Is Rondo the best NBA PG? He's almost there.

[BballTim]

I didn't read all of the op but when you say "a single assist per game", keep in mind that a player would have to take 25 shots a game in order to get to a single basket a game difference between 46% and 50% fg%.

This.

My multiplied factors were not meant to give you a quantitative difference between Rondo and the other six; purely qualitative. For instance, if you recalculate their positions based solely on their average rank in APG, ATTR, and FG%, you get essentially the same ratings, with only Rondo and Nash switched.

1. Paul 2.33
2. Nash 2.33
3. Rondo 2.66
4. Parker 3.66
5. Rose 5.33
6. Williams 5.66
7. Westbrook 5.66

I mentioned it wasn't a perfect calculation. Only ballpark.

If you normalize the statistics, you're still giving a qualitative difference.  In my example, I wasn't saying that the player with the 46% shooting percentage and 11.7 assists per game is .09 better than the player with a 50% shooting percentage and 10.7 assists per game (or comes out .09 better according to your metric.)  I was saying I think the guy with the higher shooting percentage is better, and should be reflected as such in a rating scheme.  I bet if you were to poll people on which is the better point guard, most would take the 50% FG and one less assist per game.

  Most people don't understand how fg% translates into ppg. If you asked people whether they'd rather see 15/11 or 16/10 they'd see it as more of a wash, or possibly go towards the assists.

[Rondooooooooo]

its not about who is the flashiest,

Rondo gets a lot of attention for his flashy passes, dribble moves, and steals -- and people tend to gloss over the holes in his game. So yeah.

i realize that but thats not why i consider him the best. rose is flashy as well as paul and i think thats a major contributor to why people view them as better. im not sayin just flashy plays, flashy numbers as well like PPG

Rose and Paul have carried teams on their back with little help. That's why people view them as better, and it's a legitimate point of view.

  The teams Rose has "carried on his back with little help" generally outscore their opponents when he's out of the game. He's a great player, and he scores a lot of points, but perception isn't always reality.

They don't -- neither in points per 100 possessions, nor in net point difference (this is particularly glaring in pre-Thibodeau times, by the way). Check facts much?

they carry their team with little help is not a prerequisite of a point guard. thats what the best player is for a la Lebron, Michael. They weren't point guards. I'm not disputing that Rose and Paul and great players but im saying the specific position of point guard.

So it seems what you're saying is  that Rose and Paul are not "great point guards", but "great players who play the PG position"? Excuse me if that isn't making any sense.

not at all, you're twisting everything i say. i never said they weren't great point guards, im saying rondo's a better poing guard, theyre both also great point guards but they're better overall players but i think rondo is better at the position and what the position needs than they are.

[AB_Celtic]

All I'm trying to do is show why you don't mix percentages and counting statistics the way you did.  I like that you tried to come up with something -- just wanted to help you make it better.

I know it's flawed, but I don't see any effective way to normalize the statistics. I'd be perfectly happy to listen to any concrete suggestions on how to improve it. Truly. Unfortunately, all you've told me is what's wrong with it.

[BballTim]

its not about who is the flashiest,

Rondo gets a lot of attention for his flashy passes, dribble moves, and steals -- and people tend to gloss over the holes in his game. So yeah.

  Most point guards have more holes in their games than Rondo. He's not a great outside shooter or free throw shooter, but he's arguably the best in the league in just about every other point guard skill. Aside from rebounding (where he's pretty good) only CP3 can make a similar claim.

Or to flip your argument around: Rose/Paul/Nash/[insert your favorite PG here] is not a great defender or passer, but he's arguably the best in the league in just about every other point guard skill.

  Nash isn't a great defender or rebounder and doesn't get to the rim much. Rose doesn't rebound, defend, pass, run an offense or control the pace of the game as well as Rondo. Which point guard skills (aside from scoring) do you see Rose as best in the league at?

[AB_Celtic]

...

  Nash isn't a great defender or rebounder and doesn't get to the rim much. Rose doesn't rebound, defend, pass, run an offense or control the pace of the game as well as Rondo. Which point guard skills (aside from scoring) do you see Rose as best in the league at?

That's easy. Scoring, scoring, and scoring. That's all that matters in the NBA, right? 

[ManUp]

He's 3 in my book. As far as pure point guards go, Chris Paul is number one and Deron Williams is number two. Rondo just edges out Nash who's game has been slipping production wise.

IMO, a point guard should be just as effective with his scoring as he is with his passing. Nash has the perfect balance as far as passing and shooting goes. If I had to put a ration on it Nash shoots and passes at something like a 50/50 ratio. A guy like Rondo shoots and passes at something like a 30/70 ratio. A big part of why I thought Rondo looked so good in the play-offs is because he showed a better balance between scoring and passing. If Rondo can pickup where he left off in the play-offs he'll be on his way to being number one, but not yet. 

Sidenote: Anyone who say Deron isn't a pure point is either extremely biased or hasn't seen him play in Utah. However, I will concede that D.Will has been a bit of a chucker in NJ. Still that's just 2 out of 7 years and he maintains a 9 apg average for his career.

[saltlover]

All I'm trying to do is show why you don't mix percentages and counting statistics the way you did.  I like that you tried to come up with something -- just wanted to help you make it better.

I know it's flawed, but I don't see any effective way to normalize the statistics. I'd be perfectly happy to listen to any concrete suggestions on how to improve it. Truly. Unfortunately, all you've told me is what's wrong with it.

Personally, I'd find the mean and standard deviation for all three stats, either amongst point guards, guards, or all nba players (with enough minutes to count.)  If someone  was right at the mean for a stat, they get a score of 0 for that statistic.  If they're one standard deviation above, a 1, and one below, a -1, and so on.  Then add them together.  Or weight them if you do think assists should be more important, and then add them (multiplying doesn't work, however, because of the negatives.)

You could also find the mean of the three stats, again, and then divide the player's stat by the mean.  Now if they're right on the mean they get a 1, if they're double it they get a 2, etc.  I like this method far less because it will preserve some of the differences in shooting percentage and assists, since shooting percentage is capped at 100%, whereas assists have no upper bound.  So if the mean shooting % was 48%, it'd take a 96% to get a score of 2 (practically impossible) but if the mean assists is 5, then 10 assists gets a score of 2 (very possible.)  But it gets rid of some of the differences in scale.

Either of these ways will still produce a "qualitative" answer, in that if players are .1 apart, you can say that one is better than the other, but not .1 better.

I think the standard deviation is the way to go, because it rewards a player for having an elite skill and/or punishes a player for being particularly bad at a skill.  Simultaneously, it gets rid of the differences in scale between the underlying statistics.

[AB_Celtic]

Personally, I'd find the mean and standard deviation for all three stats, either amongst point guards, guards, or all nba players (with enough minutes to count.)  If someone  was right at the mean for a stat, they get a score of 0 for that statistic.  If they're one standard deviation above, a 1, and one below, a -1, and so on.  Then add them together.  Or weight them if you do think assists should be more important, and then add them (multiplying doesn't work, however, because of the negatives.)

Yep, that'd be the way to do it.

And since I love you guys, (and since I'm bored at work) I'm doing that now.

[Finkelskyhook]

Who would you rather have on your team?

I'd take Rondo on my team over any of his peers.

[2short]

Who would you rather have on your team?

I'd take Rondo on my team over any of his peers.

that to me is the statement
ranking i'd go
1.paul
2.rondo

i am not sure i'd swap them, rondo has shown it all when it matters and is a leader

I don't think people are looking at our 6?th year pg historically.  By the time his contract is up with the team (deal by the way) his number is set for rafters. 

[saltlover]

Personally, I'd find the mean and standard deviation for all three stats, either amongst point guards, guards, or all nba players (with enough minutes to count.)  If someone  was right at the mean for a stat, they get a score of 0 for that statistic.  If they're one standard deviation above, a 1, and one below, a -1, and so on.  Then add them together.  Or weight them if you do think assists should be more important, and then add them (multiplying doesn't work, however, because of the negatives.)

Yep, that'd be the way to do it.

And since I love you guys, (and since I'm bored at work) I'm doing that now.

That's the way I did it, with TS% instead of field goal%, and it comes out:

1. Paul
2. Nash
3. Calderon (who's quite underrated, at least when looking at assists, turnovers, and shooting)
4. Rondo
5. Ginobli (because he's an amazing shooter who gets enough assists and few turnovers to score well.)
6. Parker
7. Lawson
8. Bonner (because he turned the ball over only 14 times last year)
9. Rose
10. Conley
11. Williams
...
...
...
53. Westbrook

I really like how 3 of the top 8 players are Spurs.  It's like Popovich already computed this stat.  Obviously, I did this ranking by not filtering out non-PG's, although I did get rid of players who didn't shoot enough to qualify for the FG% leaderboard.

TS% kills Rondo, who has a 48%, while the mean is 53%, putting him a full standard deviation below.  But if you think that Rondo is a better PG than Westbrook, this method is your best friend (which also has Paul Pierce above Westbrook.)

[AB_Celtic]

Calculated from 55 qualified NBA guards, courtesy of NBA.com.

Average FG%: .438
STDEV FG%: .030

Average APG: 4.565
STDEV APG: 2.394

Average ATTR: 2.393
STDEV ATTR: .624

Method: Subtract average from player's actual value, and divide by STDEV to obtain value. Add values for each of the three categories.

FG%, APG, ATTR
Rondo: +0.33, +2.98, +1.31
Paul: +1.33, +1.89, +3.18
Nash: +3.13, +2.56, +0.81
Rose: -0.10, +1.39, +0.30
Parker: +1.4, +1.07, +1.02
Williams: -1.03, +1.73, -0.31
Westbrook: +0.63, +0.39, -1.40

Final Rankings (Sum)
1. Steve Nash +6.5
2. Chris Paul +6.4
3. Rajon Rondo +4.62
4. Tony Parker +3.49
5. Derrick Rose +1.59
6. Deron Williams +0.39
7. Russell Westbrook -0.41

[dark_lord]

i would put rondo in the top 5 or 6 pgs, but i dont think he is the best

[saltlover]

Calculated from 55 qualified NBA guards, courtesy of NBA.com.

Average FG%: .438
STDEV FG%: .030

Average APG: 4.565
STDEV APG: 2.394

Average ATTR: 2.393
STDEV ATTR: .624

Method: Subtract average from player's actual value, and divide by STDEV to obtain value. Add values for each of the three categories.

FG%, APG, ATTR
Rondo: +0.33, +2.98, +1.31
Paul: +1.33, +1.89, +3.18
Nash: +3.13, +2.56, +0.81
Rose: -0.10, +1.39, +0.30
Parker: +1.4, +1.07, +1.02
Williams: -1.03, +1.73, -0.31
Westbrook: +0.63, +0.39, -1.40

Final Rankings (Sum)
1. Steve Nash +6.5
2. Chris Paul +6.4
3. Rajon Rondo +4.62
4. Tony Parker +3.49
5. Derrick Rose +1.59
6. Deron Williams +0.39
7. Russell Westbrook -0.41

No Calderon?

[AB_Celtic]

Calculated from 55 qualified NBA guards, courtesy of NBA.com.

Average FG%: .438
STDEV FG%: .030

Average APG: 4.565
STDEV APG: 2.394

Average ATTR: 2.393
STDEV ATTR: .624

Method: Subtract average from player's actual value, and divide by STDEV to obtain value. Add values for each of the three categories.

FG%, APG, ATTR
Rondo: +0.33, +2.98, +1.31
Paul: +1.33, +1.89, +3.18
Nash: +3.13, +2.56, +0.81
Rose: -0.10, +1.39, +0.30
Parker: +1.4, +1.07, +1.02
Williams: -1.03, +1.73, -0.31
Westbrook: +0.63, +0.39, -1.40

Final Rankings (Sum)
1. Steve Nash +6.5
2. Chris Paul +6.4
3. Rajon Rondo +4.62
4. Tony Parker +3.49
5. Derrick Rose +1.59
6. Deron Williams +0.39
7. Russell Westbrook -0.41

No Calderon?

No he's there. I just calculated the 7 we were talking about.

Calderon +5.78