of information such as:

  • Your planned training session durations and associated RPEs (Relative Perceived Exertion – 0 to 10)
  • Last weeks Load result, which was calculated by the spreadsheet

  • Several Muscle Soreness and Well-being ratings at the start of the week

Once this weeks planned session times and anticipated RPE’s, last weeks total Load, and current Muscle Soreness and Well-being ratings have been entered, the Injury Predictor neural network, will produce a predicted risk of injury, for each individual during the coming week.

Being aware of the risk is an important piece of information, but the question still remains as to what can be done to reduce that risk.  That is a job for the Optimiser.

The Amateur Rugby Injury Optimiser (or Optimiser) product, uses a unique search algorithm to find an “optimal” training scenario so as to minimise the risk of injury to an athlete.  This “optimal” scenario is subject to constraints stipulated by the user.

The Optimiser compliments the Amateur Rugby Injury Predictor (or Predictor) product, by working in partnership with it to evaluate different possible training scenarios.

Once the Predictor highlights an unacceptable risk of injury, it is possible to explore how that risk might be reduced to an acceptable level, without the use of the Optimiser by manually re-running the Predictor repeatedly, with the user changing inputs manually and re-examining the results, it may be possible to eventually reach an acceptable level of injury risk.  The only disadvantage to this approach is the time it might take to do this (with hundreds or more possible scenarios), and the uncertainty of not knowing whether the risk achieved was at its lowest level possible.

If the Predictor indicates that a player is at risk, those inputs relating to that player can be copied from the Predictor into the Optimiser.

The user simply enters the upper and lower ranges for those factors which could be modified – in this case the expected session loads.  The Optimiser then, in conjunction with the Predictor neural network, intelligently assesses hundreds if not thousands of possible scenarios, till the lowest possible risk of injury is determined, within the stated upper and lower load ranges.

The Optimiser displays the “adjusted loads” required to achieve this lower level of injury risk.

Both the Amateur Rugby Injury Predictor and Optimiser require Microsoft Excel and Microsoft Windows, and in the case of the Optimiser, also requires the prior installation of the Amateur Rugby Injury Predictor.

 

 


Injury Prediction & Management In A “Box” For Professional Rugby

As a rule our predictive models are custom built to match the specific requirements of a club and it’s staff.  Unfortunately not every club has the resources – financial, staff or data – that will enable such an endeavor to be undertaken.

To help clubs take advantage of this technology, we have created a “boxed” version of our technology, allowing Coaching, Medical, and Strength & Conditioning staff at Rugby clubs who have always wanted a way to maximise a teams preparation, while at the same time reducing the risk of Soft-Tissue injuries during the coming week.

This product is also applicable to individual Rugby players, who as professional athletes, may wish to monitor the risk of soft-tissue injury they may be exposing themselves to, either as part of the squads training, or as a result of their own individualised training program.

Although this product is based on our very effective artificial intelligence software, it has been greatly simplified compared to our bespoke systems, and makes no assumptions about the availability of historical data, technical or support staff.  This is a low cost of entry solution, for those clubs or individuals wishing to explore this technology, but who may not have the willingness or resources to consider a more customised approach.

In fact there are two parts (or perhaps I should say packages) to this story – a Predictor and an Optimiser.

The predictor software can analyse the risk of soft-tissue injury up to a week in advance, for one or more athletes.

The system, which utilises an Excel spreadsheet as a front-end for input and output purposes, requires a minimal amount of information such as your planned training session durations and associated RPEs (Relative Perceived Exertion – 0 to 10),last weeks Load, Monotony and Strain results, which were calculated by the spreadsheet,

and finally several Muscle Soreness and Well-being ratings taken at the start of the week

Once this weeks planned session times and anticipated RPE’s, last weeks total Load / Monotony / Strain, and current Muscle Soreness and Well-being ratings have been entered, the Injury Predictor neural network will, based on it’s training, produce a risk assessment for each individual for the coming week.

The second part of this story is the Optimiser.

The Rugby Injury Optimiser (or Optimiser) product, uses a unique search algorithm to find an “optimal” training scenario so as to minimise the risk of injury to an athlete.  This “optimal” scenario is subject to constraints which can be stipulated by the user.

The Optimiser compliments the Rugby Injury Predictor (or Predictor) product, by working in partnership with it to evaluate different possible training scenarios.

Once the Predictor highlights an unacceptable risk of injury, it is possible to explore how that risk might be reduced to an acceptable level, by manually re-running the Predictor repeatedly, with the user changing inputs manually and re-examining the results.  It may be possible to eventually reach an acceptable level of injury risk in this manner.  The only disadvantage to this approach is the time it might take to do this (with hundreds or more possible scenarios), and the uncertainty of not knowing whether the risk achieved was at its lowest level.  This is why we developed the Optimiser.

If the Predictor indicates that a player is at risk, those inputs relating to that player can be copied from the Injury Predictor into the Optimiser.

The user simply enters the upper and lower ranges for those factors which could be modified – in this case the expected session loads.  The Optimiser then, in conjunction with the Predictor neural network, intelligently assesses hundreds if not thousands of possible scenarios, till the lowest possible risk of injury is determined, within the stated upper and lower load ranges.

The Optimiser displays the “adjusted loads” required to achieve this outcome.

Both the Rugby Injury Predictor and Optimiser require Microsoft Excel and Microsoft Windows, and in the case of the Optimiser, also requires the prior installation of the Rugby Injury Predictor.

 

AFL Match Strategies – How To Win & Lose

AFL Match Strategies, are a series of predictive computer models (one per team playing in the AFL Premiership) which are redeveloped each season.  These models are developed using match statistics captured over several seasons for each premiere league AFL game played.

The software is directly applicable to every Australian Football League (AFL) Premiership competition club.

These models were designed specifically to identify relationships between on-field metrics and match results.  Once developed, their accuracy was tested by “replaying” games each club played over the last three seasons.

Clearly each team is different and unique, and warrants individual models to capture their unique qualities.  With this in mind, we have analysed the “differences” between each teams recorded metrics and the match outcome, to discover a very clear relationship (refer to the table on the left for the respective model’s predictive accuracy) between “differences” and Win / Loss for each club.

These models or Team Profiles, based on our artificial intelligence software, are able to examine the impact each of these metrics have on a teams performance.

 

 

Using the Adelaide profile as an example:
A    Summary graph, showing the relative “sensitivities” of each of the game metrics for Adelaide, not only to each other, but their impact on match outcomes.  Sensitivity is a measure of the change in the probability of Adelaide winning or losing, as the metric is varied.
B    Indication of the accuracy of the Adelaide team’s underlying model in correctly identifying Winning and Losing match outcomes, based on a given matches metrics.
C    Home & Away game inputs are the only ones that are not “differences”.  Think of the horizontal axis as being a “transition” from not a Home game to being a Home game, or not an Away game to being an Away game.  The benefit of Home Games to Adelaide is very apparent from this example.
D    All the subsequent graphs represent the remainder of the 19 inputs under examination.

Once you have Team models that accurately represents your team, or your opponent team’s behaviour, and its likelihood of Winning or Losing, then you are in a position to examine “what-if” scenarios to explore different types of Game Strategies that could influence that Win / Loss.

Possible scenarios could include:

  • What match metrics within my team could I manipulate, and by how much, so as to increase the likelihood of a Win against a particular opponent?
  • What match metrics within the opposition team, can I reasonably expect to influence, so as to reduce their likelihood of defeating my team?
  • Fully understand your own team’s sensitivities to changing circumstances against a given opponent, and the likelihood of moving from a Winning to a Losing situation (and vice versa), depending on circumstances.

The number of scenarios you could explore are only constrained by your imagination.  Team metrics can be given as much or as little latitude as you wish so as to examine their impact on the matches’ outcome.

It has been suggested that it may be easier to disrupt or block another team’s performance, rather than change your own team’s behaviour.  With these Team models it is possible to take either or both approaches.

Each team’s predictive model requires a Microsoft Windows environment (native or virtual machine) and Microsoft Excel.

 

Cricket Australia Examines Match Strategies

We have been given the go-ahead to apply our artificial intelligence software to the analysis of matches, so as to understand and identify “patterns” in our strategies that are associated with winning and losing.

Much of the work we undertake with our sporting clients relates to the prediction of injury to an athlete, and the analysis of strategies to reduce or minimise that risk.

In this case, Cricket Australia’s – Centre of Excellence engaged us to look at the relationships between match outcomes, and the various strategies employed against various opponents during a match.