We are supposed to extend the existing tennis model (any 1 of the 3) to predict the outcome of a specific match-up
We can do this by figuring out the probabilities of moves made by 2 specific players. For example, in the tennis data set, we find Tennis Player John and determine the probability that he does a BackHand_Crosscourt deuce stroke, then manually change the probability found in the pcase for deuce stroke.
We can also extend the project by keeping track of the last 2 moves made by a player, and use it to improve the accuracy of the probabilities
For mid-term presentation, we just need to
- do some improvements to the model
- do a powerpoint presentation to explain our improvements and findings
- we can just choose any sport as long as can have some data from somewhere to derive probability values
- probabilities just stay static**, so probabilities are derived from data (take from website that records every move for u or just watch some tennis match lol // since we also need to present how we get our data)
for the tennis model, we can also go about taking into account 2 previous shots in 2 ways:
- take into account their current position on the court before their next hit
- just take note what kind of shot they hit previously (maybe see if got pattern, but this way has alot of permutations đź’€)
- pcase just means probability case
- with prob; just means with probability (we are guessing it just prints the probability of the winning player...?)