🫡

Continuing the analysis of the Raptors season, here is how each player on the Raptors fit into these defensive archetypes. The first image shows their defensive ratings based on my defensive classifications from before. Brandon Ingram for example is a gambler who is one is one of the worst players in the role on the team. Collin Murray-Boyles is one of the best gamblers on the team and top 5 in the league. Graph two shows how well these players actually fit into the archetypes created, our classification of Sandro Mamukelashvili as a Gambler is highly representative of his playstyle on defense compared to players at the bottom. What this means? As you can see the team is made up of primarily Gamblers and Balanced Perimeter Defenders. Whether its scheme or play, the team relies heavily on their wings to be in passing lanes and aggressively attack space while guards are attempting to stay in front of defenders at the first level. The team lacks Paint Protectors and Roamers from their PF and C spot.

Here is a quick analysis of the Raptors based on the Four Factors and defensive classification based on defensive playstyle. Whether they guard the perimeter, paint, or how "active" they are on ball or off ball classifications compared to the rest of the league. The Four Factors are eFG%, Free Throw Rate, Turnovers Per Possession, and Offensive Rebounding Percentage which based on statistical analysis are huge indicators of team success. You can see the difference between the 24-25 Raptors and the 25-26 Raptors based on the Four Factors and then where their current roster places into my defensive classifications. The classifications are done through k-means and I landed on 5. The further on the left on the x-axis means they are never in the paint on defense compared to the opposite of always in the paint and along the y-axis is tipped passes or steals the higher on the axis being more disruptive. The Raptors don't have anyone that fit the Roamer or Paint Protector bucket.

My computer vision and machine learning project came about because of a greater artificial intelligence project meant to be a tool for basketball staff to: - Hidden gem analysis of upcoming free agents, assess teams free agent needs. - Agent negotiation simulation based on agent history and likely asking price and justification. - CBA navigation of simple questions which provides references of section, paragraph etc. and more complex assessments of the CBA such as quirks. - Team Travel schedule generation to maximize off-time, nearest 5 star hotels to game, and nearby events that might cause travel delays. - Data visualization which can output bar graphs, scatter plots, etc. of any statistic based on user question. This includes shot charts, heat maps, zone hot / cold and can be narrowed to time left on clock, quarter. - Pick valuations and team projection for future picks, also understands financial rules and each teams cap for financial related questions. - Trade scenarios that include 5 scenarios for open ended questions and detailed breakdowns of specific trades. "Find a way for the Celtics to get a PG" generates 5 scenarios of a "Fleece", "Downgrade + Upside", "Draft Capital Maximizer", and suggest multi-team deals. For specific deals it will say to accept, decline and offer justifications and counter offers. - What if questions like, "What should the Nets do with a two team expansion?" provide insights into who to leave unprotected, who to offer in trades for picks, or offer expansion teams picks to take certain players of theirs. - Scouting report generation based on ingested scouting reports, its own database, supplement play designs, and synthesize and conduct its own scouting. All of this is created using open source data currently which comes from the NBA API, scraped websites like BBallRef, Spotrac, and proprietary data of mine to train and learn from mainly from custom scouting reports. Very interested to hear what you all think but I plan on posting some of the AI back and forth as it is quite extensive. First up is the CBA rules, which is extracted from the CBA 101 document or the official 2023 CBA for official rulesets with the user question being based on player appearances. Next up in the series is showcasing the more in depth analysis of the CBA so stay tuned.

Here's a cool project I've been working on the last few months in being able to track players and automatically identify plays and catalog them. Check it out!

Hello everyone, Here is a personal project that builds on what @skalskip92 did with his computer vision project, I tweaked it to be able to run full games handling dead balls, free throws, camera cuts, etc. and offload all the data into a CSV where I store: - Shot locations with closest defender distance and name. - Distance Traveled Total throughout the game. - Velocity Information of players - Max and Average Speed - Time of Possession The cleanup is done post-processing where we fill in players information based on known information from our tracking with the NBA API to get key points to ignore free throws polluting the data along with filling in unknown players. Nothing is hardcoded besides team names, all roster and game information is gathered from reading the scoreboard to find the unique game and then it downloads roster information and provides a safety net for tracking from the NBA API. With being able to process a full game, one implementation I did was use Dynamic Time Warping to run two clips from separate games of similar looking SLOB's to see if it could identify plays which in the below sample has worked. Using this, a team could grab clips of teams running plays or sets and index them, points of attack, personnel, first and second options, frequency and success rate information to build out scouting reports and tendencies. The next step is to create play diagrams automatically based on the clip FastDraw-style for easy scouting insertion to coaching staff and players documents. This is part one of a greater project that will be showcased over the next few days where its incorporated into an all purpose basketball operations artificial intelligence. The clip below is the spliced together vision tracking analyzing both play segments but a full game scan is possible. Would love to hear feedback and see any questions you might have. Again, huge thanks to @skalskip92 his resources were great to springboard into something compelling to myself.

What's going on everyone! It's been awhile since I've been on here but I wanted to share some cool updates. I am nearly done with earning my Master's at Northwestern and am going to share some NBA related projects I've done on the side over the next few days. Hows everyone doing?

I hope all my 2K and 2K League people are doing good, just wanted to update you guys that I am about to start attending Northwestern University for my Masters in Sports Management. The NBA showed me so many great things and my plan is to grow and continue that journey.

WAIT WHAT DID THE LAKERS JUST DO?

Just an FYI to all past 2K League players or coaches I would HIGHLY recommend you downloading the past twitch broadcasts of your games or any special moments such as the draft. Somewhere down your road of life you are going to watch to revisit them and they might be gone.

Easily a top 5 experience I’ve had in the NBA 2K League. I’ll always be thankful for this chance,

Sorry going down memory lane with some of this stuff 🤣