This stat measures an accomulation of player's stats, if you're Westbrook or Harden you'd have the highest PER but if you're a role player or a defensive specialist PER wouldn't reflect your ability. It doesn't reflect doing the little things, it doesn't reflect defense at all. It wouldn't reflect Meyers Leonard being completely lost on defense compared to Olynyk being a very high basketball IQ defender, for example, or wouldn't reflect - like in my previous example - RoLo's impact in terms of defense and rebounding compared to Aldridge's. If Dame decides to shoot the ball when he really should pass it, if he does it frequently enough, it would raise his PER. Basketball is about stats maybe but also about decision making, about being in the right place on defense to close lanes, to defend the middle, these things you can only understand through the eye test. So PER is good for players who are primary ball handlers, first option on offense etc. Players who are highly regarded anyway. But we already have these players, our major building blocks in Dame, CJ and Nurkic and they too would be more effective if they play a less selfish, more team, basketball (which would probably lower their PER if they did) and surrounding them with the right role players to form a winning team is not mainly about PER or even stats in general. Also, the usage of stats, any stats, is subjective. It can tell you something but it doesn't prove anything. If I tell you I like a player by seeing him and tell you he's a great player and then you come and "prove" me wrong by selecting some stat that supposedly make him look bad, you're being a demagogue and misleading the other posters who maybe haven't seen the player and are convinced by whatever advanced stat you're presenting. It's true that I have rarely seen the Blazers this year but I don't see them being very good in the W-L stat without my opinion. I'd continue to voice my opinion on what I do know, Olynyk I've seen play and understand his impact, especially on team defense. If I've never seen him play or had any kind of impression of him, I wouldn't share any thoughts or opinions about him
First, I want to make it very clear, I don't think Kelly Olynyk is a bad player. Given the turn this thread has taken, ironically, PER says he is indeed a good player, but not an all star caliber player. My first post in this thread was not intended to trash Olynyk, I was just giving MickZagger a hard time for creating this thread, bumping others and saying Olynyk would be an all star if he was a starter based on a ridiculously small sample size of one game. Here's what I wrote: He has a career PER of 15.6 and has been above 15.0 every year of his career. That's very good for a role player who is not one of the top four scoring options on your roster. Most teams would be quite happy having a player like Olynyk coming off their bench. The -8.3 PER I quoted for last year's playoffs did, indeed, accurately reflect Olynyk's play in those 4 games. Again, 4 games is a very small sample size for calculating a complex stat like PER, but in those 4 games, Olynyk absolutely stunk. He was hot garbage. He didn't make a single 3-pointer in the series, shot .111 FG% and averaged 0.5 points and 1.0 rebounds per game. The low PER was a direct result of how poorly he played over that very limited sample size. Just like one great game does not make a player an all star, 4 bad games does not make him a worthless scrub. The truth in most cases, and in this one specifically, lies somewhere in the middle, which was my whole point. You can't base a player's worth, good or bad, on such a small sample size. That is only partially true. The defensive specialist part is most definitely true and widely acknowledged. Usually when I use PER to compare players' values, I issue a standard disclaimer that PER does not adequately comprehend a player's value on the defensive end. Most people who have heard of PER know that, but if I don't mention it, I immediately get dismissed by know-it-alls (Dan Shelton) who say PER is useless because it doesn't include defense. To which I say, no it's not useless, if you understand how it's calculated and what it's limitations are. That said, since PER is an accumulation of so many stats, and I've said this many times, if you have a PER less than 10.0, and you're not a elite, Bruce Bowen level defender, you're hurting your team because you suck at everything else. It doesn't take much to have a PER above 10.0. Just doing one thing above average, even if you suck at everything else, usually gets you there. This can actually work in the favor of some specific types of role players. Most notably, back up big men who rebound well and never shoot the ball more than 3 feet from the basket. Those types of players do two things that PER values much better than the average player - rebound and make a high percentage of their shots. It's not uncommon for these guys to have a PER > 20, which is generally considered all star level, but no one in their right mind would consider Cole Aldrich an all star. An extreme example is Enes Kanter. The last two seasons, he's posted PER of 24.0 and 23.7. That's good for 10th best and 17th best in the entire league over the last two seasons. However, no one in their right mind, who has actually watched him play, would argue that Kanter is one of the top 20 players in the league and deserves to be an all star. People who understand how PER is calculated know that it over values the things Kanter does well (elite level offensive rebounder and very efficient scorer) and undervalues the things he does poorly (defense). PER also equalizes for minutes played. This was Hollinger's attempt to standardize a players contributions and not just see the guys who play the most minutes at the top of the list. Unfortunately, I think his approach was an oversimplification and perhaps the weakest part of his whole formula. It often results in what I call small sample size superstars, guys who rack up meaningless stats in a few garbage time minutes, but it also favors bench players over starters. To truly judge a player's value, you also need to consider the quality of his opposition. Guys who start, generally play most of their minutes against their opponents' best players, guys who come off the bench, generally play a large percentage of their minutes against other teams' back ups. Of course, there are exceptions. Often a team has a token starter that starts the games, but only averages about 15 MPG, in which case, his back up that is averaging 32 MPG is the one playing in the 4th quarter with the game on the line against the other team's best. Also, players that average 18 MPG don't have to deal with the same level of fatigue as players who average 36 MPG. So, if I was devising a one-size-fits-all stat and wanted to normalize for minutes, I would include factors to adjust for starter vs. back up and minutes played. It would be a sliding scale, but players who start and average more than 32 MPG would get the most "credit" for their minutes, followed by all players that average more than 28 MPG, followed by players that start but average less than 24 MPG, followed by backups that average more than 18 MPG, followed by all players that average less than 10 MPG. Let's face it, if you average less than 10 MPG, there is a reason. Anyway, the point is PER favors specific types of role players and penalizes others., When discussing PER, it's very helpful to know and understand these idiosyncrasies. All noted and common knowledge. Nope, in most cases the exact opposite is true. Assists and missed shots (FGA - FG) both heavily factor into calculating PER. Although the acronym PER (Player Efficiency Rating) is a bit of a misnomer, it does not reward inefficient chuckers. If Dame does indeed force shots when his should be passing, both his assists and shooting percentages, and thus, his PER will suffer. He would basically become Jerry Stackhouse. Stackhouse, in his prime in DET was a great, but not terribly efficient scorer. In 2000-01, Stackhouse averaged 29.8 ppg and came in second in the scoring race to Allen Iverson. But, his PER was only 21.8, which is very low for someone with such a gaudy scoring average. Compare that to Dame's scoring average and PER this year: Jerry Stackhouse (2000-01): PPG = 29.8 PER = 21.8 Damian Lillard (2016-17): PPG = 27.0 PER = 24.1 So, in spite of Stackhouse scoring almost 3 more PPG, he had a much lower PER than Dame. Forcing up bad shots will kill your per. Stackhouse scored more, which, all other things being equal, will help your PER. But all other things aren't equal. He also had more turnovers, fewer assists, and most importantly missed a lot more shots. Stackhouse did average 5.1 AST that year, but he also lead the league in missed FG. Dame had higher shooting percentages across the board. Forcing up bad shots would not increase his PER. Agreed. PER is a tool, but it's just one tool and an imperfect one at that. That doesn't make it useless, if you understand how it is calculated and it's limitations. It can be useful for comparing players in similar roles and as a short hand way of distilling several stats into a single number. Not necessarily true. As the case with Kanter and other back up bigs, it favors different roles differently. It's not just about the players who handle the ball the most or take the most shots. If those players are inefficient at what they do (see Monta Ellis) they will have a lower PER than players who play a similar role more efficiently (or players who play a different role and play it better). I didn't agree with PapaG on much, but one of his pet peeves was players whose PER was low relative to their USG%. I feel the same way about players whose PER is low compared to their scoring average. In both cases, these are players who are using up way too many possessions and doing so very inefficiently (chuckers). In 2009-10, Monta Ellis averaged 25.5 PPG, but his PER was only 16.7. That's an incredibly inefficient 25 PPG. No one else in the modern era comes close. Many casual fans would see than 25.5 PPG and think Ellis had a good year, quite the opposite. He was GSW's leading scorer that season, and the next, but his is incredible inefficiency is why GSW got so much better after they traded him and started Klay Thompson at SG. For those that use the Curry/Ellis backcourt as "proof" that Dame and C.J. can't coexist, I say: Monta Ellis (2009-10): PPG = 25.5 PER = 16.7 Steph Curry (2009-10): PPG = 17.5 PER = 16.5 Combined: PPG = 43.0 PER = 33.2 C.J. McCollum (2015-16): PPG = 23.0 PER = 19.9 Damian Lillard (2015-16): PPG = 27.0 PER = 24.1 Combined: PPG = 50.0 PER = 44.0 Ellis was an unconscionable, historically inefficient chucker and Steph Curry was not close to the player he would become. Lillard and McCollum are both much more efficient scores than Ellis and Curry at the time they played together. Anyway, just being a primary ball handler and No. 1 option does not guarantee a high PER. You still have to be good at those things. Of course, the addition of Bogut was the other thing that helped turn GSW around and we already got our much younger Bogut in Nurkic. Surrounding them with the right role players might lower their scoring averages, but it should actually improve their PERs. If we have additional reliable scoring options, Dame and C.J,. will have to force fewer shots and defenses won't be able to leave other guys wide open to double them as much. Their shooting percentages would increase, and as a result, so would their PERs. Stats are a tool. By themselves, they don't prove anything. I use them to point out a players strengths and weaknesses. If I use a players stats to "mislead" someone about a player they haven't seen play, then it's their own damn fault for not watching that player actually play basketball. Don't blame me for someone else's lassitude. If they don't care enough to actually watch a player play, then their opinion means absolutely nothing, whether I agree with them, or not. Yes, I use stats, but I also watch a hell of a lot of basketball. I said previously, the two go hand in hand. Completely ignoring one, or the other, gives you an incomplete view of a players talents and production. Many people here thought Victor Claver passed the eye test, that he "looked like a basketball player" and would eventually put it all together. His stats were abysmal. How you "look" doesn't matter if you can't produce. That's why he's out of the league. I feel the same way about Jake Layman. Other than his first 8 minutes of garbage time against GSW, he flat out sucked. Even in the D League, where his raw numbers were good, his percentages were absolutely awful. He was an inefficient chucker playing against weak competition. Many here love the form on his shot and the way he shoots without hesitation, but at the end of the day there are no style points in basketball. If the ball doesn't go in, it doesn't matter how great your form looks. Some people here think Jake Layman passes the eye test. He could still turn it around but so far, to me, he looks an awful lot like Victor Claver 2.0. They were't very good in the W-L stat without Nurkic, but were 14-6 with him. His impact goes beyond his own stats. I've shown in other threads, how much our team 3FG% improved after adding Nurk, even though he didn't make a single 3-pointer. I've also talked about how his defensive contributions go way beyond blocked/altered shots and steals. He has great stats but his impact goes far beyond his individual stats. I have never suggested that anyone base their opinions 100% on stats. They are just a tool, but they are useful tool when used properly. Analytics wouldn't be a thing if all stats were meaningless. But in addition to analytics experts, all teams still employ scouts. There is a lot of money at stake, so you need to use every tool at your disposal when constructing an NBA roster. I don't have that level of responsibility, but I still like having access to the data the teams use to help inform my own opinions about who this team should keep and what other players we should target through the draft, free agency and trades. To me, it's all pat of being a fan. BNM
Didn't read through the whole post but your explanation on the minus PER being from small sample size makes me understand it better (and appreciate the formula better) because Olynyk is a very good player IMO. Also, I think Boston is looking to use their cap space and maybe let him go to do that. If that's the case I would want us to offer him what we can afford to pay, the MLE or whatever we have with the new salary-cap
Please take the time to read it all when you have a chance. You made an effort to understand my use of PER. So, I made an effort in return to further explain how and why I use it, and when I don't. It's a complex stat, that on it's own, is neither inherently "good" nor inherently "bad". It is what it is, but understanding what it is helps during discussions where it comes up. BNM
I've seen on one of the draft sites that they have us picking some Center who has 4+ Points and 4+ Rebounds per game but a relatively high PER, so I understand it's more than just an accomulation of stats. Still haven't seen his videos, Ike something his name is
No, I know it isn't. I barely watch the games. People can debate the finer points of needle point vs crocheting techniques (or some other boring shit) Somewhere in that massive wall of text you said it influenced how you saw the game. If I did have time to watch the games I'd just want to enjoy them. It's like the movies, I used to get turned off when a car would explode like it was loaded with C4 if someone shot the radiator. I just ignore that stuff now and movies are way better.
It's not that he sucks completely, but you said 1)We needed to draft him 2) he was the best player in the draft 3) and he was going to be the closest thing to Dirk there is. We drafted CJ instead of him, so wrong on the first one. CJ, Giannis Antetokounmpo and Rudy Gobert were in the same draft, so wrong on the 2nd one. Wrong on the third one, obviously. Keeping in mind his minutes and production have been basically flat ever since he came into the league, lets take a look at his other elimination games. In last years elimination game against Atlanta he was 0-2 from the field, zero points, zero assists but he did have one big rebound. The year before that, in the elimination game against Cleveland, he was 0-1, one point, zero assists and one big rebound. But that's great you made a thread to gloat about him now that he's had his first good playoff game at the age of 26. Props to you.
This is why you don't talk shit based on a one game sample size. 2 points on 1-6 shooting. It looks like Games 3 and 6 of the WAS series were much more typical of Olynyk's performance in the post season than Game 7. In 22 career playoff games, Olynyk has scored 2, or fewer points 9 times, including four scoreless games, two 1 point games and three 2 point games. So, he has scored 2, or fewer points in over 40% of his career playoff appearances. That Game 7 performance is truly a statistical outlier. It's the only time he's ever scored more than 14 points in a playoff game. BNM
You want a thread to bump? Forget KO, go for Deal Coming Soon! p.s. Hahahahaha, that thread is so old, the search function can't find it!
Bump, bump, bump it up: http://www.basketballforum.com/portland-trail-blazers/365887-deal-coming-soon-merged.html
