How Unreliable is the 'Eye-Test' for NBA analysis?

[Yoshun]

I hate the term "eye test." Can we just call it "observation" instead?

[User_friendly]

The “standard” eye-test for judging players is very unreliable
But the “standings” eye-test aka “his team is winning” is lazy, dumb and blind

[TheGeneral99]

I mean you need BOTH the eye test and the advanced stats (analytics). If both correspond it probably means it's right.

For example, a guy like Bradley Beal had superstar level stats in his peak but by the eye test he never struck me as a top tier player. His team success and advanced stats also didn't support him as a top tier player.

This is similar to a guys like Lavine who also have great stats but if you watch him play he isn't a high IQ player who makes players better and is turnover prone.

On a different note as a Raptor fan during Lowry's prime he might have only had a 15, 5 and 5 game but if you watched the game you would see that he dominated on both ends, taking charges, setting great screens, boxing out, getting deflections, controlling the pace, getting his teammates into their spots etc. This is why Lowry was always a top 10 player in RPM and WS between 2015 and 2020.

[slick_watts]

This is similar to a guys like Lavine who also has great stats but if you watch him play he isn't a high IQ player who makes players better and is turnover prone.

this is a great example of the limitations of 'eye test'.

[JonFromVA]

The "Eye-Test" is pretty unreliable, but when you watch a team game in and game out you start to understand context and that's invaluable when analyzing statistics.

[SNPA]

So flawed a premise.

Data. You use it as though only numbers are data. That is wrong. You also use it as though this type of data has no “bias.” Also flat wrong.

Everything not quantitive you lump into one category and call it the eye test. So everything that impacts a basketball game that can’t be put on a spreadsheet is eye test.

Everything can be expressed by numbers. Whether we are expressing it all or whether we are expressing it correctly is definitely a valid debate and rebuttal. But all data can be numbers. And numbers should play a disproportionate role; math and numbers are the fundamental tool in human development. There is no reason why basketball is any different.

Love. Show me the numbers.

[dhsilv2]

I mean you need BOTH the eye test and the advanced stats (analytics). If both correspond it probably means it's right.

For example, a guy like Bradley Beal had superstar level stats in his peak but by the eye test he never struck me as a top tier player. His team success and advanced stats also didn't support him as a top tier player.

This is similar to a guys like Lavine who also have great stats but if you watch him play he isn't a high IQ player who makes players better and is turnover prone.

On a different note as a Raptor fan during Lowry's prime he might have only had a 15, 5 and 5 game but if you watched the game you would see that he dominated on both ends, taking charges, setting great screens, boxing out, getting deflections, controlling the pace, getting his teammates into their spots etc. This is why Lowry was always a top 10 player in RPM and WS between 2015 and 2020.

This feels like a great example of where you found 3 guys who all were incorrectly rated by most people's eye tests and stats, especially those that are more advanced all told a more accurate story which could be supported by more advanced viewer's eyes too.

Like beal had good per game scoring, but even low level box score analysis would leave you debating him being a top 15 guy, and digging in you start debating if he's even an allstar. And again a bad eye test, oh yeah he scores a lot. But looking deeper...kinda gets over powered, doesn't seem to pass that well, his reads are limited, poor defense...just doesn't seem to be making his team meaningfully better, but it's nice he can create shots for himself.

[Heej]

The only time eye test matters is when you're hand tracking how the player responds to different defensive coverages or progresses thru reads on offense

Other than that it's a whole lot of recency bias until you see the overall strike rate on things

[NbaAllDay]

I hate the term "eye test." Can we just call it "observation" instead?

You could. However observation is usually deemed so as part of any scientific study or anaylsis. Where (and this is a ballpark) the majority of the actual analysis is based on 'numbers' and it's proven through putting it in action and 'observing it' Which is more just a confirmation or rejection of the numbers crunched.

It's rare that any field has 'observation' deemed as reliable source. NBA shouldn't be any different, especially given how intricate it is.

[NbaAllDay]

The majority of responses I am seeing from people who lean into the 'eye-test' over any form of stat seem to have a commonality around not trusting stats, or seeing them as biased/more biased.

Can I ask how much you understand the data available. Do you have a solid understanding of the 20+ data points used and how 'impactful' they are? This is not box score, stats. This is more to do with all the advanced data sets available.

If I were to take a guess, i'd say those that favour the eye-test often do so as they are unwilling or are turned off from the data. It could even include data sets that go against what you believe.

All of this is fine and to each their own. I just believe it's a bit disingenuous to disregard data of which you haven't even attempted to properly understand.

[SNPA]

The majority of responses I am seeing from people who lean into the 'eye-test' over any form of stat seem to have a commonality around not trusting stats, or seeing them as biased/more biased.

Can I ask how much you understand the data available. Do you have a solid understanding of the 20+ data points used and how 'impactful' they are? This is not box score, stats. This is more to do with all the advanced data sets available.

If I were to take a guess, i'd say those that favour the eye-test often do so as they are unwilling or are turned off from the data. It could even include data sets that go against what you believe.

All of this is fine and to each their own. I just believe it's a bit disingenuous to disregard data of which you haven't even attempted to properly understand.

Stats people are amazing. They really think they have the panacea and everyone else is wrong. If only the plebeians could see!

Lol.

No. It’s all just different languages to describe reality. You like to translate reality into numbers to describe it. Others like to use Spanish. Others like to use English. They are all translations of reality. Numbers just give the illusion of objectivity, and many people fall for that.

[NbaAllDay]

The majority of responses I am seeing from people who lean into the 'eye-test' over any form of stat seem to have a commonality around not trusting stats, or seeing them as biased/more biased.

Can I ask how much you understand the data available. Do you have a solid understanding of the 20+ data points used and how 'impactful' they are? This is not box score, stats. This is more to do with all the advanced data sets available.

If I were to take a guess, i'd say those that favour the eye-test often do so as they are unwilling or are turned off from the data. It could even include data sets that go against what you believe.

All of this is fine and to each their own. I just believe it's a bit disingenuous to disregard data of which you haven't even attempted to properly understand.

Stats people are amazing. They really think they have the panacea and everyone else is wrong. If only the plebeians could see!

Lol.

No. It’s all just different languages to describe reality. You like to translate reality into numbers to describe it. Others like to use Spanish. Others like to use English. They are all translations of reality. Numbers just give the illusion of objectivity, and many people fall for that.

It's very much a false equivalence using different verbal languages as an equal comparison to what we visual on the court, versus the data thats provided.

I asked you on another thread what evidence you had for Bird being a 'killer' and you said: "His Career" and refused to even elaborate outside of one anecdotal video. You are basically saying 'trust me bro'

The fact you are saying 'stats people' it's very clearly evident that you have no idea how to even interpret them. Which again, is fine, but your ignorance isn't my problem. Nor is it an excuse to justify your eyes as a reliable source.

It's also got nothing to do with wrong or right.

Think of it as a sliding scale. The more complex something is, the more data is generally required in order to better interpret what is going on.

This idea that you can make all your judgements based on what you see is an argument i'd expect from a 12 year old and I've already seen your level of commentary in a previous thread so I don't really see anything productive out of discussion further with you.

[SNPA]

The majority of responses I am seeing from people who lean into the 'eye-test' over any form of stat seem to have a commonality around not trusting stats, or seeing them as biased/more biased.

Can I ask how much you understand the data available. Do you have a solid understanding of the 20+ data points used and how 'impactful' they are? This is not box score, stats. This is more to do with all the advanced data sets available.

If I were to take a guess, i'd say those that favour the eye-test often do so as they are unwilling or are turned off from the data. It could even include data sets that go against what you believe.

All of this is fine and to each their own. I just believe it's a bit disingenuous to disregard data of which you haven't even attempted to properly understand.

Stats people are amazing. They really think they have the panacea and everyone else is wrong. If only the plebeians could see!

Lol.

No. It’s all just different languages to describe reality. You like to translate reality into numbers to describe it. Others like to use Spanish. Others like to use English. They are all translations of reality. Numbers just give the illusion of objectivity, and many people fall for that.

It's very much a false equivalence using different verbal languages as an equal comparison to what we visual on the court, versus the data thats provided.

I asked you on another thread what evidence you had for Bird being a 'killer' and you said: "His Career" and refused to even elaborate outside of one anecdotal video. You are basically saying 'trust me bro'

The fact you are saying 'stats people' it's very clearly evident that you have no idea how to even interpret them. Which again, is fine, but your ignorance isn't my problem. Nor is it an excuse to justify your eyes as a reliable source.

It's also got nothing to do with wrong or right.

Think of it as a sliding scale. The more complex something is, the more data is generally required in order to better interpret what is going on.

This idea that you can make all your judgements based on what you see is an argument i'd expect from a 12 year old and I've already seen your level of commentary in a previous thread so I don't really see anything productive out of discussion further with you.

Touched a nerve?

Find where I said quantitative has no role or isn’t useful? It obviously has a role. It’s obviously useful.

Quantitative is also a translation. It isn’t reality…I hope you can agree with that. It is a way to record reality and it is used to interpret/understand reality…just like…wait for it…any other language.

Bird is a killer. It requires no argument or defense. Ask anyone who played against him.

[scrabbarista]

Some people's eyes are better than others.

The main benefit of stats is that they capture what the eyes, for whatever reason, didn't see. That said, on an action-by-action basis, I would far rather see what took place than see it described in print - whether that description is numerical or not. The "eye-test" is ultimately superior - when you can get it - but stats are essential due to limitations on our time and ability/skill in observing the game

[NbaAllDay]

Stats people are amazing. They really think they have the panacea and everyone else is wrong. If only the plebeians could see!

Lol.

No. It’s all just different languages to describe reality. You like to translate reality into numbers to describe it. Others like to use Spanish. Others like to use English. They are all translations of reality. Numbers just give the illusion of objectivity, and many people fall for that.

It's very much a false equivalence using different verbal languages as an equal comparison to what we visual on the court, versus the data thats provided.

I asked you on another thread what evidence you had for Bird being a 'killer' and you said: "His Career" and refused to even elaborate outside of one anecdotal video. You are basically saying 'trust me bro'

The fact you are saying 'stats people' it's very clearly evident that you have no idea how to even interpret them. Which again, is fine, but your ignorance isn't my problem. Nor is it an excuse to justify your eyes as a reliable source.

It's also got nothing to do with wrong or right.

Think of it as a sliding scale. The more complex something is, the more data is generally required in order to better interpret what is going on.

This idea that you can make all your judgements based on what you see is an argument i'd expect from a 12 year old and I've already seen your level of commentary in a previous thread so I don't really see anything productive out of discussion further with you.

Touched a nerve?

Find where I said quantitative has no role or isn’t useful? It obviously has a role. It’s obviously useful.

Quantitative is also a translation. It isn’t reality…I hope you can agree with that. It is a way to record reality and it is used to interpret/understand reality…just like…wait for it…any other language.

Bird is a killer. It requires no argument or defense. Ask anyone who played against him.

At least you are consistently clueless i'll give you that Happy hunting.

[MrBigShot]

Eye test for general things. You can tell whether a guy is empty stats by his style of play. If he's a blackhole, dominates the ball when he'd never touch on a good team, doesn't play any defense ect...

Some things you need #s to be able to gauge. Most people would think Giannis is having a better season than SGA if we didn't have impact numbers to be able to look at.

[Petergrifindor]

Best method.

Nothing beats actually watching the games.

[SNPA]

It's very much a false equivalence using different verbal languages as an equal comparison to what we visual on the court, versus the data thats provided.

I asked you on another thread what evidence you had for Bird being a 'killer' and you said: "His Career" and refused to even elaborate outside of one anecdotal video. You are basically saying 'trust me bro'

The fact you are saying 'stats people' it's very clearly evident that you have no idea how to even interpret them. Which again, is fine, but your ignorance isn't my problem. Nor is it an excuse to justify your eyes as a reliable source.

It's also got nothing to do with wrong or right.

Think of it as a sliding scale. The more complex something is, the more data is generally required in order to better interpret what is going on.

This idea that you can make all your judgements based on what you see is an argument i'd expect from a 12 year old and I've already seen your level of commentary in a previous thread so I don't really see anything productive out of discussion further with you.

Touched a nerve?

Find where I said quantitative has no role or isn’t useful? It obviously has a role. It’s obviously useful.

Quantitative is also a translation. It isn’t reality…I hope you can agree with that. It is a way to record reality and it is used to interpret/understand reality…just like…wait for it…any other language.

Bird is a killer. It requires no argument or defense. Ask anyone who played against him.

At least you are consistently clueless i'll give you that Happy hunting.

I have noticed with a little push back you guys tend to fold. You actually aren’t as bad as some of your brethren. You got people on your side making absurd statements like:

“Everything can be expressed by numbers.”

It’s just comical. The give away is how the whole thing is divided into two parts… quantitive and “eye test” by which you mean literally every other factor that impacts how humans would interact. Little imbalanced, right?

Show me how you calculate for the psychology of the players and group(s)?

Show me how you calculate for the cultural elements of players backgrounds, locales and teams?

Show me how you calculate for team chemistry?

Oh…that’s right. There is no stat for those and you can’t. So instead you lump them into an absurd category (a favorite technique of stat nerds) and call the whole thing eye test and talk bad about how it’s inferior to numbers. The language of stats isn’t able to describe all elements of human behavior. This is not making widgets or A/B testing, this is a group of humans we are talking about.

If we both had the same team and access to the same basic counting stats and you got access to all the advanced stats and I got access to the team psychologist who do think would be a better coach?

[LLJ]

It's flawed, but so is not watching someone and just looking at stats. Watching games you sort of get a feel for which players tend to tangibly impact momentum shifts. But, just because you see someone do a few cool moves or score a lot of points doesn't mean they're a great player either.

If you only follow stats, you can get a sense of a player's skillset but you don't get a sense of how they affect the game. There are players who put up good boxscore stats but when you watch them they seem to only put up numbers when the game is out of reach or during times when the team is already rolling. You don't get a sense of if a player can turn it up when his team is down and needs a push to tie the game or get a lead unless you deep dive into the situational stats, but even those can be flawed if the player tends to play with certain lineups regularly.

[NbaAllDay]

Touched a nerve?

Find where I said quantitative has no role or isn’t useful? It obviously has a role. It’s obviously useful.

Quantitative is also a translation. It isn’t reality…I hope you can agree with that. It is a way to record reality and it is used to interpret/understand reality…just like…wait for it…any other language.

Bird is a killer. It requires no argument or defense. Ask anyone who played against him.

At least you are consistently clueless i'll give you that Happy hunting.

I have noticed with a little push back you guys tend to fold. You actually aren’t as bad as some of your brethren. You got people on your side making absurd statements like:

“Everything can be expressed by numbers.”

It’s just comical. The give away is how the whole thing is divided into two parts… quantitive and “eye test” by which you mean literally every other factor that impacts how humans would interact. Little imbalanced, right?

Show me how you calculate for the psychology of the players and group(s)?

Show me how you calculate for the cultural elements of players backgrounds, locales and teams?

Show me how you calculate for team chemistry?

Oh…that’s right. There is no stat for those and you can’t. So instead you lump them into an absurd category (a favorite technique of stat nerds) and call the whole thing eye test and talk bad about how it’s inferior to numbers. The language of stats isn’t able to describe all elements of human behavior. This is not making widgets or A/B testing, this is a group of humans we are talking about.

If we both had the same team and access to the same basic counting stats and you got access to all the advanced stats and I got access to the team psychologist who do think would be a better coach?

Every heard of a straw man?

I never said the eye test included 'everything else'. You keep pretending that's what I said so you can ramble on about things to support your narrative.

It's exactly what 'you people' do. Since you aren't able to have an honest discussion on its own merits.

You talk about cultural elements, players backgrounds and locales. Or Whether or not they have an effective psychologist. These things can impact a person's mindset in a number of ways. So show me how it impacts their game?

Did someone having a different cultural impact mean they were a less efficient? If so how?

They grew up in a poor neighbourhood, is there ability to pass the ball hindered?

Even if they said it was true, how could your
proove it? What relevance does it have when comparing two players impact on the court?

You claim these things have an impact yet havent given a shred of evidence as to how they impact a player. I'm not even asking for 'stats' like a nerd. Anything that's not laughable will do.

It's easier to believe that I am 'folding' or 'struck a nerve' then it is to address what's being said.

So as I said before I'll leave it there. It's clear this conversion isnt going anywhere. You are welcome to reply but that will be it for me.