Sports fans: Analytics vs the eye test? What do you trust more?

[Wolfie*]

Analytics is being more and more introduced into team sports as a way to show who individually is more efficient but often times from what I've seen, the numbers don't tell the full story base on what we call the eye test. With that being said, should we hold a lot of stock to analytics to begin with?

How does analytics not show the full story? I'll give you examples of how analytics do not tell the full story in regards to the nba. Take a look at a rookie kevin durant in 2005. At the time, he had a historically a bad +/- when he was a rookie, yet base on the eye test, you can tell he was going to be a lethal scorer base on what he was showing on the court, yet analytics showed him being an liability on the court.

Since I'm a Lakers fan, this season our Rookie Kyle Kuzma(not Lonzo unfortunately), has shown a bad +/- especially in the 4th quarter, yet if you actually watch the games closely you can tell he gives us the best chance to win when he is actually playing despite what his analytics say.

As a sports fan, what do you think of analytics? Do you hold a lot of stock into them? I'm not saying it's always wrong or it's bs because Lebron has been one of the most efficient players over the years and that it is true, even passing the eye test as well. Thoughts?

[money4100]

Player's +/- is only an indicator of how the team plays during the span that the given player is on the field, it can mean that his teammates are some scrubs that cant play, also it can mean that a player still has to grow as a team player. It could mean that teammates are not very fond of him and the team chemistry is all wrong during the time he plays. Maybe he doesnt get his teammates involved, maybe he's a bad leader. It could mean that teammates don't like when he's playing so they are sabotaging him (it might be not possible but lets assume for the sake of argument that it is). Finally it can mean he's a terrible player.
The point is, you can make millions of conclusion based on this. The data is not conclusive, especially that the +/- can be affected by a lot of things not connected to the player's abilities. When trying to check which hypothesis is more probable you have to be very careful and analyze more sets of data. In +/- example it could be +/- of other players.
Analysts rather than look on Kd's +/- could have seen his 20 ppg or his great FT percentage, from which you could predict his potential as a great shooter despite his subpar 3pt and FG percentages.

I personally think that statistical analysis is more reliable when put in a larger context. Using data you can conclude what type of teams historically had more success, or what type of players these teams that accomplished something had. You can analyze what type of plays are more reliable and find those plays that dont work.

Finally, data sets that are being stored are only getting bigger so the analysts role will only grow.

[The Big Dunn]

Depends on what type of player.

IMO analytics are good for analyzing support players. However, when we talk about franchise box office players and all stars, it isn't a good measure because analytics do not take into consideration the job of a LeBron or Kobe is much more than just winning games-it's making sure the advertisers and networks keep giving the NBA the cash.