Article Draft: High-Tech Hoops Stats for the Casual Fan – VOL 2 – PER

Discussion in 'Chicago Bulls' started by transplant, Feb 5, 2011.

  1. transplant

    transplant Global Moderator Staff Member Global Moderator

    Joined:
    Jan 6, 2011
    Messages:
    4,111
    Likes Received:
    210
    Trophy Points:
    63
    For those who hated the first in this series (TS%), damned if I didn't go draft another. While it's meant to be fairly simple and entertaining as opposed to scholarly, I don't want any factual errors if i can help it. I plan on publishing the finished product early next week. I'd appreciate your input.
    **************
    [​IMG]

    When it comes to advanced measurement of the performance of a basketball player, the Player Efficiency Rating, or PER, is the biggest, baddest, 600-pound gorilla of a stat you can find. I mean, if basketball stats were pizzas, this thing would have absolutely everything on it, including the kitchen sink and other sundry fixtures.

    The PER stat was developed by ESPN’s John Hollinger in the late 1990s. According to Hollinger, “PER sums up all a player's positive accomplishments, subtracts the negative accomplishments, and returns a per-minute rating of a player's performance.” The positives accomplishments included in the formula are points scored, offensive rebounds, defensive rebounds, assists, blocked shots and steals. Negatives include, missed field goals, missed free throws, turnovers and fouls committed. If there’s anything else you can think of that folks keep track of during a basketball game that I didn’t mention, I’m sure it’s my oversight, not Hollinger’s. Various weights are assigned to each positive and negative and then the raw score is adjusted for minutes played. Because of this adjustment, PER has the advantage of being useful in comparing the relative productivity of big-minute, small-minute players and everyone in between.

    The formula? Oh, I suppose I could copy and paste it here, but last time I looked at it my head started hurting so I think I’ll spare you the experience. However, for the truly masochistic, here’s a link:

    http://www.basketball-reference.com/about/per.html

    Just remember I warned you.

    In the first article of this series on True Shooting Percentage (TS%), I showed the current league leaders in TS%. It wasn’t a very impressive a list. My guess is that you'll like PER's leaderboard better:

    1. LeBron James-MIA 27.0
    2. Chris Paul-NOH 25.9
    3. Dwyane Wade-MIA 25.3
    4. Dwight Howard-ORL 25.3
    5. Kobe Bryant-LAL 24.9

    These may not be your personal top-5 players in the NBA, but if you watch many games, it’s probably not far off. As I see it, any performance measure that comes this close to being in sync with what my eyes tell me merits my respect if not a standing ovation.

    Another interesting feature of the PER stat is that it adjusts itself so that a score of 15.0 is the NBA average. I mean, how cool is that? Just please, don’t ask me how it does it, because that also makes my head hurt.

    Mr. Hollinger, you done good.

    OK, enough of the hearts and flowers…time to take our cheap shots.

    Measuring Defensive Performance

    Some of the greatest statistical minds on the planet have attempted to quantify an individual player’s defensive impact on the game, and for my money, they’ve all been found wanting. Hollinger included (and in fairness, Hollinger agrees). I’m sure that my experience coaching the game colors my opinions on this, but playing good defense is about a lot more than defensive rebounds, steals and blocked shots. In fact, when it comes to steals and blocks, players who emphasize these in the way they play the game often trigger their coaches’ latent homicidal tendencies. This is due to the fact that missed steal and block attempts take the player out of position, frequently resulting in easy baskets for the opposition. As I see it, if PER (or any other “all-inclusive” performance stat) doesn’t negatively recognize when these defensive stat-hounds leave their teammates in the lurch, they ought to remove steals and blocks from the formula altogether.

    But there’s more to it than that. Good defense is doing whatever you can to prevent the other team from scoring, including all the little things a defender can do to make it difficult for opponents to “run their stuff” - a hand that’s up, not down, a bump to take a cutting opponent off stride, a well-timed double-team or weakside help…the list goes on and on. The obvious problem is that much of this is subjective and cumulative in their effect. This is why I don’t think there will ever be a measurement of individual defensive performance that I’ll be able to fully embrace.

    A few more things about PER that make me go “hmmm”

    • As mentioned earlier, each of the positive and negative on-floor accomplishments are given their own weight. As nearly as I can tell, a blocked shot has the same positive weight as an offensive rebound. I’m sure Hollinger had his reasons, but as I see it, an offensive rebound always results in an extra possession while a blocked shot doesn’t necessarily result in a change of possession. I mean, you could have 5 blocked shots in a single possession and your opponent could still score on that trip. Sorry, I don’t get it.

    • I’m not too keen on penalizing a player for fouls committed since we all know that not all fouls are created equal. There are good fouls and there are bad fouls. There are fouls the head coach specifically tells a player to commit. Heck, if Shaquille O’Neal has the ball under the basket and you don’t foul him, you’re an idiot. Opportunities to win games have disintegrated because a foul wasn’t committed quickly enough.

    • As I see it, if you want to penalize for fouls committed, limit it to fouls that result in an “And 1” for the shooter and any time a player fouls a jump shooter…those are virtually always bad fouls. Oh, and fouling a shooter on a 3-point attempt should immediately drop your PER to zero.

    • The inherent beauty of seeing a player hustle into position and take a charge is completely overlooked by PER. Oh wait…that’s not quite true. The player committing the charge is penalized by both the foul and the fact that it’s a turnover...but the poor guy lying crumpled on the court who actually made the play gets nothing, nada, zippo. Mr. Hollinger, that just ain’t right.

    Of course the worst thing about the PER stat is how some people use it. In particular those who, in any discussion relating to the comparative abilities of two players, whip out the PERs and lay ‘em on the table like they’re revealing a royal flush. Hollinger doesn’t even see it that way, yet you’ll run into some insufferable know-it-alls who do...and they also make my head hurt.

    If you believe that PER is all you need to know to assess the quality of a player then you believe that:

    • Golden State’s Stephen Curry is a better point guard than the Celtics’ Rajon Rondo.

    • There are 57 shooting guards you’d rather have on your team than Oklahoma City’s Thabo Sefolosha, a NBA second-team all-defensive team selection last season.

    • Chicago small forward Luol Deng is just barely an average NBA player (PER 15.3) and is only the 14th best small forward in the league.

    • Orlando power forward Ryan Anderson outshines the Celtics’ Kevin Garnett, New Orleans’ David West, the Lakers’ Lamar Odom, Chicago's Carlos Boozer, Portland’s LaMarcus Aldridge and Miami’s Chris Bosh at his position.

    • That this season the Milwaukee Bucks would be better off with the Mavericks' Tyson Chandler or Nuggets Nene Hilario at center than their own Andrew Bogut.

    And it’s perfectly OK to believe these things…you’d just be wrong.

    Hollinger’s Player Efficiency Rating is, in my opinion, a magnificent statistic and a credit to its creator. It’s just not perfect…none of the advanced hoops statistics are. So if you run into one of those people who offers up PER as a case-closed-discussion-over trump card, don’t be afraid to give your antagonist a steely glance and say in your best Clint Eastwood voice, “Very nice…now what else ya got.”

    Tom Nossem
    Sources: basketball-reference.com, espn.com
     
    Last edited: Feb 6, 2011
  2. JayJohnstone

    JayJohnstone Active Member

    Joined:
    Feb 15, 2008
    Messages:
    1,032
    Likes Received:
    3
    Trophy Points:
    38
    Good stuff, Transplant. I never understood the fouls piece of it either. Also, in Hollinger's defense, I don't believe "Charges drawn" are an official NBA stat.
     
  3. transplant

    transplant Global Moderator Staff Member Global Moderator

    Joined:
    Jan 6, 2011
    Messages:
    4,111
    Likes Received:
    210
    Trophy Points:
    63
    They're not, but you can be damn sure teams keep them...and thanks.
     
  4. Denny Crane

    Denny Crane It's not even loaded! Staff Member Administrator

    Joined:
    May 24, 2007
    Messages:
    72,957
    Likes Received:
    10,620
    Trophy Points:
    113
    Occupation:
    Never lost a case
    Location:
    Boston Legal
    PER is certainly not the end all of any discussion, but it is the "start all" of many good ones.

    While it's true that PER doesn't give the complete picture of a player and his contributions because it doesn't consider defense, you can certainly look at a guy with a 25 PER and he's guaranteed to be a top 10 type player in the league. As pointed out in the article. The flipside is true, too. You can look at a guy with a PER of 9 and be certain he's not a top 10 player - in fact he's not likely to be a very good player at all.

    For grins, I looked at Ben Wallace's advanced stats at basketball-reference. He had 5 seasons in a row with PER over 17, peaked at 18.6. He's obviously not a great offensive player by any stretch, but somehow PER captured at least some aspect of his contribution to those Detroit teams that consistently made the ECF or better.

    On the other hand, Bruce Bowen was a 1st team or 2nd team all nba defense player with a career best PER of 9.5.

    I would make the case that for several of his seasons, Ben Wallace was a top 10 type player in spite of his low PER (compared to Kobe's 24.9) because you could see he played defense at a level unlike any we've seen, perhaps ever.

    I wouldn't make the case for Bowen, though he was starting SG for the Spurs who won three titles between 2000-2001 and when he retired in 2008-2009. Bowen's defense alone likely would not have got him the starting job - he was a 3pt specialist who hit .393 for his career and over .400 for most of those seasons in San Antonio.

    Digging only a little deeper, ws/48 tells this story. Bowen twice was better than .100, his best at .111. Wallace was consistently at .150 or better.
     

Share This Page