using PER, WS48, and Simple Rating (82games) to rank players

spurket · Post #1 » by **spurket** » Wed May 2, 2012 7:52 am

When I want to see how well a particlular player is playing, these are my favorite 3 stats to look at:
PER at espn.com
Win Shares per 48 at basketball-reference.com
Simple Rating at 82games.com

Most people feel that LeBron is the best player in the world followed by Durant and Chris Paul.
LeBron is 1st in all 3 of these stats.
Chris Paul is 2nd in all 3 of these stats.
Durant is 4th in PER, 3rd in WS48, 5th in Simple Rating
Wade is 3rd in PER, 5th in WS48, 3rd in Simple Rating

top 10 PER
LeBron
Paul
Wade
Durant
Love
Howard
Ginobili
Blake
Rose
Westbrook

top 10 Win Shares per 48
LeBron
Paul
Durant
Harden
Wade
Noah
Love
Tyson Chandler
Ryan Anderson
Boozer

top 10 Simple Rating
LeBron
Paul
Wade
Howard
Durant
Anderson
Nash
Garnett
Rose
Blake

starvinmarvin17 · Post #2 » by **starvinmarvin17** » Wed May 2, 2012 9:10 am

Where is the kobester

C-izMe · Post #3 » by **C-izMe** » Wed May 2, 2012 9:20 am

Win shares per 48 is a horrible stat to use. Simple rating is too because all it is is PER difference and one year adjusted apm. They end up ranking the best of the best correct but you also get Noah, Tyson, Anderson, etc. in the top ten. I don't use any of those but PER (and that is only for comparing boxscore stats).

DeltronZeroHero · Post #4 » by **DeltronZeroHero** » Wed May 2, 2012 9:55 am

How about actually watching the games instead of using pointless stats?

Aeternus · Post #5 » by **Aeternus** » Wed May 2, 2012 12:03 pm

DeltronZeroHero wrote:How about actually watching the games instead of using pointless stats?

Yeah, why is that people don't watch hundreds of games every year while remaining completely unbiased towards teams, players, styles and situation, while also memorizing perfectly every play made and summarizing it mentally so that they can have an objective representation of the past season?
Using stats sure is for losers. :roll:

83SixersRocked · Post #6 » by **83SixersRocked** » Wed May 2, 2012 12:04 pm

DeltronZeroHero wrote:How about actually watching the games instead of using pointless stats?

This.

Stats are great for fantasy teams, but there's more to it than that.

MMyhre · Post #7 » by **MMyhre** » Wed May 2, 2012 12:33 pm

The top 3 is kinda good though, insert Durant for CP3 and it's K. CP3 is 4-5

jinxed · Post #8 » by **jinxed** » Wed May 2, 2012 1:28 pm

You left out the best stat, Adjusted Plus/Minus.

From basketballvalue.com

Over the last two years...including playoffs (*starters only*) ..Needs two years of data to be really accurate.

1.Dirk (no one else is even in the same ballpark)
2.Nash
3.CP3
4.Millsap
5.Josh Smith
6.Pau
7.KG
8.Lebron
9.Dwight
10.Wade

*Manu would rank 3rd, but doesn't qualify for enough minutes over past two years.

bigpimpatl · Post #9 » by **bigpimpatl** » Wed May 2, 2012 1:45 pm

DeltronZeroHero wrote:How about actually watching the games instead of using pointless stats?

What's the difference between watching a player score x points with y shots at z efficiency and using statistical analysis to come up with the same conclusions?

stats are only pointless to you because you a) don't understand them or b) purposefully ignore them.

branch · Post #10 » by **branch** » Wed May 2, 2012 1:55 pm

bigpimpatl wrote:
DeltronZeroHero wrote:How about actually watching the games instead of using pointless stats?

What's the difference between watching a player score x points with y shots at z efficiency and using statistical analysis to come up with the same conclusions?

stats are only pointless to you because you a) don't understand them or b) purposefully ignore them.

Can stats measure leadership?

Post #11 » by **nate33** » Wed May 2, 2012 2:08 pm

branch wrote:
bigpimpatl wrote:
DeltronZeroHero wrote:How about actually watching the games instead of using pointless stats?

What's the difference between watching a player score x points with y shots at z efficiency and using statistical analysis to come up with the same conclusions?

stats are only pointless to you because you a) don't understand them or b) purposefully ignore them.

Can stats measure leadership?

I'm so tired of this debate. Nobody with a brain thinks that stats alone in a total vacuum tell you everything you need to know. And clearly, "watching the games" as your sole means of talent evaluation is a flawed strategy because you can't possibly watch enough games and accurately track all the players without bias.

You need both. The OP merely posted the summary stats that he likes to use. Why do the anti-stat nazis get so crazy when people talk about stats?

I personally start with PER and Usage Rate to get a good idea of a player's offensive aptitude, recognizing that it's difficult to remain as efficient when the usage rate rises. I then consult the team on/off numbers to see if I can get a read on how a player impacts the team's performance, keeping in mind that on/off numbers are fairly "noisy". And finally, I check out the counterpart PER stats to see if a player is doing a good job as an individual defender - though I take these numbers with a hefty portion of salt, particularly for big men who have more help defense responsibility than one-on-one defense responsibility.

In general, I'm pretty trusting of the numbers in their ability to gauge a player's offensive aptitude. I'm much less trusting of their ability to gauge a player's defensive aptitude. Eventually, you need to watch some games to see if a player knows how to defend.

I don't like Win Score at all. It seems to overweight rebounding and gives way too many questionable results.

chubby_1_kenobi · Post #12 » by **chubby_1_kenobi** » Wed May 2, 2012 2:13 pm

Kevin Love is so awesome.

HurricaneKid · Post #13 » by **HurricaneKid** » Wed May 2, 2012 3:11 pm

I love the anti-stat nazis. They crack me up.

http://www.insidehoops.com/forum/showth ... p?t=249631

The truth is there are things that stats can't measure that make a huge difference. Setting solid picks, taking charges, contesting shots, help D, etc. But those largely go unnoticed by most observers anyways; observers who tend to overrate spectacular dunks, miraculous shots, etc instead.

Stats don't tell the whole story but they are far more accurate than almost anyone can get through even advanced viewing. Far more accurate than most front offices even.

jinxed · Post #14 » by **jinxed** » Wed May 2, 2012 5:11 pm

Ginobili's .257 WS48 would rank 3rd if he qualified with enough minutes.

JimboSlice · Post #15 » by **JimboSlice** » Wed May 2, 2012 5:19 pm

So the moral of this story is, I guess everybody should purchase synergy and stop arguing so that they can measure everything statistically that you can see in a game besides leadership, which they'll automatically attribute to whoever has been mentioned as a leader in any interview they've ever seen on espn.

I get tired of hearing "why don't you just watch the game, stats are pointless because they aren't perfect".

These are the same people who were excited about the Thunder signing Fish for his veteran leadership and playoff experience, while ignoring his career-low shooting percentages, Scott Brooks' addiction to inflating players' minutes based on name and not production, and his league-worst perimeter defense, then also ignoring the Thunder's huge difference in win % before and after the signing as well.

Yes, watching the games is important, judging strictly off of stats and recaps doesn't tell you everything, but it helps keep things objective, and can point out some things that you would definitely miss just watching.

LApwnd · Post #16 » by **LApwnd** » Wed May 2, 2012 5:21 pm

starvinmarvin17 wrote:Where is the kobester

overrated chucker

clipocketurs · Post #17 » by **clipocketurs** » Wed May 2, 2012 11:56 pm

They should use these stats to decide the MVP.

Paul should finish 2nd in the MVP vote.

Durant and OKC have the better record because Durant has better teammates than Paul.

Wade is underrated.

droponov · Post #18 » by **droponov** » Wed May 2, 2012 11:59 pm

bigpimpatl wrote:
DeltronZeroHero wrote:How about actually watching the games instead of using pointless stats?

What's the difference between watching a player score x points with y shots at z efficiency and using statistical analysis to come up with the same conclusions?

stats are only pointless to you because you a) don't understand them or b) purposefully ignore them.

The difference is that two players can score x points with y shots at z efficiency in completely different ways - and that can have a huge influence in the value and impact of the player.

Not to mention there are just too many things stats can't capture.

droponov · Post #19 » by **droponov** » Thu May 3, 2012 12:00 am

Aeternus wrote:
DeltronZeroHero wrote:How about actually watching the games instead of using pointless stats?

Yeah, why is that people don't watch hundreds of games every year while remaining completely unbiased towards teams, players, styles and situation, while also memorizing perfectly every play made and summarizing it mentally so that they can have an objective representation of the past season?
Using stats sure is for losers.

Yeps, you definitely need to memorize perfectly every play made by the player to have an opinion on him. I mean, c'mon.

droponov · Post #20 » by **droponov** » Thu May 3, 2012 12:04 am

I don't have any use to boxscore summary stats like PER.

How much time do you need to actually look at the stats that compose PER? 15 seconds? Is that too much? If you have watched the player and the team enough, you have a good idea of context and you can weight that stats much more accurately instead of using a fit all weighting.

Using PER is akin to reading ultra-condensed versions or cliff notes of great books.