Statistical Analysis. Humor. Knicks.

Friday, December 19, 2014

Henry Abbott & The Art of Automobile Maintenance

I love sport lists. Get 10 NBA fans together and ask them who the 10 best NBA players of all time are, and you’re likely to get 10 different lists. Even getting a consensus on the best NBA player of all time proves to be difficult. Many will point to Russell’s rings, and just as many will claim Wilt was a man among boys. Some might say Jordan was clutch, while others might argue the “Big O” was the most versatile.

Lists tend to reveal a lot about the person making the list. You may have Russell at the top of your list if you think winning a team championship is the best measure of an individual. On the other hand if you think that winning a championship is more a team effort and doesn’t adequately reflect a single person’s accomplishments, then Wilt might be your guy. If you feel that today’s athletes are far superior and face tougher competition than those of yesteryear, then Jordan would be #1. While lists are subjective, it’s not as open as choosing your favorite ice cream flavor. While “pistachio” would be an acceptable answer at your local ice cream shop, saying that Bill Cartwright was the greatest NBA player of all time is just wrong.

Recently ESPN asked their writers to rate the top 10 centers of all time. Henry Abbott of truehoop.com, and ESPN newbie, filled out his form and included Bill Walton & Dave Cowens, but omitted Moses Malone. Bill Simmons made a short blurb about Malone’s exclusion in one of his columns, and Abbott felt the need to explain his reasoning at truehoop. While there might be valid reasons for ranking Moses 11th or greater, I think Abbott’s position is a bit odd.

If he had been 6-9 he probably would have made it. I’m into overachievers. But Malone’s awfully big. And strong. Not fair, but true, I’m afraid.

Three different sites list Malone at 6-10, but I don’t want to split hairs over an inch. Abbott’s point is that he would have viewed Malone’s accomplishments more favorably if he were a smaller player. But I have to ask: given all other things being equal, how would being shorter made Moses Malone a better player? I just don’t get the argument there, because it leads down a slippery but not very steep slope where smaller players get more credit for their achievement.

If Henry’s discussion ended there, you wouldn’t be reading about it here. But he continues onward.

I’ll tell you this much, though: you’re not going to convince me just with stats. I play basketball, or something close to it. I never thought being a winner was necessarily about getting the most points and rebounds. It’s about building a team.

I’d like to state the obvious and say there’s a strong correlation between getting the most points and winning. But seriously, what is the purpose of bringing out the “I play basketball” card? Is Abbott suggesting that there is a division between those that can ball and those that can multiply fractions? Or rather that only those who play basketball are qualified to understand what makes a winning team? I play basketball too. So does Dean Oliver, and I’m sure there are tons of people in the statistical community that can lace them up. Nonetheless you don’t need to be a great basketball player to understand what being a winner is about. Isiah Thomas and Kevin McHale were much better players than Jerry Colangelo, but who would you rather have building your team?

Abbott continues his thoughts on statistics:

But for now, I’m convinced that points and rebounds, as freestanding indicators of a player’s quality, are a total crapshoot. If Eddy Curry were a total ballhog, he’d shoot every time he touched it, and no doubt score more. But of course he’d really hurt his team in the process. No way to factor that in when we put points and rebounds on the altar as sacred stats.

I can see four major problems in Henry’s example. The first is that Curry is a notoriously poor rebounder. So by using points and rebounds as an indicator of Curry’s quality you’d probably get a good understanding of his value. In other words if you were in a coma for the last 8 years & I gave you a newspaper with Curry’s stats, you’d probably have a fair idea of what Eddy brings to the court.

Second is Abbott’s assumption that Curry could score much more if he were a ballhog. There is no doubt that Curry can score one on one against just about any center in the league. However, like any NBA player, Curry can’t score with two defenders draped on him. And unfortunately Curry is unable to find the open man when double teamed. Hence opposing teams find it a low risk move to double team Curry, and by using this tactic they can limit how many points he can score. If Eddy Curry were able to find his teammates for an easy bucket, then opponents wouldn’t be able to double team him as much, and therefore Curry would be able to increase his scoring averages. So Curry’s scoring average isn’t predicated on some kind of statistical altering greed, but rather it’s limited by his poor passing ability.

Third Henry assumes that if Curry scored more it would hurt his team. But Curry’s primary function is scoring. Other than grabbing offensive rebounds, Curry doesn’t do much else well. If Eddy Curry were able to drop 29 points a night instead of 19, it would benefit the Knicks. The Knicks are trying their hardest to get the ball to Eddy more, not less, in an effort to increase his usefulness to the team.

Finally Henry asserts that should Curry hurt his team by going for personal glory, that there is “no way to factor that in” with stats. O RLY? Should Curry become a “total ballhog” you’d see a steep rise in his turnovers per minute, and his PER would plummet. To see how this affects the team you could look at the team’s offensive efficiency, or you could go to 82games.com and check the offense’s +/- with Eddy on the court. So in fact, there are many ways to statistically “factor in” a player that is in over his head.

It seems to me that Henry Abbott’s main gripe is that statistics isn’t the panacea of NBA analysis. That is you can’t take a single formula & use it to find precise answers as to the net value of a player. In some cases statistics do a poor job of capturing a player’s worth. Guys like Bruce Bowen, Quinton Ross, and Raja Bell aren’t adequately represented by their statistics. But should we just discard all statistics because it gives a few players the short end of the stick? That would be like eliminating capitalism because of the poor. Statistics bring such a surfeit of unbiased data that we can live with their deficiencies.

Abbott frequently uses the term “crude” when describing statistics (“using one players’ individual’s points and rebounds as a major tool in that debate is like using a shovel as a major tool for brain surgery: so crude it hurts.”). However, I find statistics to be an accurate and elegant way to communicate information. Take for example my assessment of Eddy Curry. Curry is a highly efficient scorer (19.3 ppg, 58% FG%) who doesn’t rebound well (7.1 reb/g) especially on the defensive end (4.6 dreb/g). He isn’t among the league leaders in scoring because he doesn’t pass well (0.9 ast/g, 3.4 to/g). The Knicks aren’t doing well mainly because of their defense (26th defensive efficiency) and some of the blame points to Curry (0.6 blk/g, defense 5.7 points worse with Curry on the court). Without the numbers to back it up, my view of Curry might be skewed by my allegiances.

Finally, Henry uses this analogy:

Points and rebounds are only ubiquitous because they are so simple to measure. Any idiot with a clipboard can chart that.

Similarly, it’s really easy to tell if your car’s headlights are working. But that’s a bad bit of investigation if you want to figure out if you’re going to make it safely cross-country. For that you gotta pop the hood and get your hands dirty.

I have to agree with him on this, but I feel as if Abbott missed his own advice. Moses Malone has more points, rebounds, free throws, MVP awards, and All Star appearances than Walton & Cowens combined. The only advantage I can see that Walton & Cowens have over Malone is that they’ve won more titles. I can’t say for sure what Abbott’s exact criteria was (rings?, desire?, height?) but by ignoring the wealth of statistical information available, he certainly didn’t pop that hood open and get his hands dirty.

27 comments on “Henry Abbott & The Art of Automobile Maintenance

  1. STEVE

    This debate always returns to a major point we seem to always forget…basketball is a team game. The greatest individual players have to blend in with a team concept to win anything. Malone lifted bad teams with his magnificent play.
    When he went to a contending team in 1983, they had one of the best seasons in NBA history. Kareem is my man, but I attended the 83 finals at the Forum and Malone crushed Kareem like a grape.
    The Sixers had no answer for ABdul-Jabbar in 80 or 82.
    The second importiant thing to me is the competition at that players’ position in the era. For my dime Kareem still gets number one spot, because he played against Hall of Fame competition night in and night out in a 23 year Career. Kareem faced more double and triple teams than any center living or dead. Every game plan was to stop him first. I would take Malone over Walton and Cowens based on the brute force that he played with, the relentless attack on the offensive and defensive glass. The others may have been more complete players, but they would have needed help to check Moses.

  2. Brian Cronin

    Hey, does anyone else read firejoemorgan.com?

    Great site that does basically the same thing KB does here (which was well done, KB), go through poorly-thought-out sportswriting and picks them apart.

  3. Henry Abbott

    It’s getting a little old at this point, but just for the record, I totally agree that Moses Malone belongs in lots of people’s top tens. Put him first overall if you like in your list. I respect that.

    To be honest, I DON’T EVEN UNDERSTAND THE CRITERIA. They were not specified. Walton in 1977 was incredible. Most of his career he was out injured. Do we count that? How about Arvydas Sabonis’ Soviet years? Who knows? It’s not specified.

    Through the tangle of arguments, I feel there are some centers who are inarguably superior. In no particular order: Bill Russell, Wilt Chamberlain, Kareem Abdul-Jabbar, Shaquille O’Neal, and Hakeem Olajuwon. I add Walton to the list, because I arbitrarily decided to consider him at his peak, which is wholly unscientific, but so is this whole debate.

    After that, every single player has flaws. Pick whoever you want out of the ten or fifteen next best. Perhaps I should have left those blank. But in the spirit of the thing, I selected the ones who seemed most valuable based on my understanding of the game–and, admittedly this is a crap shoot, which at least I admit. (You think you have a scientific basis for ranking ten different players from ten different eras? I’d like to see it.)

    Artis Gilmore made it because I have talked to people who played against him, and they hated playing against him. Period. Made an impression on me. Robert Parish won a lot, didn’t make mistakes, and perfected the art of making his teammates look good. Without George Mikan there probably wouldn’t be an NBA.

    Moses Malone did a lot of incredible stuff, and if the criteria were to pick the top ten centers for a single playoff run, I’d put him in the top seven. But to me, there’s a message from this whole list about winning and teamwork. Malone played on NINE teams, and only won a title when he was on, essentially, an All-Star team. This was not Mr. Team Building Title Winner.

    Can I tell you, categorically, that these ten are better than those six or seven? Am I surprised to find convincing arguments for people who were not in my top ten? On both counts, not for one second. That’s inevitable (and, incidentally, why people make lists like this: to force those conversations and the buzz that comes with them). And no one can. But I can tell you that I put a lot of thought into it, and I stand by it.

    Interestingly, one of the voters with actual NBA head coaching experience, Jim O’Brien, made very similar picks to mine, and also left Malone out of his top ten. Haven’t heard anyone questioning his judgment.

  4. Kevin Pelton

    KB, I feel like you missed the mark here.

    If we’re talking points per game and rebounds per game — and that’s the impression Henry gives — then I do think they are crude. I don’t think they come close to telling the whole story of anything.

    Henry has been an advocate for advanced analysis, not a foe of it. I think he puts it well that the question is not “stats or no stats” as much as it is “stats that reflect context or traditional stats.”

    Yeah, the notion that Malone would be ranked higher if he was 6-9 (or that he was a stat padder, for lack of a better term) is silly, but ultimately, ranking players is a silly exercise – right, KD?

  5. Michael Zannettis

    Brian: great shout out to firejoemorgan.com It’s an absolutely wonderfully hilarious/insightful blog that takes ‘edutainment’ to new heights (or, if you’re not into snide humor, new lows.)

    Henry: Not to speak for Mr. Knickerblogger here, per se, but I believe the essence of his article is that we should be striving for accurate measurements, and that interpreting the right statistics are an incredibly valuable tool in determining the value of a basketball player.

    What I feel is often lost in the statistical community criticism is how much basketball stat heads invest to TEAM stats. We have +/-, offensive and defensive efficiency, and even opponent PER by position, if you want to go there. We know A LOT about how a certain player affects his teammates.

    It’s BECAUSE of TEAM statistics that we know that Shane Battier is an incredibly valuable defensive player. We don’t just look at his steals and blocks, we see his +/- and can conclude that, ‘Damn, that man knows how to cut angles from the weak side.’ Using statistics gives us a deeper appreciation for glue guys, who make their teams better.

    And if I was a GM and wanted a glue guy to make some passes, play some defense, and knock down an open three, statistics is the first thing I’d look to. How else would I know that Battier is better at those skills than say, Quentin Ross or Bruce Bowen or Mickael Pietrus?

    So maybe it comes as no surprise that the most statistically oriented front office in the league was the same one that traded away flashier players (Swift, Gay) for the selfless Battier.

  6. Henry Abbott

    Ooh, and one other point. I feel like perhaps you missed my main point about stats. I am in fact making an argument FOR the kind of smart work people like you, Knickerblogger, do with stats. (My Eddy Curry story was strictly about points and rebounds, in contrast to PER or whatever other more holistic new breed stat you want to discuss.) I actually solicited opinions on that exact post from two people: Dan Rosenbaum and David Berri. Dean Oliver commented too. Understanding, appreciating, and promoting their work is a consistent theme of TrueHoop.

    A lot of NBA fans out there still believe in points and rebounds above all else, and I believe that’s foolhardy. I’m trying to seperate thoughts about old school basic stats from refined new school stats, which to me have great promise with very real implications in the way basketball decisions are made.

    My point was more that statistics aren’t ready to replace humans in ranking the top ten centers of all time. For instance, you suggest some notion of points, rebounds, free throws, all-star appearances, and MVP trophies? Do those five things really add up to have any lasting meaning? My point is: even if we trust heartily in stats, it’s still a crap shoot.

  7. Rick

    The reason that statistics (EqA, VORP, WARP, WHIP, etc.) are so valuable in baseball is that baseball is essentially a series of one-on-one plays that can be assigned values. A batter at the plate either makes an out or does not make an out. A pitcher either strikes out a player, walks/hits a player, or the ball is put into play. In basketball, we cannot break down the plays as such. Like quantifying a catcher’s ability to call a game, how can we accurately measure a player’s ability to set a pick and roll? How can we determine that Iverson is a valuable (not “great,” for the sake of this argument) player at 39% shooting simply because he “makes people open” or “is an excellent rebounder?”

    “There is no doubt that Curry can score one on one against just about any center in the league. However, like any NBA player, Curry can?t score with two defenders draped on him. And unfortunately Curry is unable to find the open man when double teamed.”

    While I agree that he is a poor (read: lazy) rebounder and lacks the passing skill of a center like Shaq, he still (and you assert that he cannot score on double teams) shoots 58% from the field. 58%! While field goal percentage is certainly not the best metric available to us, it certainly says more than than PPG or AST/TO. That kind of efficiency, double-teamed or not, lazy rebounder or not, 7′-horse-with-blinders-on or not, cannot go unnoticed.

    And I agree that PPG and RPG are very crude indicators of a player’s value.

  8. KnickerBlogger Post author

    In all honesty I don’t care so much about the list. I think reasonable people can disagree on such things. I’d guarantee that a post on APBRmetrics would merit just as hearty a discussion.

    What bothered me was the holding of statistical analysis to a higher standard than non-statistical methods. Mind you, I think championships and awards are a good place to start when evaluating a player’s greatness, but there are pitfalls to this method as well. Hollinger does a good job in three sentences in showing how Russell’s titles are biased. (“Russell played alongside no fewer than eight Hall of Famers who were in their prime — Sam Jones, Tom Heinsohn, Bill Sharman, John Havlicek, Bob Cousy, Frank Ramsey, Bailey Howell and K.C. Jones. Russell didn’t have all seven at the same time, obviously, but he never had fewer than three at his side and in most years he had four or five. Imagine for a moment a team with one of our other top centers in the middle and four Hall of Famers surrounding him, and the 11-in-13 thing seems a little less mystical.” )

    People are seem to be more willing to use team methods (titles, wins, wtc.), that seem to incorporate so much more than one individual’s accomplishments, when those methods are highly flawed, but at the same time shun statistics for lesser infractions. Statistics have their place, but they can’t explain everything. And observational analysis can give you a greater understanding, but it tends to be too biased. There’s plenty of room for both.

  9. KnickerBlogger Post author

    Hmmm I think I get it now. Since there was no criteria you chose a more liberal definition. Additionally the Eddy Curry analogy was for those that only look at points and rebounds. Of course how would they know if he’s hurting the team (“our points per game is going up”).

    I re-read your post with all your comments in mind, but something like this sticks out:

    The truth is nobody knows what’s most essential, out of five players acting in concert, in getting wins. Does the team with the center that gets the most points and rebounds get the most wins? Is that the best measure of anything? Was Wilt better than Russell, when the proper goal of every one of their matchups was to win? What does a team need to win?

    From Dean Oliver’s four factors we know that shooting percentage is most vital, as is not turning the ball over, getting offensive rebounds, and converting from the free throw line.

    I feel like Stan in the latest South Park episode “I get it. I don’t get it.”

  10. Kevin Pelton

    Yes, but keep in mind we know that shooting percentage is most important *for the team*. Getting from individual shooting percentage to team shooting percentage, as the usage v. efficiency debate reminds us, is a lot more complicated.

  11. KD

    Ranking players is real work. Tough, tough work. Interchanging the 12th and 13th week by week … so tough.

    Had I been in on this vote, I would have killed what little cred I’ve established on the internets. Because the criteria wasn’t actually written out, Walton would have been my number one — because, if I were able to put an all-time team together for me to coach, he’d be my pivot:

    C: ’77 Walton
    PF: Russell
    SF: Bird
    G: Jordan
    G: Pippen

    … and we’d triangle it up.

  12. KnickerBlogger Post author

    KD – that’s a good idea to have Russell and Walton, so when Bill’s out for 40 games to follow the All-Time Grateful Dead team you still have a center.

  13. jon abbey

    “Hollinger does a good job in three sentences in showing how Russell?s titles are biased. (?Russell played alongside no fewer than eight Hall of Famers who were in their prime ? Sam Jones, Tom Heinsohn, Bill Sharman, John Havlicek, Bob Cousy, Frank Ramsey, Bailey Howell and K.C. Jones. Russell didn?t have all seven at the same time, obviously, but he never had fewer than three at his side and in most years he had four or five. Imagine for a moment a team with one of our other top centers in the middle and four Hall of Famers surrounding him, and the 11-in-13 thing seems a little less mystical.? )”

    but part of the reason those guys are in the Hall of Fame is because they won so many titles, so that line of reasoning seems at least somewhat flawed to me.

  14. dave crockett

    “Could one be so bad an announcer as to invalidate their entire Hall-of-Fame playing career?

    I think Bill Walton can.”

    Never in a million years would I go on record defending Bill Walton as an analyst. I mean, I rarely hit mute on a game but when he’s on…

    Having said that, when I lived in Arizona Walton would do some college games for FoxSportsNet. He adopts a very different on air persona for the college game. I found the difference stunning. He knows he’s being an a-hole on the NBA telecasts. But he seems to be less ridiculously critical of the college kids and more committed to providing insight.

  15. Captain Bringdown

    Aside..

    I couldn’t just let this sit – can I say that I find the whole basketball stats-to-capitalism analogy in poor taste? Dealing with poverty is a much more pressing need than fairly evaluating Raja Bell’s contribution on a basketball court, and your comment comes off as kind of glib.

    Seems to me that capitalism and statistical analysis are both first and foremost tools, a means to an end, each (in their own way) terribly effective at what they do well, and terribly ineffective at what they do poorly. We should, it seems to me, try to reduce the negative consequences of each to get something finely polished and supremely versatile, not just “live with the deficiencies.” Just struck me the wrong way.

    /poli sci whining

    Great site, though.

  16. Rick

    I don’t find the stats/capitalism comparison in poor taste. That’s hypersensitivity, I think.

    ===============
    But should we just discard all statistics because it gives a few players the short end of the stick? That would be like eliminating capitalism because of the poor.
    ===============

    I also do think it’s a bad analogy, since capitalism leaves many more than a “few” (try majority) of its players with the short end of the stick.

    But that’s a discussion for another board.

  17. Matthew

    I’d take prime David Robinson over O’Neal, Russell, and Walton any day of the week. He was one of the league’s best scorers and one of the best defenders. The other 3 were only ever great at one of the two, and average to bad at the other.

  18. Rick

    “Could one be so bad an announcer as to invalidate their entire Hall-of-Fame playing career?”

    This reminds me… anyone remember that article blasting Clyde as an awful announcer? Anyone agree that he might be the best out right now?

  19. jon abbey

    Bill Raftery is the best, sadly he mostly does college these days. Clyde is better than he used to be certainly, but there’s still plenty of room for improvement there.

    “ONIONS!”

  20. Rick

    I remember it was close and late in one of the games in the last month, and Jamal had an open look for a mid-range jumper, and Clyde yelled “Shoot it!”

    I love home-team bias. National broadcast teams are just so boring. Kenny Smith is definitely a great #2 for MSG.

  21. A Knicks Fan

    This is a list of the best current NBA point guards:

    10. Chauncey Billups
    9. Jason Terry
    8. Stephon Marbury
    7. Nate Robinson (he’s my fav)
    6. Deron Williams
    5. Larry Hughes
    4. T.J. Ford
    3. Chris Paul
    2. Jason Kidd
    1. Steve Nash

    But of course that’s just coming from MY prospective…Nobody cares about MY opinion. And I’m really disappointed about tonight’s loss (3/16/07) They had THREE attempts to tie or win…That can’t happen. And they can’t let the other team come from 18 points behind. Oh well. What’s done is done.

  22. jon abbey

    when did Tony Parker change his name to Nate Robinson?

    I don’t think Kenny Smith has been back since botching that FT call at the end of the Seattle game, maybe that was always the plan anyway. I kind of like him, but that was just inexcusable, hopefully someone went off on him at least for that.

    did anyone here have any hope that Channing Frye would make that one at the end (easiest of those three shots by far)? he wasn’t even close. Francis drives me nuts, but at least he has balls, he almost hit that one at the buzzer, leaning and double-clutching. Curry looked awful on the next-to-last possession too, so frustrating.

Comments are closed.