Is Ennis the bet frosh in college basketball?

[jordoo]

Offensive and Defensive efficiency are highly correlated with Winning. Different player actions on the court effect OE and DE. By measuring their effect on OE/DE, and measuring OE/DE effect on winning, we can create a measure of how much a player's actions on the court contribute to winning. It is not a hodge podge of math, it is a simple logical extension to Wins/Losses using an intro level college stats class.

But its completely arbitrary really. Its a made up stat using other stats not something real and quantifiable so in other words a hodge podge of actual measurable stats used to create a new not real stat. It is not logical because Ennis didn't win 2.9 of our 16 games or whatever. We won 16 and he contributed to each one just like when we lose he would be part of that. I find using stats and math to make something up funny. Really stats do not tell the whole story anyways and one of the beautiful things about our team this year is that we are more than the sum of our parts and the guys seem to fit together well in a way that makes it easier on each other. This wins thing is taking measurable stats and trying to figure out which players are actually contributing the most to wins I get it, but its still silly to me because often times players are more efficient because of the better players around them that get more attention and wind up being less efficient. Don't these metrics rate efficiency quite heavily?

1st my understanding of bball is outdated.
2nd I am not able to understand intro level college stat classes.
What's next? I am attacking the particular statistic not you. If you want to value it go for it I find it to be silly and a waste of time because it tells us nothing new and tries to quantify something that isn't really quantifiable numbers wise IMO.

 

[Cusefannotindc]

I'm not really a huge fans of wins produced; as mentioned it loved Landry Fields a while ago Dwight Howard no doubt used to be probably the second or third best player in the league. But I also believe it likes Ronnie Brewer better on a per 48 minute basis than Melo, right?

Looking at Basketball-Reference, Fields's numbers are never better than average. NBA Geek only has the this season and last now for some reason. BR shows Melo with better numbers last year than Geek. Above average, but not super star numbers. In the model, Melo's scoring is inefficient, too many missed shots and he doesn't get a lot of rebounds.

 

[Cusefannotindc]

But its completely arbitrary really. Its a made up stat using other stats not something real and quantifiable so in other words a hodge podge of actual measurable stats used to create a new not real stat. It is not logical because Ennis didn't win 2.9 of our 16 games or whatever. We won 16 and he contributed to each one just like when we lose he would be part of that. I find using stats and math to make something up funny. Really stats do not tell the whole story anyways and one of the beautiful things about our team this year is that we are more than the sum of our parts and the guys seem to fit together well in a way that makes it easier on each other. This wins thing is taking measurable stats and trying to figure out which players are actually contributing the most to wins I get it, but its still silly to me because often times players are more efficient because of the better players around them that get more attention and wind up being less efficient. Don't these metrics rate efficiency quite heavily?

1st my understanding of bball is outdated.
2nd I am not able to understand intro level college stat classes.
What's next? I am attacking the particular statistic not you. If you want to value it go for it I find it to be silly and a waste of time because it tells us nothing new and tries to quantify something that isn't really quantifiable numbers wise IMO.

You really don't seem to understand this. It is not arbitrary, there is a wealth of data on teams wins/losses and player statistics which can easily be used to calculate quantitatively the effect had on W/Ls. The argument is not Ennis won 2.9 games, it is that his stats have been worth 2.9 wins to the team. Obviously players on the court matter as well, I suspect CJ being on the court, despite being less efficient scorer and turning over the ball often is demanding enough attention to improve the other players stats. It would be difficult to prove though because publicly available pbp data doesn't include line up changes and CJ already plays 90% of available minutes so there is not much time he isn't on the floor.

 

[jordoo]

You really don't seem to understand this. It is not arbitrary, there is a wealth of data on teams wins/losses and player statistics which can easily be used to calculate quantitatively the effect had on W/Ls. The argument is not Ennis won 2.9 games, it is that his stats have been worth 2.9 wins to the team. Obviously players on the court matter as well, I suspect CJ being on the court, despite being less efficient scorer and turning over the ball often is demanding enough attention to improve the other players stats. It would be difficult to prove though because publicly available pbp data doesn't include line up changes and CJ already plays 90% of available minutes so there is not much time he isn't on the floor.

Didn't you say wins produced is based mainly on individual stats? So you add up every ones individual stats and get a percentage of how many of the teams wins those stats have produced correct? If that's the case then I am correct its not an actual measurable stat. Guys could have there best games of the year in losses and those awesome stats contribute to their wins produced stat when they didn't win in those games or are the stats only collected from winning games. Stats are also not all equal for the roll a player plays on a specific team and do not tell the whole story they are great tools to help you understand what's going on but not the be all end all. I mean I could go on and on but basically no one has produced any wins by themselves with their stats. They have contributed to wins but not produced them.
Your explanation on CJ just shows how truly arbitrary this stat is because all things are not equal or provable by your own admission above. I don't need to understand how that stat works to know its not real when you have to combine a bunch of other stats and perform a mathematical equation to get your answer, no matter how simple that equation is. You've simply taken real stats (which can be great indicators but do nit tell the whole story) and come up with a formula that gives you a new made up stat trying to indicate the players over all value against said teams wins. Basketball isn't as simple as breaking it down into math IMO.

 

[strumpfasaurus]

Don't these stats kind of go out the window in a play off setting where the sample size is so small especially in an NCAA tournament format.

 

[jekelish]

Wins produced is not a metric I put much stock in. With baseball you can use WAR because while it's technically a team sport, everything in the game is based on individual performance and matchups. There's no team element except on something like a double play.

Basketball you can't really calculate it the same way. If anything, you could do +/- like they use in hockey but for the most part you can watch a game, if you know what little things to look for, and figure out who is contributing to wins or losses without any sort of equation.

EDIT: And by the way, there are so many things in a basketball game that can't be quantified or placed into any sort of formula because they aren't tracked and don't show up in a box score. What about setting screens, for instance? Or making a pass that leads to an assist? Or on a rebound, being a guy who tips the ball to keep it alive but doesn't get credit for a rebound? Or defensively, rotating over and cutting off a driving lane, or forcing a travel or a bad pass, but not getting credit for a steal there?

Or on the flip side, what about missing a rebound on defense that leads to an offensive put back? Or blowing a rotation that allows for a backdoor layup? The ways that you can impact a game that do NOT have stats are far more numerous than the ways you can impact the game in ways that DO have stats. It's impossible to create any sort of equation that sums up a player's value on a basketball court.

 

[jordoo]

They pretty much go out the window in team sports because they don't accurately reflect the affect the other players have on said players successes and failures. IMO

 

[Cusefannotindc]

Didn't you say wins produced is based mainly on individual stats? So you add up every ones individual stats and get a percentage of how many of the teams wins those stats have produced correct? If that's the case then I am correct its not an actual measurable stat. Guys could have there best games of the year in losses and those awesome stats contribute to their wins produced stat when they didn't win in those games or are the stats only collected from winning games. Stats are also not all equal for the roll a player plays on a specific team and do not tell the whole story they are great tools to help you understand what's going on but not the be all end all. I mean I could go on and on but basically no one has produced any wins by themselves with their stats. They have contributed to wins but not produced them.
Your explanation on CJ just shows how truly arbitrary this stat is because all things are not equal or provable by your own admission above. I don't need to understand how that stat works to know its not real when you have to combine a bunch of other stats and perform a mathematical equation to get your answer, no matter how simple that equation is. You've simply taken real stats (which can be great indicators but do nit tell the whole story) and come up with a formula that gives you a new made up stat trying to indicate the players over all value against said teams wins. Basketball isn't as simple as breaking it down into math IMO.

It is not tied to a team's win. It is based on how many wins should have been produced based on a player stats.

The idea that you can't combined multiple stats to create a simple measure of all them is an absurd statement. I assume you greatly dislike slugging percentage in baseball too then?

 

[jordoo]

It is not tied to a team's win. It is based on how many wins should have been produced based on a player stats.

The idea that you can't combined multiple stats to create a simple measure of all them is an absurd statement. I assume you greatly dislike slugging percentage in baseball too then?

Based on how many it should? Again I say that is silly. Come on this is arbitrary and the more you explain it the more it becomes so.

Simple measureable stats are not combinations of other stats combined into a formula making a stat that is measuring something that someone determines should produce x amount of wins.

Perhaps I'm too simple to understand but I call BS when I see BS.

 

[jekelish]

Seriously, can we stop pretending this is baseball? Of course based on the stats you should win more games if Miguel Cabrera is in the lineup, because baseball is entirely about individual matchups and the team aspect doesn't factor in. Basketball is not the same thing. Again, you cannot quantify the things that happen away from the ball that help the team win or cause them to lose. Therefore this is entirely moot.

As ImperialOrange pointed out when he talked about the first half in his post about rewatching it, Cooney ran UNC defenders, mostly Paige, ragged. So while he wasn't hitting, the benefits of the way he was moving away from the ball was not showing up in the stat sheets. He wore out Paige, who in turn had a crappy game. Cooney doesn't get any credit for that sort of thing in the box score, and therefore it doesn't factor into this nonsense "stat" that tells you who should win and when.

 

[Cusefannotindc]

Don't these stats kind of go out the window in a play off setting where the sample size is so small especially in an NCAA tournament format.

Since the NCAA is single elimination, it is much more likely the lesser of two teams will win one game than a series of games.

 

[Cusefannotindc]

Based on how many it should? Again I say that is silly. Come on this is arbitrary and the more you explain it the more it becomes so.

Simple measureable stats are not combinations of other stats combined into a formula making a stat that is measuring something that someone determines should produce x amount of wins.

Perhaps I'm too simple to understand but I call BS when I see BS.

You clearly are. There is nothing arbitrary about it, it is based on past data. People aren't just picking numbers out of air, they are looking at years of past data and using simple statistical tools to measure their effects and outcomes.

 

[Orangezoo]

Statistical models are far from an exact science. Considering in basketball there are very few constants every game efficiency models are a fun way to compare ratings but often are over valued due to the fact they correlate well with something as simple as wins and losses. If you were to take the top 100 teams in terms of something as simple as their record you could fairly easily predict the correct 80 pct who will go to the tournament. Maybe even 90 pct. For this shear fact you are almost guaranteed to find that efficiency ratings are a decent predictor. Wins and efficiency are closely related.

In many ways its like saying the humidity levels are high so that means it is going to rain. You can have some accuracy of predicting the weather just based on that but its not some fancy mathematical way of analyzing it. It is a hobbyist thing and nothing more.

The reason that efficiency models or any other models are not the sole way to analyze teams is due to matchups , type of defense etc etc. Two good def 3pt fg defenses could play polar opposite defenses. Those things matter when you look at matchups.

I am a much bigger believer in the eye test than efficiency ratings or models. And I have a degree in mathematics and deal with modeling and forecasting every day at work. Where efficiency and predictors help are in bubble team discussions and seeding weaker conference teams where the eye test against good competition is limited. Its a secondary tool and nothing more.

 

[Knicks411]

In Fields rookie year he ranked 17th in the NBA in WP/48. Among players who played 1,000 minutes, he was 12th.
If you get a lot off defensive rebounds and don't turn the ball over a lot, WP/48 seems to love you. So if you aren't very active on offense, you will probably have low TO numbers, but someone needs to create shots.

Someday, someone is going to come up with a great way to weigh usage vs efficiency

 

[Cusefannotindc]

Seriously, can we stop pretending this is baseball? Of course based on the stats you should win more games if Miguel Cabrera is in the lineup, because baseball is entirely about individual matchups and the team aspect doesn't factor in. Basketball is not the same thing. Again, you cannot quantify the things that happen away from the ball that help the team win or cause them to lose. Therefore this is entirely moot.

As ImperialOrange pointed out when he talked about the first half in his post about rewatching it, Cooney ran UNC defenders, mostly Paige, ragged. So while he wasn't hitting, the benefits of the way he was moving away from the ball was not showing up in the stat sheets. He wore out Paige, who in turn had a crappy game. Cooney doesn't get any credit for that sort of thing in the box score, and therefore it doesn't factor into this nonsense "stat" that tells you who should win and when.

Simply because we can't measure something perfectly doesn't mean we shouldn't. Basketball game is a high scoring game with few players on the court and many games, after Baseball it is the sport advanced statistics can best be applied to.

 

[jordoo]

You clearly are. There is nothing arbitrary about it, it is based on past data. People aren't just picking numbers out of air, they are looking at years of past data and using simple statistical tools to measure their effects and outcomes.

But they are doing it to make up a stat that doesn't exist nor measure anything real. Show me how its a real stat? How its provable that so and so produced x amount of wins. Let me ask you this if you add up everyone's WAR on a team does it equal the teams actual wins?

 

[Orangezoo]

Simply because we can't measure something perfectly doesn't mean we shouldn't. Basketball game is a high scoring game with few players on the court and many games, after Baseball it is the sport advanced statistics can best be applied to.

Its hobbyist and you validate that in this statement. It is fun to play with the numbers. And you can occassionally find some variables with strong correlations to outcomes. But it does not mean that those statistical models make good predictors.

 

[jordoo]

Simply because we can't measure something perfectly doesn't mean we shouldn't. Basketball game is a high scoring game with few players on the court and many games, after Baseball it is the sport advanced statistics can best be applied to.

So its not measurable as I've been saying. Thanks man. Taking actually measurables putting them into a formula no matter how simple it is and coming up with a stat that is not real nor perfectly measurable. So actual measurables + Math = something that's off and not actually accuarate. How surprising.

 

[Cusefannotindc]

But they are doing it to make up a stat that doesn't exist nor measure anything real. Show me how its a real stat? How its provable that so and so produced x amount of wins. Let me ask you this if you add up everyone's WAR on a team does it equal the teams actual wins?

Saying this isn't real is like saying your car insurance rate isn't real. You pay based on a formula that calculates how much it should cost to cover you.

I don't know of a site that calculates them all, I quick did the top 5, all of them have around 15 total WPs, Ohio St is closer to 17. But it is possible to win games while playing inefficiently, like SU did yesterday.

 

[Cusefannotindc]

Its hobbyist and you validate that in this statement. It is fun to play with the numbers. And you can occassionally find some variables with strong correlations to outcomes. But it does not mean that those statistical models make good predictors.

You should probably tell that to Daryl Morey and John Hollinger then. If you model you know you never have a r^2 of 1, all models operator at a level below 100% accuracy.

 

[Cusefannotindc]

So its not measurable as I've been saying. Thanks man. Taking actually measurables putting them into a formula no matter how simple it is and coming up with a stat that is not real nor perfectly measurable. So actual measurables + Math = something that's off and not actually accuarate. How surprising.

Why don't you come over to the thread where we have been picking against the spread and your eyes can match up against my math.

 

[jordoo]

Saying this isn't real is like saying your car insurance rate isn't real. You pay based on a formula that calculates how much it should cost to cover you.

I don't know of a site that calculates them all, I quick did the top 5, all of them have around 15 total WPs, Ohio St is closer to 17. But it is possible to win games while playing inefficiently, like SU did yesterday.

Care insurance come on dude of course I know how car insurance works this is sports prognosticating and its apples to oranges.

Again you are simply backing up my original argument that its pointless to go so far in bball stats and silly because after a certain point you could have learned more watching games.

 

[jekelish]

For your next trick are you going to design a javelin that takes advantage of Lamar's limp wristed throwing style?

Better yet, ask any basketball coach in America if they actually think any of these formulas matter. See how many think it's remotely useful.

 

[jordoo]

Why don't you come over to the thread where we have been picking against the spread and your eyes can match up against my math.

Spread is about creating action and making money not being accurate. If this stuff helps in betting than great but its forecasting and its simply again backing up my original post which you refuse to admit. Its not all numbers man just admit it and some stats are kind of silly just on the face of it.

 

[jordoo]

For your next trick are you going to design a javelin that takes advantage of Lamar's limp wristed throwing style?

Better yet, ask any basketball coach in America if they actually think any of these formulas matter. See how many think it's remotely useful.

LMAO