Alright, so today I’m gonna walk you through this little project I was messing around with: “jokic mavericks”. Sounds kinda cryptic, right? Well, let me break it down.
It all started when I was trying to figure out how to predict NBA player stats. Like, could I build a model that could somewhat accurately guess how many points Jokic would score in a game, or how many assists Luka would dish out? Classic data nerd stuff, you know?
First thing I did was grab a bunch of data. I’m talking years worth of NBA game stats, player info, team records, everything I could get my hands on. I scraped some data off the web, downloaded some publicly available datasets, the whole shebang. It was messy, let me tell you. Data cleaning is like 80% of the job, am I right?
Once I had the data in a somewhat usable format, I started playing around with different machine learning models. I tried linear regression, decision trees, even threw in a random forest for good measure. I’m no expert, I just like to tinker around and see what happens.
The “jokic” part came in because I was initially focusing on predicting Nikola Jokic’s stats. He’s such a unique player, and I thought his impact on the game would be interesting to model. Then I expanded it to include Luka Doncic, hence the “mavericks” part.
I spent a solid week just tweaking features, trying different combinations of variables to see what would give me the best results. Like, does the opponent’s defensive rating affect Jokic’s scoring? Does the number of games played in a week impact Luka’s assists? You get the idea.
Honestly, the results were… mixed. Some predictions were surprisingly accurate, while others were way off. I think the biggest challenge was accounting for unpredictable factors like injuries, player matchups, and just plain old luck. Sports are inherently random, which makes them tough to model.
I ended up building a simple web app using Flask to display the predictions. It was nothing fancy, just a basic interface where you could select a player and a game, and it would spit out the predicted stats. More of a proof-of-concept than anything else.
Here’s a quick rundown of the steps I took:
- Data Acquisition: Scraped NBA stats from various sources.
- Data Cleaning: Wrangled the data into a usable format. Lots of pandas involved.
- Model Selection: Experimented with different machine learning models.
- Feature Engineering: Created new features from existing data.
- Model Training: Trained the models on historical data.
- Prediction: Used the trained models to predict player stats.
- Web App: Built a simple Flask app to display the predictions.
So, yeah, that’s “jokic mavericks” in a nutshell. It was a fun little project that taught me a lot about data science, machine learning, and the challenges of predicting the unpredictable. I might revisit it someday and try to improve the model, but for now, it’s just another project in the “done but not perfect” category.
Would I use this to bet on games? Absolutely not. But it was a cool exercise in data analysis and prediction.
