The Golden State Warriors – Data & Talent = Heart!
The Golden State Warriors – Data & Talent = Heart! A blog about the use of data analytics by an NBA franchise that led to the end of a 40 year NBA finals championship drought.
The GSW – Data & Talent = Heart !
The Warriors and …. Data?
They Golden State Warriors (GSW), have won 2 NBA championships in the last 3 years, made it to the NBA finals every year consecutively for three of those years and in the process set a new regular season record with 73 wins, (dethroning 1995-1996 Bulls’ record). But how did the GS warriors reappear in the NBA finals after a dormant 40 years?! The recipe is still being analyzed and debated by media talking heads and analysts, while all agree on one ingredient – data science.
Context:
The GSW ‘Data-era’ is rumored to have begun with the change of ownership in 2010. Many cite the ownership’s venture capital and tech savvy background as the engine behind the adoption of a numbers strategy, (which may or may not have led to the team’s tagline: “Strength in numbers”). The franchise began experimenting with a data analytics focused approach and was one of the few early adopters of a new camera based technology called SportsVU, (cameras situated above the arena capturing plays frame by frame). To avoid ‘betting the team’ the franchise began experimenting on the Santa Cruz G league. Assistant GM Kirk Lacob led a data focused approach to analyze the team’s data: games, practices etc. The strategy was later adopted by the GSW front office.
Creating and capturing value?
NBA analysts and journalists agree that the GSW posses great players talent, strong coaching and medical staff, and a dedicated front and back office. But even the official franchise position has been that they aim to be early adopters in the field of data insights. Lacob has stated that the reason they wanted to be pioneers in this field was to use data insights as an enhancer to their decision making processes. Among others, the GSW franchise has used data to gain insights from box score, geospatial analysis, and most recently wearable technology. A few examples that have been publicly discussed are:
- Defending opponents three point shooting: The franchise worked with MOCAP Analytics, and shared their findings on their perceived root problem with the coaching staff. These included heat maps, and measurable quantifications of indicators such as ‘pick and roll’, ‘ball touches’ etc.
- Optical tracking has been used to create customized workout for players. Models generate an Estimated Possession Value, (EPV), at any given moment based on all 10 players’ position and possession of the ball. This has also been effective when engaging in trade negotiations to illustrate and prove value.
- Rest – The GSW used reliable data to manage players minutes as evident in the 2017 season. Lebron James and Kyrie Irving led the chart in total season minutes at 1st and 15th places. The first Golden State Warrior player, Klay Thompson, only shows up at 28th while Steph Curry and Kevin Durant only show up at 38th place. This was instrumental in their finals series.
- Wearables and readiness – The coaching staff is presented with a readiness rating for each player, which is an synthesis of quantitative and qualitative health indicators. These include daily surveys and workload measured by wearables. Among the input sources are SportVU, (mentioned above), Catapult data for practice workloads and Omegawave heart variability data to test neurological stress.
- The Three point decision: Many have dubbed this shift as simple math and that the real innovation here was acknowledging it, and more importantly following through to execute consistently. The new franchise owners focused on what some have called the ‘3-point line market inefficiency’ – shooting percentage for players right within or without the line were almost identical. The average shot value of a shot beyond the line was worth 1.09 vs. a value of 0.76 within the line, leading to a ~43% rate of return for players shooting the 3pt shot. The rest is ‘Steph and Klay history’.
Smooth Sailing From Here On Out?
Internally, I think the GSW have proven to be adopters of early technology and that they are constantly on the look out for an analytical edge. However, looking at the league wide adoption of SportVU and teams like the Houston Rockets leading the league in 3-point shot attempts, I see a catching up trend the Warriors should be worried about. Could the rest of the NBA be soon a leveled playing field? Or will talent, coaching, injuries and luck all make it very hard to judge this exact question and also serve as distractors for teams to fully adopt the analytical approach? Time will tell.
Sources:
- https://www.nytimes.com/2016/04/03/magazine/what-happened-when-venture-capitalists-took-over-the-golden-state-warriors.html
- https://www.wsj.com/articles/the-golden-state-warriors-have-revolutionized-basketball-1459956975
- http://dataconomy.com/2017/07/golden-state-warriors-data-science/
- https://www.stats.com/sportvu-basketball/
- https://blog.chartio.com/posts/how-analytics-drives-the-golden-state-warriors
- http://bigdata-madesimple.com/big-data-meets-basketball-golden-state-warriors-up-their-game/
- https://www.popsci.com/math-wizard-kirk-lacob-how-golden-state-warriors-won-73-games
- https://channels.theinnovationenterprise.com/articles/how-tech-helped-the-nba-s-golden-state-warriors-win-a-championship
- http://www.reltio.com/blog/2017/6/data-drives-golden-state-warriors-to-another-nba-title
- http://www.reltio.com/blog/2015/6/warriors-get-a-data-driven-assist-to-win-nba-championship
- http://www.ibmbigdatahub.com/blog/taking-data-analytics-hoop
- https://www.sfgate.com/warriors/article/Golden-State-Warriors-at-the-forefront-of-NBA-5753776.php
- https://www.teamrankings.com/nba/stat/three-pointers-attempted-per-game
Thanks of sharing this post – super interesting. You raise an interesting point – what happens in the end game when everyone has this technology in the NBA. If it then becomes about scouting raw talent, that can be digitized too. What can’t (at least in LIVE basketball) is the players themselves – I think that there are some forces in the sport that lead players to stay with the same team instead of transferring each season: it takes time to get the team to play well together as a team and also fans like to have some consistency in the roster. So maybe we need institutions in form of the NBA and the draft system to save us and help make sure that the it doesn’t become a winner takes all market … where good teams accumulate wealth, buy good players, get more wealth … and become unbeatable?
Interesting post. The Warriors are a model franchise and an exemplar of how to build a team in today’s NBA. They have used advanced analysis to better inform in-game and out-of-game strategies. However, they have also aggregated the best starting 5 in NBA history, largely through timing, good drafting, and fortuitous contracts (Steph curry being on cheap deal after his ankle injuries, Draymond Green in the second round, Durant becoming a free agent during the one year the salary cap jumped).
With a roster that possesses an unprecedented level of talent, it is unclear how much the other things they do matter, and how much of what they do was built around the talent they aggregated. For example, the Warriors take a lot of threes and were one of the main catalysts for the 3-pointer becoming the weapon it is in the NBA today. However, they also drafted the best shooter in NBA history (Steph Curry), and a few years later one of the best ever in Klay Thompson. Curry was picked by the previous ownership, and no one knew he’d be the player he was. Many might ask, did the Warriors initiate the move towards high dependence on 3-pointers, or did they simply adjust to the talents of their best players? Either way, they get a lot of credit. However, I do not think they have intentionally used data to drive the strategy of a franchise as much as others like Daryl Morey has in Houston.
Thanks for sharing this great post! I share your concerns regarding the leveling of the data playing field as more teams adopt this technology. I wonder how far up the innovation curve the GSWs are compared to teams like Houston and if they will be able to sustain a competitive advantage. I also wonder who owns the data generated by each player: the franchise or the player? Will a player and his agent be able to take his data set with him to new teams? As more and more teams adopt this technology, I’m sure player contracts will need to account for this question and others.
Good Informative post.. Thanks for sharing this info with interesting point. Keep maintaining this.
Data Science Course is in Big Demand now with #1 Place in National and International Job Market. We also Provide Data Science Classroom Training in Hyderabad and Data Science online Training for rest of the world audience.
https://socialprachar.com/data-science/