Introduction In a <a href="http://bbs.clutchfans.net/showthread.php?p=2417689">previous thread</a>, I reported on the possible discovery of a rule that seemed to be unbreakable in the NBA world. Some of my data was trashed by the NBA year-numbering problem, so I am starting over here in order to avoid confusion. I am also taking the opportunity to expand my analysis to almost the entire history of the NBA, all the way back to 1951; it seems the rule has never been broken! I will go through my arguments again, hopefully with more clarity this time. I will also talk a bit about how this rule applies to the Rockets. And because this message is way too long as is, the data will appear in followup messages. The Rule The unbreakable rule is this: you must have at least one significant rookie in a 3-year period if you want to win a championship in the 4th year (unless you are in the middle of a dynasty). This rule has never been broken in the history of the NBA. The converse is equally important: no team that has ever gone through a 3-year drought of significant rookies has ever won a championship in the 4th year (except for those teams that were already in the middle of a dynasty). Definitions -- A "significant rookie" is someone who plays at least 1000 minutes in his rookie year. -- A team has a "dynasty" going if the maximum gap between championships is two consecutive non-championship years. The two-year maximum gap makes sense for Michael Jordan's absence from the NBA in the middle of the Bulls dynasty, and also applies nicely to the great Celtics dynasty of the 1960s. -- A team is in a "drought" if it has had no significant rookies for 3 years or more, unless that team is already in the middle of a dynasty. Discussion It is notable that the rule has never been broken in the 57 years for which the NBA has existed. We can rule out chance as the explanation. That the converse holds is just as important. There have been 40 droughts in the history of the NBA, and not once -- repeat, not once -- has the converse ever been false. Again, this is far beyond any reasonable chance. When you combine the rule and the converse, you have a very strong "if and only if" relationship: you win a championship if and only if you managed to load up on one or more significant rookies in the previous 3 years (unless you are in the middle of a dynasty). The rule makes sense, because the development of a significant rookie speaks well of the GM for his astuteness in finding talent, and of the coach for his patience and teaching ability. Any team with a management duo like that is well positioned to win it all. The "middle of dynasty" exception is also sensible: a team that has just won a championship looks pretty set. Its rookies, however good they may be, are unlikely to get a whole lot of playing time and therefore won't pass my test for significance. Implications for the Rockets When the 2006-7 season starts, it will have been 4 years since we drafted Yao in 2002. Luther Head played over 2000 minutes in 2005-6, so he might qualify as a significant rookie; if he does, we are safe. However, some people object to this, because Luther played so much only because the team suffered an incredible rash of injuries. JVG himself has said as much. If the objectors are correct, we are past the 3-year limit and in serious danger of missing the championship train.
Now for the data. Here are the significant rookies picked up by each of the championship teams. Notice that every championship team or dynasty, without exception, managed to load up on one or more significant rookies in the 3 years before the dynasty began. A word about notation. I have adjusted all championship years to indicate the calendar year in which that season began. Thus Magic Johnson was drafted in 1979, and his first championship year is also listed here as 1979, despite the fact that it actually happened in the spring of 1980. I have only covered the NBA; I am too tired to include the ABA, though a quick look indicates that the 3-year rule should also apply there. <dl> <dt>Minneapolis Lakers: 1949 dynasty (1949 1951 1952 1953)</dt><dd>Clyde Lovellette (1953)</dd> <dt>Rochester Royals: 1950</dt><dd>The franchise started in 1948, so the 3-year rule does not apply here.</dd> <dt>Syracuse Nationals: 1954</dt><dd>George King (1951), Wally Osterkorn (1951), Bill Kenville (1953), Johnny Kerr (1954), Dick Farley (1954)</dd> <dt>Philadelphia Warriors: 1955</dt><dd>Mark Workman (1952), Danny Finn (1952), Jack George (1953), Walt Davis (1953), George Dempsey (1954), Tom Gola (1955)</dd> <dt>Boston Celtics: 1956 dynasty (1956 1958 1959 1960 1961 1962 1963 1964 1965 1967 1968)</dt><dd>Frank Ramsey (1954), Dwight Morrison (1954), Jim Loscutoff (1955), Dick Hemric (1955), Tom Heinsohn (1956), Bill Russell (1956), Thomas Sanders (1960), John Havlicek (1962)</dd> <dt>St. Louis Spirits: 1957</dt><dd>Jack Stephens (1955), Al Ferrari (1955), Win Wilfong (1957)</dd> <dt>Philadelphia 76ers: 1966</dt><dd>Lucious Jackson (1964), Billy Cunningham (1965)</dd> <dt>New York Knicks: 1969 dynasty (1969 1972)</dt><dd>Cazzie Russell (1966), Walt Frazier (1967), Phil Jackson (1967), Dean Meminger (1971)</dd> <dt>Milwaukee Bucks: 1970</dt><dd>Greg Smith (1968), Dick Cunningham (1968), Kareem Abdul-jabbar (1969), Bob Dandridge (1969)</dd> <dt>Los Angeles Lakers: 1971</dt><dd>Bill Hewitt (1968), Eldo Garrett (1969), Rick Roberson (1969), Jim Mcmillian (1970)</dd> <dt>Boston Celtics: 1973 dynasty (1973 1975)</dt><dd>Dave Cowens (1970)</dd> <dt>Golden State Warriors: 1974</dt><dd>Jamaal Wilkes (1974), Phil Smith (1974)</dd> <dt>Portland Trailblazers: 1976</dt><dd>Bernie Fryer (1973), Bill Walton (1974), Lionel Hollins (1975), Bob Gross (1975), Johnny Davis (1976), Robin Jones (1976)</dd> <dt>Washington Bullets: 1977</dt><dd>Mitch Kupchak (1976), Larry Wright (1976)</dd> <dt>Seattle Supersonics: 1978</dt><dd>Dennis Johnson (1976), Bob Wilkerson (1976), Jack Sikma (1977)</dd> <dt>Los Angeles Lakers: 1979 dynasty (1979 1981 1984 1986 1987)</dt><dd>Tom Abernethy (1976), Earl Tatum (1976), Norm Nixon (1977), Magic Johnson (1979), Kurt Rambis (1981), James Worthy (1982), Byron Scott (1983), A.c. Green (1985)</dd> <dt>Boston Celtics: 1980 dynasty (1980 1983 1985)</dt><dd>Cedric Maxwell (1977), Jeff Judkins (1978), Larry Bird (1979), Gerald Henderson (1979), Kevin Mchale (1980)</dd> <dt>Philadelphia 76ers: 1982</dt><dd>Andrew Toney (1980), Marc Iavaroni (1982)</dd> <dt>Detroit Pistons: 1988 dynasty (1988 1989)</dt><dd>Joe Dumars (1985), John Salley (1986), Dennis Rodman (1986)</dd> <dt>Chicago Bulls: 1990 dynasty (1990 1991 1992 1995 1996 1997)</dt><dd>Horace Grant (1987), Scottie Pippen (1987), Stacey King (1989), B.j. Armstrong (1989), Toni Kukoc (1993)</dd> <dt>Houston Rockets: 1993 dynasty (1993 1994)</dt><dd>Robert Horry (1992), Sam Cassell (1993)</dd> <dt>San Antonio Spurs: 1998</dt><dd>Tim Duncan (1997)</dd> <dt>Los Angeles Lakers: 1999 dynasty (1999 2000 2001)</dt><dd>Travis Knight (1996), Kobe Bryant (1996)</dd> <dt>San Antonio Spurs: 2002 dynasty (2002 2004)</dt><dd>Tony Parker (2001), Charles Smith (2001), Emmanuel Ginobili (2002), Beno Udrih (2004)</dd> <dt>Detroit Pistons: 2003</dt><dd>Mateen Cleaves (2000), Zeljko Rebraca (2001), Mehmet Okur (2002)</dd> <dt>Miami Heat: 2005</dt><dd>Caron Butler (2002), Rasual Butler (2002), Dwyane Wade (2003), Udonis Haslem (2003)</dd> </dl> Number of years: 57 Number of lone championships or dynasties: 26
Here is the data to support the converse: that no team going through a 3-year drought of significant rookies has ever won a championship in the 4th year (unless that team was in the middle of a dynasty). The data supports the conclusion unanimously. I am again using the "draft year" convention everywhere for year numbers. <table cellpadding=2 cellspacing=7> <tr><td>Atlanta Hawks</td><td>1986, 5</td><td>1996, 3</td></tr> <tr><td>Boston Celtics</td><td>1949, 5</td></tr> <tr><td>Charlotte Hornets</td><td>1993, 6</td></tr> <tr><td>Chicago Bulls</td><td>1972, 3</td></tr> <tr><td>Cleveland Cavaliers</td><td>1975, 3</td></tr> <tr><td>Dallas Mavericks</td><td>1987, 4</td><td>1996, 7</td></tr> <tr><td>Denver Rockets</td><td>1983, 3</td></tr> <tr><td>Detroit Pistons</td><td>1973, 3</td><td>1982, 3</td><td>1996, 4</td></tr> <tr><td>Fort Wayne Zollner Pistons</td><td>1949, 3</td></tr> <tr><td>Golden State Warriors</td><td>1971, 3</td></tr> <tr><td>Houston Rockets</td><td>1985, 3</td><td>1989, 3</td></tr> <tr><td>Indiana Pacers</td><td>1997, 4</td><td>2002, 3</td></tr> <tr><td>Los Angeles Clippers</td><td>1991, 3</td></tr> <tr><td>Miami Heat</td><td>1996, 3</td></tr> <tr><td>Milwaukee Bucks</td><td>1983, 9</td><td>1997, 6</td></tr> <tr><td>Minnesota Timberwolves</td><td>2000, 5</td></tr> <tr><td>New York Knicks</td><td>1949, 6</td><td>1992, 12</td></tr> <tr><td>Orlando Magic</td><td>1994, 4</td></tr> <tr><td>Philadelphia Warriors</td><td>1949, 3</td></tr> <tr><td>Philadelphia 76ers</td><td>1989, 3</td><td>1998, 3</td></tr> <tr><td>Phoenix Suns</td><td>1989, 3</td><td>1996, 3</td></tr> <tr><td>Portland Trailblazers</td><td>1986, 3</td><td>1990, 5</td><td>1998, 6</td></tr> <tr><td>Sacramento Kings</td><td>2001, 4</td></tr> <tr><td>San Antonio Spurs</td><td>1976, 3</td><td>1993, 4</td></tr> <tr><td>Seattle Supersonics</td><td>1991, 9</td></tr> <tr><td>Utah Jazz</td><td>1986, 3</td><td>1997, 4</td></tr> </table> This all the data I have. Feel free to fire away!
By the way, I could not find the "minutes played" stat for individual players until 1951. This is why my analysis begins on that year, even though the NBA started in 1949.
So, if your championship year is, say, 2006-2007, that means you have to have a significant rookie in 2003-2004, 2004-2005, or 2005-2006... or can that significant rookie come in the year of 2006-2007, the championship year itself? Seems like that latter is the case, since Sam Cassell counted as a significant rookie according to you for the 1993 Championship team. So you really just need 1 rookie playing 1000 minutes in one of the last 4 seasons to qualify for this. Right? Just want to be clear. Also, two of the Piston's "significant rookies" were Cleaves and Rebraca.. Are you serioiusly arguing that without these two, they don't win the championships? Even Okur only averaged 3.7 ppg in the championship playoffs. It seems "insignificant rookie" Tayshaun Prince was far more important. As for whether Luther Head counts or not... if he does not, are you going to go back and delete every "signficiant rookie" who only got enough playing time because of either injury or the fact that the team sucked at their position? If you go back and do that, does your theory still work? In the end, what happens when you go back and examine "significant trade acquisition" or "significant FA contributor"? Have teams won without getting oen of those in the previous 3 years?
Good Analysis. I was about to reply that the first 2 championships you cited (above) did not adhere to your 3-yr rule (i.e. Lovellete in '53 could not account for championships in 49, 51 and 52), but then I saw that you had no data available prior to 1951. (one assumes that Minn had significant rookies between 46-49 to account for the '50 championship, likewise in 48, 49 or 50 for the Royals.)
I would point out that correlation does not necessarily equal causation. So I would question the "unbreakable rule" connotation. Interesting analysis and discussion though.
now you need to do a statistical analysis of all NBA teams seeing what the average number of teams that have a significant rookie in the previous 4 years, them compare that data to the data you already have, because i have a feeling that most NBA teams have a player drafted in the previous 4 years playing somewhere in the rotation.
Everything you said was great. And props on the extensive research and analysis. You had everything perfect- until your last sentence. Everything can be broken- the Red Sox curse, "no team has ever won the middle 3 games of the NBA Finals at home," Roger Maris' record, etc. If I'm an NBA GM, am I really going to say, "Well, better make sure I have that rookie in order to win the championship?" My own basic theory, very simplified, says that you have to have a dominant superstar and only 1 "weak" link in your starting 5 in order to win a championship- or at least put yourself in the position to do so. If you have more than I weak link, you have to have 2 dominating superstars and a fairly decent 3rd player who is a specialist in a key area (e.g. Rice with the Lakers, Rodman with the Bulls). This rule is, of course, subject to be broken (2004 Detroit Pistons), but probably not often. I think having a player who is fearless in the most crucial of moments is a critical element for a championship team. With only 5 players on the floor at a time, one player has much more of an impact than in other sports. I think that, as long as our 2 main guys are healthy, in one of the next 5 years, we're going to have the right combination of players and luck to win a championship.
Props on the extensive analysis, but I think the point about causation is important. One shouldn't take this theory to mean that you should favor rookies over veterans. Sure, we could trade Yao for Brandon Roy, who is almost sure to be a significant rookie, hence giving us the magic significant rookie in the last 3 years. I see good rookies as something of a matter of luck (though coaching and training certainly help them develop), especially with reference to the salary cap. You can only get so much veteran talent and still stay under the cap. When you get lucky with a rookie, you're getting a lot more than what you pay for.
Nice thread and analysis terse. What I find remarkable about it is that the advent of free agency still hasn't broken the rule. On the other hand I think it's always important to find and develop young talent. Head can play and Yao was an overall #1 which really doesn't say much how we've been able to develop young players in the las 5 years. Food for thought.
...whether they won a title or not. Along with the "correlation does not equal causation" I think you're not quite there with your theory.
terse, I mentioned this in your other thread...but you're STILL doing your analysis completely from one side. You have data to back up these facts: 1. All championship teams have had rookies/newcomers play signficant minutes 2. No teams that have had "3-year droughts" ever won the championship the fourth year However, these two facts are NOT converses of each other. There's no "if and only if" conclusion to be drawn here. The true converse of the first statement is: All non-championship teams have not had rookies/newcomers play signficant minutes. In the same manner, the converse of #2 is: Teams that did not have a "3-year drought" proceeded to win the championship in the fourth year. Just like before...although your statistical analysis shows that all championship teams have had good rookies, it doesn't conclusively demonstrate that these rookies are a prequisite to winning a championship.
Yes, the 3-year rule includes the championship year itself. So if Novak or JL3 pan out, and JVG doesn't adopt his usual "let 'em rot" attitute towards them, we just might win it all this year. No, I am not arguing anything about Okur. However, Rebraca was involved in the Rasheed Wallace trade, and that was indeed critical to the Pistons' championship run. See what happens when you keep and develop your rookies? They become valuable pieces that you can trade for serious talent. If you do what the Knicks did for 12 clueless years, you will win zero championships.
I am willing to try it, but you will have to find the database that lists the reasons why each rookie got his playing time. Feel free to do it yourself.
I am only pointing out the correlation. I discussed a reasonable explanation for it (GM's astuteness), but I don't really care -- I am only pushing the correlation. I am also suggesting that the correlation probably has predictive value because it has been so incredibly strong for the entire lifetime of the NBA.
Well.. Rebraca wasn't just another salar to match the contract... He hasn't "developed" (hell, he was in his 30s by then and played less minutes each year. ATL really didn't want him. They only played him 51 total minutes for the remainder of the season and let him go as a FA as soon as they could. Not to say the Pistons don't know how to draft-- Tayshaun is a great pick, though not "significant" under your formula. Darko, as we've seen in ORL, is not a bad player, but just never got the court time and lost his desire in DET. Look, I agree it's important to make the most out of your draft (or luck into something good at least), but this in no way says trading away picks is a bad thing. Spurs have traded away a lot of recent first rounders, and they are a first class team. As for the Knicks... first, as I said, their problem was that they drafted a lot of busts who were not worth anything in NY or anywhere else. The fact that they didn't leave the bench as rookies or they got traded really didnt' hurt the team... the fact that the Knicks wasted picks on them in June did. Tayshaun didnt' get 1000 minutes as a rook, Ben Wallace didn't even get 200, Mike James barely got 100 minutes, Charlie Ward got a grant total of 44... all these guys became at good NBA players or better. If you get the right players and keep them around, then the talent and desire will rise to the top eventually.
Well, I am not really pushing the "causation" aspect. The correlation stands so strongly on its own that it's probably predictive. I think you need to strike a balance between rookies and vets. You're right about Brandon Roy. We would have gotten him if it weren't for our two unnecessary wins near the end of the season. JVG's selfish stat padding dropped us one or two spots in the draft, and that cost us Brandon. Some luck is helpful, for sure. But take a look at Bill Russel's Celltics dynasty, the Showtime Lakers, and Tim Duncan's Spurs: they kept loading up with good rookies even as they continued to win championships. These teams probably had the worst drafting positions imaginable, yet they continued to draft well. (Manu Ginobili was a 2nd rounder, and Tony Parker almost was!) To me, that suggests excellent management.
