New AI software can predict rugby injuries

Gudnews said:
What's BS about this?
Click to expand...

No doubt they can work out averages better, and pinpoint dangers / needs for improved focus on some issues. But predict the future?? Come on, man...

The article itself mentions it's results of a once-off and uses careful wording. Also, results are not specific to who or when as I understand it.
Put it down to clickbait headings if you prefer.
 
Brian_G said:
No doubt they can work out averages better, and pinpoint dangers / needs for improved focus on some issues. But predict the future?? Come on, man...

The article itself mentions it's results of a once-off and uses careful wording. Also, results are not specific to who or when as I understand it.
Put it down to clickbait headings if you prefer.
Click to expand...
This article was thrown in here because it contains the word AI.
 
Paper is here:

https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0307287
The classification accuracy (i.e., accuracy with which injured/non-injured group membership is determined) of the models that emerged from the feature selection stage was assessed via naïve bayes [61], J48 decision tree [62], support vector machine [63], and K-nearest neighbour [64] classification algorithms
Click to expand...
Nothing wrong with the methods listed that are actually doing the classification, but they are not new. Nor are they even close to AI. They are basic statistical models.

Heck, even the classifier that did the best in their paper was the Naive Bayes, which is basically just projecting a player's data onto multiple gaussian distributions that are derived from the training set (calculating the mean, median for each parameter) and adding them and seeing what has the highest likelihood. Even when I did ML at uni, we had more sophisticated models ffs.

AI in the current context, given what we have right now in 2025, the year of our Lord and Saviour Donald Trump, would be being able to put all the factors that you know about a player into an LLM, which has a generalized understanding of the world, and it being able to give you a 1-10 risk score.
 
Last edited:
konfab said:
Paper is here:

https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0307287

Nothing wrong with the methods listed that are actually doing the classification, but they are not new. Nor are they even close to AI. They are basic statistical models.

Heck, even the classifier that did the best in their paper was the Naive Bayes, which is basically just projecting a player's data onto multiple gaussian distributions that are derived from the training set (calculating the mean, median for each parameter) and adding them and seeing what has the highest likelihood. Even when I did ML at uni, we had more sophisticated models ffs.

AI in the current context, given what we have right now in 2025, the year of our Lord and Saviour Donald Trump, would be being able to put all the factors that you know about a player into an LLM, which has a generalized understanding of the world, and it being able to give you a 1-10 risk score.
Click to expand...
The world currently thinks that a sensor that triggers an action at a certain time is AI :ROFL:
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter