I wouldn't disagree if you felt that Marbury was a better point guard at age 21, but if you do feel that way, the difference is very small. Davis was a backup that year and Francis, as noted, was in college, so Arenas has a pretty good head start on being as good as the top young point guards in the NBA.

"I want you to name a point guard at 20 years of age that puts up numbers that were better."
Well, Arenas turned 21 last season, so that's what age I consider him to have been. Still, few players have been more successful at that age than Arenas - not all that many are even in the NBA at 21, let alone starting, let alone playing well as a starter. I've sorted the top age-21 seasons (based on age as of the end of the regular season, April 17) by my Value Over Replacement Player calculation. Where does Arenas rank?

Name		       VORP    RelEff	Year	
Tracy McGrady		349	126.3	2001
Shaquille O'Neal	339	124.3	1993
Tim Duncan		328	123.2	1998
Shareef Abdur-Rahim	288	122.2	1998
Kevin Garnett		278	119.6	1998
Clark Kellogg		278	122.9	1983
Isiah Thomas		262	119.3	1983
Chris Webber		255	123.0	1994
Pau Gasol		254	119.4	2002
Kobe Bryant		245	122.1	2000
Elton Brand		244	118.5	2000
Paul Pierce		243* 	121.4	1999
Rashard Lewis		230	119.7	2001
Bernard King		226	116.8	1978
Gilbert Arenas		226	118.3	2003

*Pierce's season pro-rated to 82 games

That's thirteen players - one of them, Hall of Famer Isiah Thomas, a point guard. In fitting with the proliferation of young players in the NBA, just three played before 1990, with five (not including Arenas) in the 'oughts' so far. It should strike you looking at these players that most of them have been highly successful NBA players. Most of the players on this list have achieved superstardom. The only exception amongst the veteran players is Clark Kellogg, whose career was destroyed during his fourth year in the NBA by a knee injury. Kellogg, Pau Gasol, and Rashard Lewis are the only players on this list not to have made an All-Star team⁵ (how Kellogg didn't make the All-Star team as a rookie while scoring 20 points and grabbing 10 rebounds per game is utterly beyond me), and the latter two have several years left in that pursuit, though being Western Conference forwards won't help them for the next decade or so.

(As a side note, conspicuous in his absence is Magic Johnson, one of the greatest early entrants of all time. Per-minute, Johnson would rank at the top of this list, but torn cartilage knocked him out for 45 games. He still makes a second 21-year-old point guard better than Arenas.)

While I disagree with those who think age is the single biggest indicator of a player's future success - performance is – it certainly is extremely important. A player who is very successful in the NBA at a young age, as Arenas has been, is highly likely to become an All-Star caliber player. You may disagree with my method for rating players, or my opinion of Arenas' success so far, but that much you cannot doubt. Successful young players go on to be very successful veterans.

Let's try a study to see just how important this is. Honestly, I am doing this study just as I type this article.⁶ What I'm doing is separating all rookies who have had moderate success - defined as 100 VORP or better. We'll compare how much players had improved four years later, looking at age. There are 121 players who met my criteria whose rookie season fell between 1979 and 1999 (can't be later because those players haven't reached their fifth seasons). 115 played five years in the NBA. I used Relative Efficiency (explained above) to compare the players so that non-permanent injuries/minute changes didn't come into play. Here's the table, broken down strictly by age:

Age	#	 Yr1	 Yr5	Change
19	1	110.7	124.0	 12.0
20	6	115.3	118.5	  2.7
21	12	116.8	121.8	  4.3
22	50	116.2	114.4	 -1.6
23	37	115.7	113.7	 -1.8
24	8	117.3	117.2	 -0.1
25	2	117.2	110.9	 -5.4
26	3	116.1	100.2	-13.7
30	1	120.5	103.5	-14.1
31	1	131.5	120.9	 -8.1

At this point, this doesn't necessarily tell us a great deal. While the age-19 players look impressive, they're not as much so when you consider that Kevin Garnett was the only such player. The next step is to aggregate them into "young" (19-21), "normal" (22-23), and "old" (24-31):

	 #	 Yr1	 Yr5    Change
Young	19	116.0	120.8	 4.2
Normal	87	116.0	114.1	-1.7
Old	15	118.2	114.1	-3.4

It now becomes apparent that age plays a huge role in determining these players' development. While the young players had plenty of room to grow on average, neither the 22-23-year-olds nor the older players could even sustain their success on average. Of course, many of these players did in fact grow significantly, but the important issue is the average. The difference is most apparent at the extremes. The 19-21 list reads like a Who's Who list of All-Stars. Of the 19 players, only four - Joe Smith, Benoit Benjamin, Wayman Tisdale, and Kellogg - never played in an All-Star Game. On the other hand, the players 25 and over had significantly less success. Two - Dino Radja and Abdul Jeelani - were out of the league within five years. Dean Garrett was a one-year flameout, and while Sam Mitchell had a lengthy and successful career, he only matched his rookie success once, at age 35. The two most successful old rookies were foreigners - Dikembe Mutombo, Arvydas Sabonis, and Sarunas Marciulionis (though the latter's career was basically over by his fifth year as well).

Arenas doesn't actually even make this list - he played too little as a rookie - but the general point is very applicable to him. He has room to build on his early success.

If this discussion of player development has done nothing to convince you of Arenas' value, my best argument comes in the form of a comparison. The two players being compared put up the following primary statistics last year:

Name	    PPG   APG   RPG   SPG  TOPG   FG%
Player A   18.3   6.7   4.3   1.5   3.5  .431
Player B   21.0   6.2   6.2   1.7   3.7  .435

Pretty similar. The similarity is even more apparent when we look at secondary statistics. These are some of the ones I use:

Name	    P48  ScEff  Score  Pass   Pos/m  R48   Eff
Player A   25.0  .540    2.82  3.18    0.58  6.5  .509
Player B   24.6  .541    2.78  2.54    0.56  7.2  .513

(By way of explanation, ScEff is PTS/(2*fga+.88*fta), Score is ScEff*PTS/MIN*10, Pass is AST/TO*AST/MIN*10, Pos/m is possessions per minute, (fga+.44*fta+to)/min.)

Now that's awfully damn similar. Player B is a better rebounder, but Player A is a better passer. By my statistics, they appear to be basically identical in terms of scoring despite Player B's 2.7 ppg advantage. Obviously, Player A didn't play as much as Player B. As a final basis of comparison, three more categories in which these players are basically identical, and one very important one in which they are not:

Name	    Pos   Ht   Wt   Birthday
Player A    PG    75  191   1/06/82
Player B    PG    75  195   2/21/77*
*Just for the record, Player B was listed with a 
birthday of 2/21/78 entering the NBA Draft. It
since appears to have changed to 2/21/77. I'm not
exactly sure which is correct.

So what do we know about these players? They are both point guards, even though their numbers may more closely replicate those of shooting guards. They are not great passers, but they are very good scorers and rebounders and very good players overall. However, one of them can be expected to improve his game a lot more in the coming seasons than the other, who at age 26 is already nearing his peak.

Obviously, Player A is Arenas. If you follow the NBA, you should have a pretty good idea of who Player B is - Houston's Steve Francis. Really, I'm not sure there are two players in the league more closely matched than Arenas and Francis. I'm working on a system to determine player similarity that mostly uses the second set of figures, along with some others, and Francis easily comes out as Arenas' most comparable player, with a similarity score of 99.64. The next best is Sam Cassell, at 98.50. Vice versa, Arenas is Francis' best comparable (though Bobby Jackson and Chauncey Billups are in the ballpark).

Last year, as you probably recall, Francis was eligible for a contract extension. The Rockets gave him the maximum deal possible, $85 million over six years. That's an average of over $14 million per season, starting next year, far more than Fegan said Arenas was worth. Did anyone so much as bat an eye at this decision? Okay, John Hollinger of AlleyOop.com did say "I might have wanted to see him play migraine-free for a year before I committed this kind of money, but Francis is clearly a stud when healthy." Was there anyone else who questioned that decision at all? Actually, I did for News@Hoopsworld, but that's neither here nor there.⁷

Let's take a look at this:
1. Gilbert Arenas and Steve Francis were very similar players last year
2. Gilbert Arenas was five years younger (maybe)
3. Steve Francis is judged to be worth $14 million a year over the next six years by most everyone.
4. Shouldn't Gilbert Arenas be worth a similar amount?

Really, most of the criticisms of Arenas after Fegan's demand are true of Francis as well:
Not a true point guard – check
Hasn't played for a playoff team - check (Francis' teams have been closer, but Arenas' was a lot worse before he broke into the starting five)
Immature at times – check

The only thing that doesn't fit is that Arenas hasn't "proved himself", though I think that's a complete crock anyway. Arenas has been terrific when he has gotten the chance to play; it's not his fault the Warriors decided to ignore him for the first four months of his rookie season. We can't be certain that Arenas is going to develop into an All-Star point guard (or shooting guard, if eventually he's moved back to that position). We couldn't be certain that Francis' migraines weren't going to derail his path to superstardom. We can never be certain. All available evidence, however, points to Arenas becoming a very good point guard.

In the end, my conclusion is simple. If I were the general manager of an NBA team that needed a point guard and had $7 million in salary-cap space, I would be happy to sign Arenas for that amount. Maybe you wouldn't. The evidence suggests that, for whatever warts he has, Arenas is going to become a historically good player. If the alternative is signing someone over-hyped or over the hill, I'll take Arenas, thank you very much.⁸

2. Here's what I wrote on my Sonics message board: "I like Gilbert Arenas. While everyone talks about Kidd as a free agent point guard next year, I think that kid is going to be a great pickup for somebody ... ." Okay, maybe not as strong as I remembered it, but keep in mind people were talking about Steve Logan starting over Arenas in the summer of 2002. And if you're saying, "Who?" that's exactly the point.

3. He did this again in Washington. Call me crazy, but I still think it's pretty funny. I probably wouldn't feel that way if he played for the Sonics.

4. You can't read this article online anymore, but here's a Usenet post that repeats it.

5. Lewis made it in 2005. Gasol's got a pretty good shot at doing so in 2006, thanks both to playing more minutes to pad his traditional statistics and a scraggly-looking beard that has made him seem a thousand times tougher.

6. I honestly did. Believe me. Though I will tell you if it didn't back up my point, it might not have seen the light of day.

7. Here's what I wrote: "Francis is merely adequate as a distributor, below-average from three-point range, and his lackluster defense was a significant reason why the Rockets were one of the league’s worst teams on that end of the court."

But I then followed with: "Francis is without question one of the brightest young stars in the NBA and the piece around which the Rockets are rebuilding. Not extending him is almost unimaginable; that doesn’t mean it wasn’t worth discussing at least a little."

In hindsight, I was too kind to Francis and everyone else way too kind. Francis clashed a bit with new Rockets coach Jeff Van Gundy and was traded to Orlando as part of a package for Tracy McGrady in the summer of 2004. Anyone suggesting at this point that Francis is better than Arenas would be laughed out of the room in less time than it took you to read this sentence.

8. So let's talk about what eventually happened. Arenas signed with the Washington Wizards for a deal starting north of $7 million. He then had one of his "Gilbert being Gilbert" moments in explaining the deal, saying he went with the Wizards over the L.A. Clippers because he flipped a coin 10 times and it landed on the Clippers eight, but he wanted to go against the odds with the Wizards. (Whether this story is true is debatable at best.)

Arenas did not re-sign with the Warriors in large part because he would have had to take a smaller salary for at least one season before Golden State earned his "Bird Rights" and was able to go over the cap to re-sign him to a deal like the one offered to him by the Wizards. In the 2005 CBA negotiations, the league created what is known informally as "the Gilbert Arenas Rule," which provides that second-round picks who become free agents after their first or second season cannot earn more than the mid-level exception in the first two years of their new deal, allowing their old team to match any offers made to them.