Tuesday Morning Quarterback – 49ers

Welcome to Tuesday Morning Quarterback! Go Niners!

The Xcalar team love sports, almost as much as we love big data. The teamwork, the competition, the thrill of victory, and the agony of defeat. But as data freaks, we also love that sports is now as much about the data as the sport itself. Play-by-play data, historical analysis, ML predictive analysis, what-ifs, fantasy teams – sports is just big data wrapped in cool uniforms.

So it is only natural that we are introducing our Tuesday Morning Quarterback event. Each Tuesday morning, we will shoot the breeze on the past weekend’s/Monday’s games and use data to back each other’s otherwise blustery prognostications.


Did you miss this morning’s Tuesday Morning Quarterback event? Check out the video!


The Technology

Between the Monday night’s game and Tuesday morning’s webinar, we only had a few hours to put the analytics together. Fortunately, we have this analytic pipeline fully automated on AWS and Xcalar.

Operationalizing Sports Data from S3 to Xcalar to Tableau

  1. At around 10pm, we received the play-by-play data from NFLSavant
  2. This was dropped into S3
  3. AWS Cloudwatch and Lambda automatically detect the new file and kick-off Xcalar’s NFL data flows that bring in the play-by-play data along with the NFL’s team roster data.
    • Building the dataflow was not an easy task, much of the play-by-play data was embedded in text fields – we use a mixture of python, SQL and tabular manipulation to extract and harmonize the data.
    • As an example, the play-by-play data referred to players as NUMBER-FIRST_INITIAL.LAST_NAME
      • e.g. 10-J.Garoppolo for Jimmy Garoppolo
      • We used Python to transform our roster data into this form (10-J.Garoppolo) and used this to parse out the play-by-play data.
  4. The results of the analytics was then automatically pushed into a Tableau dashboard

With Xcalar, using a mix of Python, SQL, Tabular transformation to determine 49er passing and rushing success.

 


49ers vs ClevelandBrowns – A Crushing Rushing Game

49ers played a rush-focused offense.

For our first Tuesday Morning Quarterback session, we took on last night’s 49ers vs Cleveland Browns game.
This was a highly anticipated game with both teams hyped up – 49ers with Jimmy G and Brown’s with OBJ (Odell Beckham Jr).

Analysts were expecting a rush-focused offense for the 49ers…. They were not disappointed because the Niners dismantled the Cleveland Browns defense (says Xcalar’s Omar) (see below and to the right).

Rushing plays were more successful than passing plays.

Monday’s game vs. historical data

We also specifically looked at Jimmy G’s favorite teammate with whom he executes most plays. For Monday night, we can see that Jimmy G looked to Breida and Coleman the most. For the season to date, Wilson, Tabor, and Mostet were the preferred teammates.

Jimmy G’s Favorite Teammates (Monday vs Season to Date)


This was really our first foray in to sport commentary; while the event is “Tuesday Morning Quarterback” we will be tackling a wide variety of sports statistics including the NBA, English Premier League, Baseball, E-sports, and even Cricket! Some of these webinars will be about taming all the wild sports data, others will be about operationalization, next Week, Tuesday Morning Quarterback will be using ML to predict game outcomes! You’re not going to want to miss it!

until next Tuesday,
The Xcalar Xports Desk