top 20 player WNBA all time

[stwainfan]

The top 20 at 20: a statistical analysis of the WNBA’s best all-time players

 

[Wally East]

Three problems with the methodology (and I would say this regardless of who topped the rankings).

1. Equal weight of categories. Scoring is the most important category. It's how you win games (well, that and preventing the other team from scoring but defense isn't considered here other than steals and blocks). To giving points per game equal weight to steals per game or, for some reason, minutes per game, doesn't reflect what's actually important about winning a basketball game. (Further, if Player A and Player B both average 15 ppg but Player A does it in 30 minutes per game and Player B does it in 35 mpg, Player A is more valuable.)

2. Use of rankings. Using the ranked order as a determination is weak. It gives no consideration to the space between 1 and 2 or 2 and 20. Britney Griner is first and averages 3.5 bpg. Lisa Leslie is second and averages 2.2 bpg. Candace Parker is third and average 1.9 bpg. Ranked ordered just puts them 1, 2, and 3, equally spaced.

1 could average 12 blocks per game and 2 could average 3.0 bpg while 3 averages 2.9 bpg. Ranked order sees the space between 1 and 2 and 2 and 3 as equal.

3. No accounting for efficiency. Not all 20 point games are created equally. If Player A takes 12 shots and Player B takes 24 shots and both average 20 points (assuming the same number of free throws just for the sake of the example), Player A had a MUCH better game and gave her team a much better chance to win. Missing shots is a negative for your team.

Basketball-reference.com has a list of career leaders for win shares per 48 minutes. Cynthia Cooper is first, which, duh. Lauren Jackson and Maya are second and third. (Here's a link to the explanation of Win Share.)

 

[pinotbear]

Any ranking that puts Diggins in the top 20, with her vast list of time served, accomplishments, and durability... isn't worth the paper it's printed on. There's a number of things you can question in this list, but, geez, she's played 2.3 seasons, shoots under 40%, under 30% outside the arc, hasn't won squat, ... how da hell does she come in as one of the 20 best players?

 

[UcMiami]

Wally - spot on. One of the stupidest methodologies I have ever seen. To take statistics and then rank them and use the rankings then as the values for averaging is just DUMB. I have no problem with counting the statistics equally (except minutes, which if anything should be counted as a negative value - fewer minutes to amass the stats = better production while on the floor) - a steal is easily equal to a single point except in the least efficient offenses, as are rebounds. And of course these stats are going to end up weighted towards forwards and posts since rebounds tend to be the second highest number after points

 

[baggerbob]

Wally - spot on. One of the stupidest methodologies I have ever seen. To take statistics and then rank them and use the rankings then as the values for averaging is just DUMB. I have no problem with counting the statistics equally (except minutes, which if anything should be counted as a negative value - fewer minutes to amass the stats = better production while on the floor) - a steal is easily equal to a single point except in the least efficient offenses, as are rebounds. And of course these stats are going to end up weighted towards forwards and posts since rebounds tend to be the second highest number after points

Bill Russell never ranked high in any Stats in college, Olympics, or pro's, but was probably the greatest winner of all time. I preface that, by saying I'm not a Celtics fan, never have been, but winning is the only Stat that counts.

 

[CocoHusky]

Three problems with the methodology (and I would say this regardless of who topped the rankings).

1. Equal weight of categories. Scoring is the most important category. It's how you win games (well, that and preventing the other team from scoring but defense isn't considered here other than steals and blocks). To giving points per game equal weight to steals per game or, for some reason, minutes per game, doesn't reflect what's actually important about winning a basketball game. (Further, if Player A and Player B both average 15 ppg but Player A does it in 30 minutes per game and Player B does it in 35 mpg, Player A is more valuable.)

2. Use of rankings. Using the ranked order as a determination is weak. It gives no consideration to the space between 1 and 2 or 2 and 20. Britney Griner is first and averages 3.5 bpg. Lisa Leslie is second and averages 2.2 bpg. Candace Parker is third and average 1.9 bpg. Ranked ordered just puts them 1, 2, and 3, equally spaced.

1 could average 12 blocks per game and 2 could average 3.0 bpg while 3 averages 2.9 bpg. Ranked order sees the space between 1 and 2 and 2 and 3 as equal.

3. No accounting for efficiency. Not all 20 point games are created equally. If Player A takes 12 shots and Player B takes 24 shots and both average 20 points (assuming the same number of free throws just for the sake of the example), Player A had a MUCH better game and gave her team a much better chance to win. Missing shots is a negative for your team.

Basketball-reference.com has a list of career leaders for win shares per 48 minutes. Cynthia Cooper is first, which, duh. Lauren Jackson and Maya are second and third. (Here's a link to the explanation of Win Share.)

Sincerely, thanks for taking the time to attack the methodology and propose a reasonable alternative as opposed to the reflex action to deem it automatically wrong because the players at the top wore orange as collegians.

 

[Wally East]

Bill Russell never ranked high in any Stats in college, Olympics, or pro's, but was probably the greatest winner of all time. I preface that, by saying I'm not a Celtics fan, never have been, but winning is the only Stat that counts.

Whoa whoa whoa Let's not be crazy. He wasn't the most prolific score, sure. Over the course of 9 seasons, he was the league's leader in rebounds per game 5 times. He was frequently among the top 10 in assists.

 

[UcMiami]

Just as a follow on to my comment about rebounds - typical box score for a team:
Points = 80
Rebounds = 45
Assists = 20
Steals = 10

So an 'average' 40 minute player should amass 16 points, 9 rebounds, 4 assists and 2 steals

And on most teams points will be distributed fairly evenly between forward/centers and guard/wings while rebounds will be weighted heavily towards forward/centers and assists towards guards - steals end up pretty even. If the weighting is 75/25 the 'average':
Forward center will have:
16 points, 13.5 rebounds, 2 assists and 2 steals = 33.5
Guard will have:
16 points, 4.5 rebounds, 6 assist, and 2 steals = 28.5
Wing/Swing will be close to average:
16 points, 9 rebounds, 4 assists, and 2 steals = 31

That sort of disparity shows up fairly consistently when looking at raw statistical totals.

I also don't like any analysis that doesn't take into account efficiency: assists minus turnovers for example, and points adjusted for shooting percentage.

And rebounds have always bothered me a bit as defensive rebounds which account for the 80% +/- are mostly the result of team effort while offensive rebounds tend to be more individual effort.

Wally - I have never been a fan of 'win percentages' because I think it unfairly penalizes great players on bad teams.