Masai worked with Dean Oliver. Will he prioritize 4 factors?

[Patman]

Can I just say that I developed an NCAA bracket this year that was based purely on the 4 factors theory and GOT SO BLOWN OUT OF THE WATER THAT I CANNOT SHOW MY FACE NEAR ANY NCAA POOLS AGAIN.

Not saying my experience matters for purposes of constructing an NBA team. Just relaying my experience in using the theory. And venting. Always venting.

Did you consider 4 Factors on both sides of the ball? You can have great 4 Factors on offense, but be putrid on D. Also, people almost always forget how small of a sample size the NCAA tourney is. It seems like a lot of games, but it's really not, it's only magnified. The average team in the tourney plays 2 games!

[Double Helix]

I must be a complete tool... but can someone explain what the 4 Factors Theory is? Does it combine O and D, or is it 4 Factors from each skill set?

Offence and defence. It's all about the differential. So, it's technically possible for a great defensive team with mediocre offence to have the same EFG% differential as a great offfensive team with poor defence.

The Four Factors are easy. The concept was pioneered by Dean Oliver in his 2004 book, Basketball on Paper. He analyzed thousands of box scores in order to figure out what about them really mattered. Was it rebounds? Turnovers?

He found that four factors, when weighted properly, could predict the winner of essentially every basketball game ever played.

To the surprise of no one, the most important factor is shooting. No duh, right? But the traditional use of FG% didn’t work in an era with a 3-point line. What does FG% matter when some shots are worth more than others? So he uses effective FG% (or eFG%), and all that does is add in the value of a 3-pointer. A made 3 is worth 50% more than a made 2, so that’s factored into the equation. Instead of FGM / FGA it’s (0.5*3PTM + FGM) / FGA.

Oliver calculated that eFG% accounted for about 40% of a team’s success.

Next on the list of importance (25% of a team’s success) was how well a team took care of the ball. But turnovers aren’t the best way to measure this, otherwise it’s worse to turn the ball over 15 times in a 70 possession game than 13 times in a 53 possession game. In the first example the team turned it over on 21% of their possessions (the NCAA average). Where in the second example (which had better ball security in the traditional sense) the team turned it over on 24% of their possessions.

All Oliver did was add possessions to the mix. TO% is simply a pace-independent way to measure ball security.

Accounting for 20% of a team’s success is offensive rebounding. But again, comparing raw numbers between games is meaningless, because it doesn’t account for how many opportunities a team had to actually grab an offensive board. OR% takes that into account.

Finally, getting to the line can explain 15% of a team’s success. FT Rate is the number of free throws attempted divided by the number of field goals attempted. Higher numbers are good.

[tempests_dawn]

Thanks DH. That was a great Coles notes version what the Four Factors mean

[RedX]

I must be a complete tool... but can someone explain what the 4 Factors Theory is? Does it combine O and D, or is it 4 Factors from each skill set?

Offence and defence. It's all about the differential. So, it's technically possible for a great defensive team with mediocre offence to have the same EFG% differential as a great offfensive team with poor defence.

The Four Factors are easy. The concept was pioneered by Dean Oliver in his 2004 book, Basketball on Paper. He analyzed thousands of box scores in order to figure out what about them really mattered. Was it rebounds? Turnovers?

He found that four factors, when weighted properly, could predict the winner of essentially every basketball game ever played.

To the surprise of no one, the most important factor is shooting. No duh, right? But the traditional use of FG% didn’t work in an era with a 3-point line. What does FG% matter when some shots are worth more than others? So he uses effective FG% (or eFG%), and all that does is add in the value of a 3-pointer. A made 3 is worth 50% more than a made 2, so that’s factored into the equation. Instead of FGM / FGA it’s (0.5*3PTM + FGM) / FGA.

Oliver calculated that eFG% accounted for about 40% of a team’s success.

Next on the list of importance (25% of a team’s success) was how well a team took care of the ball. But turnovers aren’t the best way to measure this, otherwise it’s worse to turn the ball over 15 times in a 70 possession game than 13 times in a 53 possession game. In the first example the team turned it over on 21% of their possessions (the NCAA average). Where in the second example (which had better ball security in the traditional sense) the team turned it over on 24% of their possessions.

All Oliver did was add possessions to the mix. TO% is simply a pace-independent way to measure ball security.

Accounting for 20% of a team’s success is offensive rebounding. But again, comparing raw numbers between games is meaningless, because it doesn’t account for how many opportunities a team had to actually grab an offensive board. OR% takes that into account.

Finally, getting to the line can explain 15% of a team’s success. FT Rate is the number of free throws attempted divided by the number of field goals attempted. Higher numbers are good.

thanks dude

[joseph235]

I must be a complete tool... but can someone explain what the 4 Factors Theory is? Does it combine O and D, or is it 4 Factors from each skill set?

Offence and defence. It's all about the differential. So, it's technically possible for a great defensive team with mediocre offence to have the same EFG% differential as a great offfensive team with poor defence.

The Four Factors are easy. The concept was pioneered by Dean Oliver in his 2004 book, Basketball on Paper. He analyzed thousands of box scores in order to figure out what about them really mattered. Was it rebounds? Turnovers?

He found that four factors, when weighted properly, could predict the winner of essentially every basketball game ever played.

To the surprise of no one, the most important factor is shooting. No duh, right? But the traditional use of FG% didn’t work in an era with a 3-point line. What does FG% matter when some shots are worth more than others? So he uses effective FG% (or eFG%), and all that does is add in the value of a 3-pointer. A made 3 is worth 50% more than a made 2, so that’s factored into the equation. Instead of FGM / FGA it’s (0.5*3PTM + FGM) / FGA.

Oliver calculated that eFG% accounted for about 40% of a team’s success.

Next on the list of importance (25% of a team’s success) was how well a team took care of the ball. But turnovers aren’t the best way to measure this, otherwise it’s worse to turn the ball over 15 times in a 70 possession game than 13 times in a 53 possession game. In the first example the team turned it over on 21% of their possessions (the NCAA average). Where in the second example (which had better ball security in the traditional sense) the team turned it over on 24% of their possessions.

All Oliver did was add possessions to the mix. TO% is simply a pace-independent way to measure ball security.

Accounting for 20% of a team’s success is offensive rebounding. But again, comparing raw numbers between games is meaningless, because it doesn’t account for how many opportunities a team had to actually grab an offensive board. OR% takes that into account.

Finally, getting to the line can explain 15% of a team’s success. FT Rate is the number of free throws attempted divided by the number of field goals attempted. Higher numbers are good.

Thanks man, I'm really into baseball sabermetrics, and am trying to get into basketball lately too, this helps a lot.

[Indeed]

I suppose we have the defensive end of the 4 factors, except avoid fouling. Meanwhile, on offense, our weakest area is eFG%. It means we lack shooters or playmakers, and we lack both. Our isolation game plan doesn't do good on the 4 factors as well.

If we are based on these stats to build our team, I think we are far from there. We are pretty bad on both end. I think we lack a 3+D wing and a playmaker off the bench.

[firekil]

Now that The Kings parted ways with Oliver, does anyone see us make a push to hire him?

[vini_vidi_vici]

Fairly good question, I oft wonder how antiquated boxscore stats guys are (myself, a fan of some of them), with all the SportsVu/etc.. stuff out there?? I think Dean is incredibly influential in how the game has changed, I just think its evolving at such a quick rate. And as such, without privy to what hes looking at/working with, is he worth the hire??

When I saw what they did with that ghost data, my mind was blown. I still occasionally watch player movement on like say some opponants FGAs vs only the raptors, and watch it get torn apart, or even the ASTs (like Walls for instance) are telling.

[Morse Code]

Divac is a terrible exec. Why do all these great ballers refuse to admit the validity of advanced stats? They need to know when to stay in their lane a bit more. Being a great driver doesn't mean you know how to build cars.

[artsncrafts]

Dean Oliver was with the Nuggets during Ujiri's first time in Denver.

Hey man, TommyUDO (RIP) used to be a great frequent poster on our board, I appreciate reading the insights on the Nuggets, thanks. Hope you drop in more often.

While I appreciate the thoughts, I am no Tommyudo, I do lurk on several teams boards, especially since the Nuggets board has been so decimated since the Melo deal (way too many Nuggets posters got banned for posts on Knicks board) and just general apathy from the fan base due to continual 1st round exits and an owner that seems more interested in the bottom line then the are in winning the championship.i

The Nuggets are the Raps of the west so you are welcome to wallow in our sorrows.

[OAKLEY_2]

Divac is a terrible exec. Why do all these great ballers refuse to admit the validity of advanced stats? They need to know when to stay in their lane a bit more. Being a great driver doesn't mean you know how to build cars.

But it still requires driver training to know if a car or any cars have intrinsic value.

[vini_vidi_vici]

Divac is a terrible exec. Why do all these great ballers refuse to admit the validity of advanced stats? They need to know when to stay in their lane a bit more. Being a great driver doesn't mean you know how to build cars.

I initially thought he was rejecting it too. But apparently its just a Dean Oliver thing, and they are going to hire another guy.

[tweet]https://twitter.com/CarmichaelDave/status/625380620701536256[/tweet]

Funny reading the comments back, thye dont even know who he is.

[tweet]https://twitter.com/CarmichaelDave/status/625381198240452608[/tweet]
[tweet]https://twitter.com/CarmichaelDave/status/625381431083012096[/tweet]

[Morse Code]

Divac is a terrible exec. Why do all these great ballers refuse to admit the validity of advanced stats? They need to know when to stay in their lane a bit more. Being a great driver doesn't mean you know how to build cars.

But it still requires driver training to know if a car or any cars have intrinsic value.

To an extent yeah. Don't have to drive professionally to know though. It's a science just like most things.

Divac is a terrible exec. Why do all these great ballers refuse to admit the validity of advanced stats? They need to know when to stay in their lane a bit more. Being a great driver doesn't mean you know how to build cars.

I initially thought he was rejecting it too. But apparently its just a Dean Oliver thing, and they are going to hire another guy.

[tweet]https://twitter.com/CarmichaelDave/status/625380620701536256[/tweet]

Funny reading the comments back, thye dont even know who he is.

[tweet]https://twitter.com/CarmichaelDave/status/625381198240452608[/tweet]
[tweet]https://twitter.com/CarmichaelDave/status/625381431083012096[/tweet]

Fair enough. I just read the wiretap and ran with it. Nice to see he's not opposed to them in general. I wonder what aspect of Oliver's philosophy he didn't agree with or support.

[firekil]

Apparently he told Oliver not to go to Summer league and he went anyways. Not that egregious of an offence if you ask me.

[ILLuMiNAUGHTY]

Wow. 2 year bump. Impressive!

[cammac]

SAC is a train wreak from the owner the executive pool and coaching.
Yet they just might make a run for the 8th playoff spot they have talent .

[Latrell]

Apparently he told Oliver not to go to Summer league and he went anyways. Not that egregious of an offence if you ask me.

If that is indeed true , the Kings really are mess of epic proportions.

[Eating a Book]

[tweet]https://twitter.com/GetNBAUpdated/status/625325046328750080[/tweet]

[tweet]https://twitter.com/netw3rk/status/625379814615203840[/tweet]

[0 - 100]

[tweet]https://twitter.com/YourManDevine/status/625654679167332352[/tweet]

High profile former ESPN analytics guru, Dean Oliver who was hand picked by owner Vivek Ranadive, as Director of Player Personnel, and Head of Analytics a season ago, and who has multiple years left on his contract with the Kings - was told by new Kings VP Vlade Divac not to attend NBA Summer League.
Oliver was in attendance anyway and was seen with King coaches. Oliver was not permitted to be there in his official capacity with the team.
Divac has told confidants that he is strongly opposed to the use of analytics in evaluating players, and has halted all communication with Oliver.
Bill Herenda of CSN California and several others confirmed the report on Twitter.

The news comes as a surprise only insofar as Oliver initially seemed a terrific fit for the outlook of Kings owner Vivek Ranadive, who made his fortune as founder of the data-driven technology company TIBCO. Yet Oliver's method has clearly not held much sway as the Kings have added players like Rajon Rondo (now the statistical equivalent of an oversized "Do Not Enter" sign) and generally put together a team of ill-fitting parts (though that would be clear to most old-school minds, as well). Whether due to personality clashes, philosophical differences, or some other combination of factors, the Kings haven't looked to have had much use for Oliver and his approach. He had little sway, even in an organization with so many differing views.

It's likely that Oliver's role became smaller as Divac's increased in importance. If the Sportando report is correct, then Divac has no time for analytics (and his offseason moves suggest as much), then there is no point for Oliver to be involved with the Kings and this resolution makes a great deal of sense. For that matter, Divac appears to be in control of many of the team's decisions at this moment, from the decision not to trade Cousins to the signing of Rondo to the organization's reported refusal to hire George's son Coby Karl as an assistant coach. Additionally, Marc Stein of ESPN reports that Divac has extended an offer to former Kings and Serbia teammate Peja Stojakovic for a full-time front-office role. Given the duo's history, Stojakovic's inclusion in the organization would consolidate Divac's power.

Kings really are a mess. No direction, knee jerk decisions.

[OakleyDokely]

Another year in the lottery for the Kings.