what a title – first the youtube video demo/pitch for the “bowling analytics” product …
statistics vs analytics
Yes, there is a difference between “statistics” and “analytics” – maybe not a BIG difference but there is a difference.
“Statistics” is about collecting and interpreting “masses of numerical data.” “Analytics” is about logical analysis – probably using “statistics”.
Yeah, kinda slim difference – the point being that there is a difference between “having the numbers” and “correctly interpreting the numbers.”
“Data analysis” becomes an exercise in asking questions and testing answers – which might have been how a high level “statistician” described their job 100 years ago – i.e. I’m not dogmatic about the difference between “statistics” and “analytics”, just establishing that there are connotations involved.
Analytics and Sports
Analytics as a distinct field has gained popularity in recent years. In broad strokes the fields of “data science”, “artificial intelligence”, and “machine learning” all mean “analytics.”
For a while the term “data mining” was popular – back when the tools to manage “large data sets” first became available.
I don’t want to disparage the terms/job titles – the problem is that “having more data” and having “analysis to support decisions” does not automatically mean “better leadership.”
It simply isn’t possible to ever have “all of the information” but it is very easy to convince “management types” that they have “data” supporting their pet belief.
e.g. I always like to point out that there are “trends” in baby name popularity (example site here) – but making any sort of conclusion from that data is probably specious.
What does this have to do with “sports” – well, “analytics” and sports “management” have developed side by side.
Baseball’s word for the concept of “baseball specific data analysis” dates back to 1982 – about the time that “personal computers” where starting to become affordable and usable by “normal” folks.
My round about point today is that most “analytics” fall into the “descriptive” category by design/definition.
e.g. if you are managing a ‘sportball’ team and have the opportunity to select players from a group of prospects – how do you decide which players to pick?
Well, in 2022 the team is probably going to have a lot of ‘sportball’ statistics for each player – but do those statistics automatically mean a player is a “good pick’ or a “bad pick”? Obviously not – but that is a different subject.
The team decision process will (probably) include testing players physical abilities and watching the players work out – but neither of those 100% equates to “playing the game against other skilled opponents.”
That player with great statistics might have been playing against a lower level of competition. That player that has average “physical ability test scores” might be a future Hall of Famer because of “hidden attributes”
i.e. you can measure how fast an athlete can run, and how high they can jump – but you can’t measure how much they enjoy playing the game.
MEANWHILE back at the ranch
Now imagine that you are an athlete and you want to improve your ‘sportball’ performance. How do you decide what to work on?
Well, the answer to that question is obviously going to be very sport AND athlete specific.
However, your ‘sportball’ statistics are almost certainly not going to help you make decisions on how/what you should be trying to develop – i.e. those statistics will be a reflection of how well you have prepared, but do not directly tell you how to prepare.
Bowling
Full disclosure – I am NOT a competitive bowler. I have participated/coached other sports – but I’m a “casual bowler.” i.e. if I have misinterpreted the sport, please let me know 😉
Now imagine that someone has decided that they want to improve their “bowling average” – how should they approach the problem?
- Step 1 would be to establish a baseline from which improvements can be measured.
- Step 2 would be to determine what you need to “work on” to improve your scores from Step 1.
- Step 3 would be to establish a session of “practices” to work on the items from Step 2.
- Step 4 would be to re-test the items from Step 1 and adjust steps 2 and 3 accordingly.
Sure, I just described the entire field of “management” and/or “coaching” – but how well a manager/coach helps athletes through the above (generic) process will be directly reflected in wins/losses in competition.
Remember that the old axiom that “practice makes perfect” is a little misleading:
Practice does not make perfect. Only perfect practice makes perfect.
-Vince Lombardi
Back to bowling – bowling every week might be fun, but won’t automatically mean “better performance.”
Keeping track of your game scores might be interesting, but also won’t automatically mean “better scores.”
I’m told that the three factors for the “amateur bowler” to work on are:
- first ball pin average
- single pin spare %
- multipin spare %
In a “normal” game there are 10 pins possible each frame. The bowler gets two balls to knock down all 10.
If your “first ball pin average” is 10, then you are a perfect bowler –and knock all the pins down every frame with your first ball.
To be honest I haven’t seen any real data on “first ball pin averages” – it probably exists in much the same manner that “modern baseball statistics” can be derived from old “box scores” – but I’m told that a first pin average around 9 is the goal.
If you consistently average 9 pins on your first throw – then you have a consistent “strike” delivery.
Which then means that IF you consistently knock down 9 pins – you will have to pickup “single pin spares” on a regular basis.
Then “multipin spares” are going to be an exercise in statistics/time and fate. Obviously if you average 9 pins on your first ball, the number of “multipin spare” opportunities should be relatively small.
SO those are the data points being tracked with my “bowling analytics” application.