AI algorithm’s top 11 NBA players

[RogueBorg]

The NBA has been around for quite some time, so understandably, quite a few legends have taken the floor throughout the league's history. Everybody will have a different opinion on who the best players in the league's history are, as it is virtually impossible to create one list that can be set in stone and viewed as the correct answer. Well, an AI algorithm created the perfect all-time list using pretty much every stat known to man, and the results are somewhat shocking.

"A data scientist put together an algorithm with every single statistical thing known to man in the NBA, and here was the list:

11. Shaquille O'Neal

10. Larry Bird

9. Magic Johnson

8. Wilt Chamberlain

7. Hakeem Olajuwon

6. Bill Russell

5. Kobe Bryant

4. LeBron James

3. Kareem Abdul-Jabbar

2. Tim Duncan

1. Michael Jordan

So here's my question; are we underrating Tim Duncan?"

Right off the bat, a few things jump out at you when looking at this list. Tim Duncan as the second-best player of all-time, is way higher than pretty much anyone ever puts him, while LeBron James falling to fourth is way lower than many folks would put him. In some cases, fans believe that James is the greatest player of all time. He's still adding to his accomplishments, but fourth feels way too low for him here.

Aside from that, some fans may be surprised to see Hakeem Olajuwon come in ahead of guys like Wilt Chamberlain and Magic Johnson, with Chamberlain, in particular, seeming to be a bit too low on this list. Even Bill Russell, who has garnered a lot more praise for his NBA career since his untimely passing, maybe too high on this list for some folks. The players themselves aren't the problem, but the AI's list here certainly has some interesting placements of the players they included that will likely catch fans' attention.

https://www.msn.com/en-us/sports/nba/ai-algorithm-s-top-11-nba-players-of-all-time-list-will-make-lebron-james-fans-furious/ar-AA1cITYk?ocid=entnewsntp&pc=U531&cvid=5138cc9e65ab447790ff935589e972f3&ei=14

[ArbitraryWater]

What ****ing stats are those :oldlol:


cause this is the weirdest top 11 Ive ever seen

[FKAri]

Nothing out of the ordinary other than Duncan.

There are so many random judgement calls that would need to be made when compiling such a list. Like longevity vs peak. How was that weighed? What about performance variance? Performing in the finals vs regular season? These are all judgement calls with no right answer. Play around and get vastly different results. Get a result you like and publish an article.

[Full Court]

No way should Kobe ever be ranked above Shaq.

[FultzNationRISE]

Nothing out of the ordinary other than Duncan.

There are so many random judgement calls that would need to be made when compiling such a list. Like longevity vs peak. How was that weighed? What about performance variance? Performing in the finals vs regular season? These are all judgement calls with no right answer. Play around and get vastly different results. Get a result you like and publish an article.

Yup.

Reminds me of the story of John Hollinger formulating PER.

First time he did it he compiled all the stats he thought relevant, and weighted them the way he thought was most appropriate, tallied up the results and found that… David Robinson was the greatest player of all time.

He decided that wouldnt be accepted so he tweaked the formula until it declared MJ to have the highest PER, and then made that the formula for PER.

Which basically makes the whole formula meaningless and the least scientific thing ever, since he decided what the outcome should be first and then came up with values that would support that conclusion.

Intangibles are, by definition, unquantifiable, so theres no statistical formula that can tell you who’s better than who so long as intangibles are part of the game. Which in basketball they are.

I could see why someone thought having an AI rank things would make a good, topical article but in reality… it’s meaningless. The list bears no more value than any random human opinion.

[ILLsmak]

Computers don’t understand basketball lol time to accept it. Nor do numbers represent it. Colored text.

-Smak

[ShawkFactory]

What ****ing stats are those :oldlol:


cause this is the weirdest top 11 Ive ever seen

Is it? After Jordan, the difference between all of these guys is miniscule. Any order could realistically be argued.

[post]

Yup.

Reminds me of the story of John Hollinger formulating PER.

First time he did it he compiled all the stats he thought relevant, and weighted them the way he thought was most appropriate, tallied up the results and found that… David Robinson was the greatest player of all time.

He decided that wouldnt be accepted so he tweaked the formula until it declared MJ to have the highest PER, and then made that the formula for PER.

Which basically makes the whole formula meaningless and the least scientific thing ever, since he decided what the outcome should be first and then came up with values that would support that conclusion.

Intangibles are, by definition, unquantifiable, so theres no statistical formula that can tell you who’s better than who so long as intangibles are part of the game. Which in basketball they are.

I could see why someone thought having an AI rank things would make a good, topical article but in reality… it’s meaningless. The list bears no more value than any random human opinion.

searched for that story online before and couldn't find it

i'd be somewhat curious to read more about it

perhaps it's in a book he wrote

i probably don't care enough to track it down through inter library loan if possible or paying 30 dollars for it

i'll say this though

odds are if what you are saying is true it was close between jordan and robinson like it still is

my guess would be the scoring component was given extra weight and the advantage robinson had was eliminated

when i first saw per years ago and noticed jordan was #1 for career i thought well how wrong could it be