how you rate player stats ?
in offseason there comes a lot of thoughts about basketball, one thing offcourse is rating players who are, who will be the best..
statitistics are one of things you have to check at least once in a while to get clear picture on how good player is regulary..
my question is how you rate statistics when compare players or when trying to find out how good he is..
I usually like to get total of players PTS+REB+AST+STL+BLK just then look at his percentages turnonvers..(and then offcouse here comes how good of the leader, how clutch, is he a good defender and so on..but let take that out of the picture, the questions is only about stats)..
so know I'm thinking is it really equal between points and other stats,.. isn't how you score and the how good allaround player you are?? or wise verse 20ppg 10apg looks better for PG then 23/7, or 20/10 big man might be more efficient then 23/7 and so on...
your thoughts guys??
Martin Manley recently committed suicide. He had planned this and did it on his birthday, leaving behind a site. He also came up with "efficiency rating", which is something very commonly used. It is not "player efficiency rating" or "PER", but a fairly simple formula that goes as follows:
EFF = [(Points+Rebounds+Assists+Blocks+Steals)-(Missed FGs + Missed FTs + Turnovers)] / Games Played
The main difference in final total obviously would be minutes played, but in the link I posted it is kind of crazy how close a guy like LeBron James' EFF comes to his PER, with efficiency rating being a much more simple formula. Either way, found the death of this man sort of fascinating, as well as crazy. His site is a major trip and while he makes it seem like his suicide was not caused by what normally might cause suicide, you have to wonder what the hell he was thinking behind everything he seems to have written.
As far as how I rate player stats, I think whether the player helps his team win games and stay competitive is important. Not to say that certain players are not in difficult situations to do so, just think that at the end of the day, you want to be on the positive end of the scoreboard. Otherwise, statistics are mostly a gauge as to how well one does against their match-up. I think offensive statistics are a much easier way of deciding the winner, while defense is something you kind of have to see with your own two eyes as opposed to getting a lot from the stat sheet.
Overall, I rate players on how they do on an average game, which I guess does require average statistics. I feel some people get carried away with individual game performances and that consistency is crucial. Guys like LeBron James, Kevin Durant and Chris Paul are consistently going out there and killing it, which to me makes them the three best players in the league (with Kobe Bryant probably coming in at 4th right now). These three also have the highest EFF and PER. After that, it is kind of statistics combined with a feeling. Carmelo Anthony is certainly a player I feel is up near the best, despite what many others may think, though I am not sure he is right after the first three, even if he is in PER. Brook Lopez is good, just not a top 5 or even top 10 player in my mind.
It is kind of interesting, basketball-reference.com used to have efficiency rating which I think they might have called "game score" when you looked up player statistics over a season. Seems like they took it off. I liked having that when looking at individual game performances, or even having the average for a series. With EFF, 20+ usually means you are having a very good game, with 30+ being excellent. I see the value of PER, which is ESPN's go to formula, just feel like it has its flaws and that sometimes John Hollinger would defend it to the death even in situations where they might not have been reason to do so. In the end, watching these players and seeing how they effect a game is still tremendously important. Statistics and analytics are good tools, yet people still realize how important the eye test is plus getting to know the players mentality and attitude.
I'm all about high shooting percentages, low turnovers, and improved production. The EFF and PER ratings didn't really rise to popularity until a few years ago, so I never really got that in to them, but I can see their advantages when looking at a player.
I've always been good with numbers and tracking trends, so when looking at player stats, I look for trends and improvement. I'm more apt to look at a player who constantly improves one or two aspects of his game as he adjusts to the league than a guy who comes on strong, but puts up pretty much the same shooting numbers every year.
It's also pretty easy to see which fringe players are going to be out of the league and which ones will get another chance after their rookie contracts if you track improvement. I also weigh post All-Star stats heavily in rookies because it separates a chunk of the season in which their early adjustment period has passed. Some guys produce at the same level late in the years, but the guys looking to blow up as sophomores really get a lot better the 2nd half of the year, and not just in April when teams rest everyone.
Xavier Henry, for example, still has some ( althought limited at this point) upside, but has produced at virtually the same level since his rookie year. In addition, he's has never been a guy who has really upped his game and production when he got more minutes late in the year. Some guys who don't get a lot of burn are ready to break out, but guys like him have a hard time getting a 2nd contract because they haven't proven much. Henry is a good defender, which doesn't show up in his stats, but because he hasn't gotten a lot better, hasn't gotten much more efficient, and hasn't looked like the best player on the court when in garbage time against 8th-12th men. I'm not surprised it's early September and he's not with another team.
stats are a great guideline, but should be taken like a grain of salt. the problem with eff is that it under values defense. +/- isn't bad, but you can't go all by that either because sometimes there are players that might be starters on other teams that play against a bunch of bench warmers, but it should be looked at. add win shares to that equation and your closer. but you must include the eye test and what roll a player has on that team to. sometimes a player is designated on the team as a stopper and shutting down the oposition's best player, taking a lot of energy out of the player sacrificing stats like iguodala's role often. sometimes a player is just a utility player to. good example, who would you rather have prime? antawn jamison or robert horry? who has the better stats? horry specializes in clutch shooting, can play both ends of the floor, shoot it from anywhere, unselfish, and one of the best inbound passers in the game. he also helped win games whether he was replacing shaq at center for a game in the finals as the go to interior option, or the 4th or 5th option on the court due to not having that me first ego unless it's necessary. antawn jamison on the other hand was arguably the best pure scorer in the league at one point of his career and shot a good field goal percentage. he was an average rebounder and didn't do much of anything else. it would be harder to build a championship team with jamison than horry imo. stats are a good tool to measure how good a player is, but not always the set in stone proof. in fantasy basketball, stats are everything. but in the real basketball world it's not always true. if stats were everything, wilt chamberlain wrote the stat record and nobody compares. bill russel was pretty good in that era to and he deserves at minimum consideration of being the best player from that era. 11 championships! iv'e also seen teams that score a ton of points and play no defense and give up even more of an astounding amount of points and don't get near .500 in the win column. that team scores a lot, players pack a lot of stats, there in turn are more possessions in the game, but it wouldn't measure how good there players are correctly if they are statistically better than most of the league averaging 120 points a game and giving up 130 and finishing 0-82.