This is funny.:grin: It's only about 5-6 seconds long but you'll hear it. (sorry about the commercial/sorry if posted) http://youtu.be/EbyKeIv766g?t=4m1s
The Rockets seemingly chose to forgo the bolded big 3 when they apparently passed on Josh Smith. By all advanced metrics Josh Smith is a great defender and rebounder, although flawed offense. To me analyzing big 3s is kind of pointless. One, stars are so rarely available you really have to grab whoever's available. And two, adding a star to 2 top-10 players in Harden and Howard makes the Rockets championship caliber regardless of fit. Go through the 2013 all-stars. Ignore the centers on the team and not-truly-all-stars like Lee and Deng, you can't find a single player that wouldn't look awesome on this team.
I would definitely see Asik as being in Cluster 10 under this methodology even though he is a center. I wonder if the strength of the 8-10-12 Big 3 comes from the ability to have either the 10 or the 12 on the court at all times. Edit. I wasn't looking at the 8-12 coefficient. The average 8-12 coefficient is better than the 8-10-12 coefficient so the third star to add to Harden-Howard would most likely be something besides a 10. There just isn't enough data to say what that is besides a cluster 7.
The cluster analysis appears to be based on how well a fit would improve, decrease the expected production of a team. Player quality does matter. For instance, jtr did an analysis to predict what the Rockets net point differential would be based on player ORTG and DRTG. For arguments sake, assume the coefficients were calculated using the same methodology as jtr did, one would then add the 7.5 (Big 2 8-12) wins to the 53 wins to say the Rockets are actually expected to win 60 games. The quality of player would affect the projected win total before the coefficient is added. The third best player being a Cluster 10 may actually be considered a detriment as the 8-10-12 coefficient is less than the 8-12 coefficient.
I don't think it's pointless, I think the point that can be drawn from the research is that if you could choose, you'd like your star players to be as least redundant as possible. Having a star 1, 3, and 5 would yeild the most balance roster of star power. Does that mean you'd not pursue a star 2 just because you have a star 1? No, like you said you take whatever star you can get and build around that. But if you could you'd like to invest your cap room in positions that don't overlap.
For example, suppose the Asik-Howard experiment does not work out, and the Rockets are shopping for a power forward later this year (or seeking to develop one internally). According to this study, based on historical data, they would be better off going after a defense oriented power forward rather than a well rounded one. That , to me, is a kind of surprising and interesting data point. Cluster 10: Limited 4’s; very strong rebounders, defense oriented: Dennis Rodman, Ben Wallace, Buck Williamsan Cluster 5-: Dynamic, well-rounded power forwards, strong rebounding, dynamic 3’s: Chris Webber, Pau Gasol, Kevin McHale
The paper is really less about Big 3s than it is about "fit". I think you'd like it. Here's an excerpt, as well as links to the paper, and an ESPN TrueHoop article about the paper. http://www.sloansportsconference.co...rt-Ayer-Sloan-Sports-Analytics-Conference.pdf http://espn.go.com/blog/truehoop/post/_/id/38016/you-me-synergy-the-big-2-and-the-big-3 "There have been many instances in basketball where a team, with perhaps a collection of new acquisitions, underperforms relative to the perceived talent on the team. Most observers will intuitively conclude that this team, while talented, just doesn’t fit well together. Conversely, there are also many instances where a team, with perhaps relatively modest top level talent, exceeds expectations. This team, most will conclude, is put together well, i.e., the players complement each other, they “fit.” Similarly, at the player level, there are many cases in which a new player on a team, perhaps acquired through free agency, who, though talented, fails to live up to expectations. This can be attributed to many things: lack of effort, erosion of skills, poor scouting. However, many times a player’s underperformance is attributed to a poor fit with the team. This is to say, that it is not enough for a player to have valuable basketball skills to fully reach his potential; the player must also be on a team which is constructed in a way that is complementary to those skills. This paper aims to provide some insight into team construction and player fit by analyzing combinations of player types (specifically, 2 and 3 man combinations, known commonly as “Big 2’s or Big 3’s), and determining which combinations lead to increased wins, while accounting for talent level and coaching skill."
For me, the paper is interesting because the Rockets potentially do so well under its analysis! And because it may gives insight into what D Morey is thinking (it was presented at his conference). The paper says the two best core player combinations are (1) Cluster-8: Multi-faceted, high scoring wings, with high assists for their position and are great 3 point shooters: Paul Pierce, Danny Ainge Cluster 12: High scoring post players, high rebounds, high blocks: Shaquille O’Neal, Hakeem Olajuwon, David Robinson Cluster-7: High scoring, high assist, high steals, high turnover point guards, who don’t shoot 3s: Kevin Johnson, Isaiah Thomas (2) Cluster-8: Multi-faceted, high scoring wings, with high assists for their position and are great 3 point shooters: Paul Pierce, Danny Ainge Cluster 12: High scoring post players, high rebounds, high blocks: Shaquille O’Neal, Hakeem Olajuwon, David Robinson Cluster 10: Limited 4’s; very strong rebounders, defense oriented: Dennis Rodman, Ben Wallace, Buck Williamsan The Rockets come very close to having one, if not two, of those combinations in Harden-Howard-Asik Harden-Howard-Lin and Harden-Howard-Parsons
It is talking about all of those as star level players. Asik, Lin and Parsons aren't yet. We'll see what happens this year, though.
Even Hitler was TERRIBLY upset that Dwight left LA http://captiongenerator.com/7177/Hitler-reacts-to-Dwight-Howard-leaving-LA
That's the thing. Quality is the dominating factor in any dissection of championship probabilities. And star-quality players add so much value that any fit issue is so miniscule in comparison. A list of top players excluding centers and players with age/injury issues: Lebron, Durant, Westbrook, CP3, Aldridge, Love, Rose, George, Curry, Melo, Bosh, etc. Wouldn't all of them look incredibly good on this team? I agree that if it's possible, you obviously choose the better fit. But when was the last time the Rockets or really any team had their pick of multiple star players? Yeah, but I don't really care to analyze fit beyond the Rockets. And I can analyze Rockets "fit" based on our roster moves and Morey's opinions that gets leaked out. And Morey has always maintained that you get the star first and figure things out later. Not the other way around.
Agreed that when it comes to star players, fit doesn't really matter. However, this paper, despite the deceptive title of "big" 2 and "big" 3, really has nothing to do with star players. In fact half of the clusters- 1, 3, 4, 9, 10, 14, are spelled out expressly to be "limited roleplayers". The value of the paper really has to to with quantifying the synergies that are created through different combinations of player types. Most interestingly, by far and away the largest boost in wins you get is through the 7-8-12 combination, with a +13.6 compared to at best in the +5 range for all other positive combos. The Rockets have those 3 players on the team right now, in theory. If you read the description and the stat averages, Jeremy Lin is clearly a 7 and Dwight Howard is clearly a 12, both given slightly higher usage compared to last year. The question is whether James Harden is a 2 or 8. While Harden would seem to be a 2 on the surface, especially comparing his stats last year to the cluster averages, my contention is that he is actually a 8 with high usage. The key distinction between cluster 2 and 8 isn't actually the PPG difference, but rather 8 being a prolific 3PT shooter, while 2 only attempts 1 3PT a game. This is important because in an analysis of player synergies, how each player type spaces the court is crucial. Cluster 2 is describing a scoring guard who relies on isolation plays- think Kobe or MJ. Harden meanwhile excels in the flow of the offense, with 3PT shooting being a major weapon of his. If you look back at Harden's OKC averages, they fit cluster 8 to a T. I think if you decrease Harden's usage% from 29 to a healthy 25ish, increase Jeremy's slightly, and return Howard's back to his Orlando usage of also 25ish, it would fit their roles perfectly to the 7-8-12 combination. Now Jeremy isn't a star player and some may be offended by his grouping with the other 2(even though we are only talking about player types, not player quality), but the +13.6 wins this combination creates through its synergy works out to be a star player on its own. Of course I have my questions to the validity to these findings, but if the benefits are anywhere near the magnitude that the paper purports, then it is worth investigating.
Quality is the first criteria. You get a top 15-20 guy, you do it. If you can put Lebron, Wade, and Bosh together, you do it. Fit is the second. Since Lebron is on Miami, it will be tough to beat Miami with quality alone. If it comes to a decision between Rondo and Ty Lawson, Fit would be the deciding criteria for picking Lawson over Rondo (Granted, I want to see Lin, Beverley play this year). Outside the top 15-20 guys, fit becomes much more important.
This I can agree with. I feel I responded in a manner that assumed a "big 3" scenario and not "what goes best with Howard and Harden." So there was some misunderstanding there. Personally I would think the cutoff point where talent becomes more important is around the top 40~50ish mark. Or basically your average 2nd guy on an average team. But in practice this may not be feasible due to the salary becoming an issue, as these players are generally vastly overpaid relative to max-players and rookie contracts. Yes, this part is curious. In particular because the Rockets clearly went with Beverley/Douglas to end numerous games. And that the "traditional" view of a typical PG next to a star wing has been the Fisher/Chalmers/Kenny Smith types. Theoretically I can see why a Lin-type PG would work with Howard/Harden. And that's one of the things I'm really curious to see when the season starts.
Is Lin a low usage 7 or a high usage or high minute 9? It seems usage is the main thing that separates 7 and 9. Beverley, Parsons, Brooks appear to be Cluster 4. I would say the combo guard really deserves its own cluster for the Brooks, Beverleys of the world. Asik Cluster 10 Motiejunas, Jones..a poor shooting Cluster 11s Smith an efficient scoring Cluster 14 or maybe 1. Harden 8 Howard 12
Beverley is a 3. Parsons next year should be between a 4 and 11. The difference between a 7 and 9, as well as the difference between any of these clusters, has to go beyond usage alone. Remember, the point of the analysis is to see how different player types interact with eachother. If the only difference is usage, then we are essentially talking about the same player type, which would make this a very shallow thesis. The essence of these clusters is that each is an unique type of player, even if the stats seem similar. Cluster 9's player type is a PG who moves the ball, sets up the offense, while being limited offensively himself. Cluster 7's player type is a PG who is attacking, and his strength is in scoring. Regardless of one's opinion of Lin's quality as a player, his style of play is clear. His strength is in attacking the defense to score or to setup a teammate, a prototypical 7.
Dwight Howard finds greener grass with Houston Rockets http://www.usatoday.com/story/sports/nba/rockets/2013/09/19/dwight-howard-houston-los-angeles-lakers-greener-grass/2840881/
Regarding last year, the 7-8 "big" 2 combo(Lin/Harden) on its own do not produce any meaningful synergy, according to the paper. Only when you add a 12(Howard) into the equation, does the trio generate a substantial positive effect. In other words, the coaches last year didn't really have any added incentive to keep Lin on the court, as far as the "big" 2 "big" 3 analysis is concerned. Also, while Lin's help defense may have been effective over the course of a game, the coaches at times clearly preferred Beverley/Douglas's individual defense to close out games, which is understandable.
