The NBA Offseason List - article forecasting Rox or the Spurs in the finals (charts/stats incl)

You can't really take those WP stats seriously when Batum or Hickson produced more wins than Lillard and Aldridge COMBINED!

I would be ashamed to post stats that defied logic.

 

I was hoping for a jcdenton graph.

 

You do need to look seriously at the formula for deriving WP. It does value defensive contributions much higher than most other measures. Thus Asik. Thus Noah. Lillard was a nonentity on defense. A concrete block set on the floor. LMA was not much better. There are two sides in basketball - offense and defense. Which is more important? Neither. If you are scoring 20 ppg but a stiff on defense how valuable are you? PER is much less PERfect than you think. for obvious reasons.

 

Wages of Wins fail the common sense test on too many things. And come up with things that just never come close to being right.

Rockets at a nominal 67 wins basically makes no sense. That would take a scenario where Harden plays like Prime Kobe, Howard plays like Orlando 2009~2010 version, DMo/Jones stepping up or trading for Aldridge/Love, and nearly no injuries and no letting up the gas to rest for the playoffs.

 

How do they take into account how good a player is on defense?

 

Did you read my post #18 in this thread??? They do publish their formula. It has its weaknesses. I tried to point them out.

No prediction of wins next season will ever factor in injuries. Injuries are asynchronous events that only the most advanced models take into account. I would never ever advocate for a trade involving LMA. He is about as overrated as it comes.

 

I explicitly listed their calculations in my previous post.

 

No you didn't.

 

http://wagesofwins.com/how-to-calculate-wins-produced/ again.

 

Actually, most predictions do take into account injuries. In fact, most statistical projections do this by taking account of injury history and age. But even then, that's my projection for 2013 IF everyone else have their normal injury risks. And every year at least one or two contender face some major injury issues.

In a world where injury is literally not a factor, I still won't see the Rockets win 67 games because the opposition would also be tougher than a typical NBA schedule. For example, likelihood of beating Cleveland in real NBA season should be much higher than beating a hypothetical Cleveland team with Irving/Bynum/Varejao all healthy?

 

It doesn't matter how he derived those values. If the end results are off, the formula is flawed.

I mean seriously...Do you actually believe Hickson produced more wins than Lillard and LMA combined??

You are so obsessed by LMA that you actually lost your perspective.

 

They defy conventional wisdom, but not logic. That's the point. How good are some of these individual Portland players when their team consistently has a losing record? Conversely, why do the really good teams highly value players like Thabo Sefolosha and Nick Collison, or Robert Horry back in the day, who do not accumulate traditional stats like points and assists? It's because some teams are better at evaluating players than others. The Wages of Wins guy is trying to properly evaluate players rather than relying on conventional wisdom. I find it incredible that so many Rockets fans fail to appreciate that their management team is on the cutting edge in this department.

That said, the Wages of Wins guy really only pays attention to the numbers. He had that Honeycutt guy on Houston's roster about a week ago until somebody corrected him.

Don't you ever wonder why Reggie Evans inevitably finds himself playing major minutes for good teams despite the perception that he's just some journeyman nobody? Or Matt Barnes? It's because these guys are good ballplayers! Morey feels like he's found guys like that in Ronnie Brewer and Omri Casspi, and the Wages of Wins guy agrees with him. Same deal with Isaiah Canaan.

 

I think you need to learn what the word explicitly means.

You didn't explicitly list their calculations like you said.

You posted a link.

Anyway, I assumed you read all about their calculations since you said you understand it so well, and I thought you could give a summary of how they calculate defense.

 

First of all, I love hearing us projected to be in the Finals!! Love it. But obviously there are sine big mistakes in there. The Win share is projected by age*minutes played etc but Lin is listed as playing only 16 mins per game? Then of course his win share is going to be low. But I'll take us reaching the Finals, lol!! Go Rockets!

 

If they do take into account the statistical likelihood of an injury I would love to see an example. I believe it only exists in the analysis of trades, because it is a asynchronous event, independent of most modeling techniques. (However it is model-able in statistical models of forests of data using perl optimized mathematical libraries or Matlab.)

Again, the WoW calculations for the best possible season include ... hell, read post # 18 again. I would recommend that you reread 4:

 

Should I think of you as my 5th grade students who need the information on the page presented, or should I think of you as as an educated adult who can delve down into the presented links? I may sum it up tomorrow if people have enough interest.

 

You said you understood this article very well and you were going to explain it.

Then I ask a question and you refer me to a link? If you were going to post links, then why even bother explaining anything at all?

I could have just read through everything myself. But since you said you have already read everything and understand it, I thought you could give a simple answer to what they take into account when calculating defense without me having to go back and read a long article about how they constructed their statistical model.

 

You are right. I do apologize. However the entire process for defense is to complex to completely detail, if that means explaining the equations and statistical adjustments made.

One aspect of defense as is modeled by WoW is based on team defense. The team defense is then allocated across its individual players based on minutes. However the adjustment to a players individual WP is small.

Player defense is however weighted heavily. The variables include steals, defensive rebounds, blocks, opponents FG%, opponents 3 point FG%, fouls, and opponents turnovers. These are individually weighted depending on a statistical analysis of the last 25 years or so of all available NBA data.

This system is less accurate than some other systems that use 5 man rotations and the individual contributions to attempt to extract more exact data for the individual players. The more exact evaluation is known as Drtg and may be searched for on the internet.and can

 

It depends on what you are using to value a player. If you're using a stat like PER to determine player value, or any other common method (pts/gm, how flashy a players crossover is) then Lillard and LMA look great. WP, however, does not reward inefficient scoring as PER does. Lillard and LMA were both mediocre in their shooting efficiencies. Therefore, the team would be better off if players with higher shooting percentages were taking those shots. You can score 20 a night all day, but if you do so inefficiently WP will laugh at you.

In the case of Lillard, he's a rookie, and I would expect him to get better with experience. LMA however is so incredibly overrated it hurts. Hickson on the other hand is a very effective rebounder and efficient scorer.

Read these articles from the WoW guys that will explain their perspective a bit.

http://wagesofwins.com/2012/03/04/wayne-winston-simplifies-pers/
http://wagesofwins.com/2013/08/23/the-uselessness-of-usage/

 

Wages of Sin is to Death as Wages of Wins is to ______________.