5 NBA Stats That Lie and How You Can Force Them To Tell the Truth

LeBron James is statistically one of the best basketball players we've ever seen. Does that mean he's one of the best basketball players we've ever seen? It all depends on how you feel about stats.

Some feel that stats are the most pure way to judge the value of a player. They argue that basketball is a team sport, and that players shouldn't be judged based on what their teammates do. The best way to judge individual performance is by stats. These are the "stats don't lie" crowd. 

Others feel that the ultimate goal is to win, so if a player's play isn't helping the team to win, then all the stats in the world are meaningless. These are the "stats lie" crowd. 

The truth is that stats tell us exactly what they are designed to tell us, nothing more and nothing less. That doesn't mean they are wrong. It just means they are incomplete. 

The problem isn't what they do tell us; it's what they don't tell us. It's not the stats that do exist; it's the stats that don't exist. I don't suggest throwing out the following five stats, but that new stats are tracked which, combined with the existing stats, will give us a fuller version of "the truth." 

Blocks

It may surprise you to know that Dwight Howard did not have the most blocked shots last season at 198. That distinction goes to Serge Ibaka. The most blocks per game came from Andrew Bogut with 2.6. What does that really mean though?

There are several things that you can look at. Ibaka blocked 2.09 percent of the Thunder's opponents shots. Dwight Howard blocked 2.83 percent of the Magic's opponents shots. Of course that's measuring their total shots, and not just what happens when they're on the court. 

On the other hand, Howard spends more time on the court than Ibaka, having played more than 800 minutes more. 

That's why Basketball-Reference's "block percentage" which shows the percentage of opponents' attempts blocked while the player is on the court is the best estimate of blocking ability. In that category, it's neither Ibaka nor Howard who leads the NBA. 

Javale McGee blocked 6.7 percent of all shots while he was on the court, the best in the NBA. 

There's still one thing that is missing from the equation though, which is what happens after the shot is blocked. Some players have a tendency to block shots out of bounds. Those don't result in a change of possession. 

Others will block the shot to a teammate or even in Woody Paiges words, "bleal" the ball, block plus steal, meaning they just catch the shot attempt. 

The missing stat here is blocks resulting in change of possession per 100 possessions. This would give the most accurate measure of not only the frequency of blocks, but also the result of the blocks. 

Steals

Chris Paul led the league in both total steals and steals per game. Does that make him the game's best thief? How much is really in a steal? Rajon Rondo is second. Both Rondo and Paul made the NBA's All-Defense team. 

Monta Ellis is third and Russell Westbrook is fourth. Are they all on the same page defensively? 

It's almost impossible to tell anything by just the number of steals per game, or for that matter, even steal percentage, where the top four remain the same, albeit in a slightly different order. 

Steals can be at the same time the best and most dangerous defensive play in basketball. 

It can be the best because it can lead to transition points, easily the most efficient offense in the game. On the other hand, if you miss on the steal attempt it can expose your team to points scored by the other team. 

Hypothetically, say there is a player who goes for five steal attempts in a game. Twice he gets the steal and those steals lead to field goals in the other direction. The other three times, he misses and gets burned for six points. 

Two steals look like "good defense" because they don't show the missed steal attempts and the cost of missing. 

That's why the best way to judge steals is "net points off of steal attempts." In this case it would be a minus-two. 

The net points would give the full story, both the good and the bad. 

Rebounds

Kevin Love had more rebounds than anyone in the league last year and he had the most rebounds per game last year. Does that mean he's the league's best rebounder?

There's a few things to consider. First, consider that there were 7,596 missed field-goal attempts in Timberwolves games this year. Compare that with 7,263 for Chicago. That's a considerable difference. Portland had only 6,962 missed field-goal attempts this year. 

There were more rebound opportunities in Minnesota games than any team in the NBA. Does that account for Love's huge rebound rate? Well, there's rebound percentage which shows the percent of rebounds a player grabs which are available. 

Among players who played at least 1,500 minutes Marcus Camby grabbed the highest percentage of rebounds while on the court at 24.1 percent. Love is second at 23.6 percent. That's better but there's still some issues to deal with. 

Consider San Antonio or Chicago compared to Minnesota. Joakim Noah leads Chicago with 18.5 percent rebound percentage. Carlos Boozer is second on the team with 17.6 percent. On San Antonio, DeJuan Blair led the team with a rebound percentage of 18.9 percent and Tim Duncan with 18.3 percent. 

So not only does have Love have more opportunities, he also has no one on his team to compete with for rebounds. 

Someone smarter than me needs to figure out some way of normalizing the team aspect of rebounding. When Blair and Duncan or Boozer and Noah compete for the same rebound, their numbers are reduced. For instance, consider that while Boozer was out at the beginning of the year, Noah averaged 15.0 rebounds per game. 

One way of figuring would be to take the team's rebound rate and subtract the player's rebound rate to determine what the teammates' rebound rate is. Then using that result to figure the total rebounds a player grabs after accounting for the team. 

In other words, for Joakim Noah, he grabbed 18.5 rebounds of his team's 52.1 while he was on the court. That means his teammates grabbed 33.6 rebounds and the opposition grabbed 47.9 rebounds, or Noah grabbed 18.5 of the 66.4 rebounds available after accounting for teammates, or 27.8 percent.

Kevin Love got 23.6 of 71 rebounds available to him after adjusting for teammates, leaving him with a 32.2 percent rebound rate. 

Marcus Camby got 24.5 of the 72.1 rebounds available to him for a percentage of 33.9 percent of the available rebounds. 

While this helps to a degree to mitigate the difference in teammates, it doesn't account for rebounds that a player may have gotten if the teammate were not on the court at the time. 

Therefore a stat that might be helpful is "Deferred Rebounds" or a count of rebounds a player could have had but deferred to a teammate who also had good position. While players like Love shouldn't be penalized for getting rebounds, players that share the court with exceptional teammates should be accounted for as well. 

Assists

Assists are very difficult to put a finger on. An assist can mean almost everything or almost nothing but since they are all treated the same, it's hard to tell what they mean.

In essence, there are three kinds of assists. There's the assist where a player merely makes an easy pass to a player wide open for a shot. There's the assist where a player makes a difficult pass to an open player for a shot, such as a crosscourt pass or a jump pass to a player who is open and makes a shot.

In both of those cases, while the latter is certainly more difficult than the former, it is still dependent on the ability of the player to create a shot for himself off the ball.

The third type of assist is a kick assist, where the player breaks down a defense and kicks the ball to a player that was made open by the passer's drive. In such a case it is the ball-handler and passer who created the shot. The shooter merely had to stand around and shoot when the ball arrived.

The problem with assists is that all three are treated the same. A player like Rajon Rondo will have all three kinds of assists but there's no distinction between one type and another. Derrick Rose and Chris Paul are exceptional at breaking down defenses. Deron Williams has exceptional court vision.

The problem is that just "assists" doesn't really tell the whole story. It treats every type of assist the same. Dumping off a simple pass isn't the same thing as creating offense for your teammates. They shouldn't be treated as the same thing. 

There should be three categories: Easy Assists, Difficult Assists and Created Assists. Honestly it's hard to give a hard definition of the difference between them, but you know it when you see it. 

Points

Kevin Durant is the back-to-back scoring champ, but does that make him the best scorer in the NBA?

Strangely, scoring might be so deceptive that it actually has a strange kind of honesty to it. In one sense yes, there are "volume shooters" who simply shoot a lot and generate scoring simply by the sheer number of shots they take.

While there is a point where "overshooting" can be an issue, determining exactly where that point is can often be a problem. For instance, take Chris Paul and Derrick Rose. Paul runs the offense in the classic way that a point guard is supposed to run it and is technically a far more efficient scorer than Derrick Rose.

However, Rose's Bulls score more efficiently than Paul's Hornets. This is in large part because of the tremendous defensive attention that Rose draws to himself. As a result, Rose's teammates have a seven percent bump in field-goal percentage when he passes out of isolation.

So a certain kind of paradox unfolds. The more Rose shoots, the better his teammates get. This same phenomenon is probably present in most volume shooters as the logic isn't likely to change. Double-teams and triple-teams mean open players somewhere else. The trick is finding the balance between scoring and passing.

Therefore, volume shooting can actually mean a more efficient offense.

Conversely, sometimes people will say that a certain player has a higher field-goal percentage and then assume that if that player were to take 25 shots per game he would maintain the same field-goal percentage.

This is contrary to a known observation called a "skill curve" which stipulates that the higher a player's usage goes the lower his offensive rating goes. That's a direct result of the same logic. As a player's role on offense grows, the defensive attention he gets grows and that in turn causes his efficiency to come down.

That's why I say that there's a strange sort of honesty to points. Players that score more points do so because they can.

As I said earlier though, there is a point of "overshooting" where a player simply is shooting too much. The point at which the benefit that comes from his teammates being open isn't enough to compensate for the decline in efficiency by the increase in usage.

Finding out precisely what that point is is a difficult thing to determine and there's a considerable amount of debate about it. The thing is, it's not the same for every player.

Another factor to consider is where the points come from. Some, like John Hollinger, have made "long twos" sound like a dirty word. They think that every shot needs to either come at the rim, where it's a higher-percentage shot, or from behind the arc. However, we've seen what happens to a team that overcommits to that kind of strategy.  

Players like Kobe Bryant, who hit 51 percent last year from 10-15 feet, are highly prized for a reason—they can score efficiently in that area between the paint and the arc. Yes, he shot only 38 percent form 16-23 feet, but that still leaves about a 42 percent rate in the "long-two range."

Once again though, it's not just points he scores that are important but the attention he commands in scoring them. Players who can score in the long-two range can break down defenses.

In some ways, points can be overstated and lie in that sense, but in many ways, even the lies can lie. Sometimes points can say too much and sometimes not enough. 

Perhaps one stat that could be tracked is "Forced Shot Attempts." These would simply be counted when a player forces up a bad shot when a teammate is open and time is left on the clock.

There could also be "Rushed Attempts" which would count when a player receives the ball at the end of a shot clock or game clock with less than five seconds to shoot. 

In those cases, rushing the good basketball play is to shoot, but it's bad for your stats. You could find a better indicator of a player's actual scoring ability by only looking at his non-rushed attempts. 

With all of these stats, I don't mean to imply that the stats aren't important. They are. I think that there's a story to tell and there's the part that the major stats tell and there's the part that they don't tell. It would be foolish to ignore either part.