This seems to be a popular topic for Rockets fans nowadays. There are different camps. Some feel that T-Mac should be our main facilitator or distributor, as he was under JVG. Some feel he should be more of a finisher with far less touches. Is he better as a jump shooter coming off screens, as an ISO player on the mid-post, as a pick and roll player at the top of the key? Then, you have to also decide how his role should change over the course of a game. Does he need to dominate in the fourth quarter, as so many people were frustrated he didn't do in game 7 of last year's playoffs? Or should he play a more team-oriented game down the stretch, as many have argued after the recent win against Milwauke? I wanted to contribute something more objective to this discussion. I took all of T-Mac's games as a Rocket in which he played at least 30 minutes (205 games in total). Then, I calculated how various stats correlated to wins (no regard for point differential, just whether we won or lost). That includes all the typical box score stats, those stats per minute, his eFG%, his TS%, his Usg Rate, also Hollinger's composite game score rating (GS, raw and per minute). Correlation coefficient can vary between -1 and 1 ... where negative means it tends to lead more to a loss and positive means it tends to lead more to a win. That doesn't imply causation, though (e.g. offensive rebounding is a good thing, but the more you get the more your team tends to be missing shots ... hence it correlates more to losses). The following table shows how these stats rank in importance, based on this analysis: Code: [B]Stat Correl [/B] Win +1.000 | FTA +0.111 GS_min +0.307 | DRB_min +0.103 TS% +0.279 | 3P +0.098 GS +0.261 | TRB_min +0.091 eFG% +0.253 | TOV_min +0.071 PTS_min +0.214 | Shots_min +0.069 AST_min +0.203 | DRB +0.066 STL_min +0.185 | TRB +0.052 FG_min +0.170 | BLK_min +0.044 FT_min +0.161 | TOV +0.031 STL +0.158 | FGA_min +0.023 PTS +0.156 | BLK +0.022 FTA_min +0.148 | Usg +0.012 AST +0.148 | ORB_min -0.011 PF_min +0.138 | Shots -0.017 FT +0.129 | ORB -0.024 3P_min +0.125 | 3PA_min -0.032 PF +0.123 | FGA -0.055 Usg_min +0.117 | 3PA -0.079 FG +0.116 | MP -0.159 I found this kind of interesting, personally. Firstly, I notice that a lot of his per minute stats tend to correlate better with winning. So, assuming he plays enough minutes (in this case, 30), what he does per minute seems to be a little more important than what he does overall. Next, this shows that free throws tend to be more important than 3-pointers. Some may consider this to be common sense, but it's nice to reaffirm it with evidence. Assists and Points have a similar correlation to wins, and it's strong relatively speaking. Shot attempts (combine FGAs and FT trips) correlates weakly to winning; my interpretation is he can either shoot you to a win or shoot you to a loss ... not much indication that he does one more than the other. I look at Usg as well, which combines FGA, FTA, AST, and TO into a number which relates to how involved he is in the offense. It's fairly positive, which one may expect, but not overwhelmingly so. It's interesting that MP correlates the most to losing of all the stats. I think this could reflect poorly we've played in close games over the years (where T-mac would tend to play more minutes); or maybe it's just that when there's a blowout (less minutes for T-Mac) we tend to be on the winning side. Anyways, I thought these numbers might be of interest in this discussion about how T-Mac should be used. I'm not sure what's the best way to interpret them, but I gave it a try above.
Durvasa, nice effort, but your stats do not show anything that further empowers the view that T-Mac is a cancer, stagnates the offense, doesn't throw it to Yao, or dominates the ball too much. Oh, and that he only shoots highly contested 22 footers with less than 5 minutes left in the game. Therefore this is inconclusive. /sarcasm Nice stuff. Not sure what conclusions to draw, other than the obvious ones you stated.
I'm using Excel, and I don't think it provides that for correlations, unfortunately. Another possible problem with these stats is they mostly involve games played under JVG. Maybe (or maybe not) the way T-Mas has to play to be successful under Adelman is different? I think if I just focus on this season's games, though, the sample size wouldn't be nearly enough.
T-mac and Yao are the all stars of the Rockets They should make other teameats better and get more points.
T-Mac needs to control the point like Magic Johnson, score bucket loads like Michael Jordan, be as driven and as clutch as Larry Legend, show off the low post game of Kevin McHale and conquer the paint like Wilt Chamberlain. After all... aren't we paying him 20 million a year to do just that?
Having watched him play this year, it seems McGrady is a more effective scorer when he is coming off of picks, catching and shooting from 17 or 18 feet or catching and instantly penetrating. He seems to be effective also running PnR when he is being aggressive with the ball. He isn't as effective a player when he is thrown the ball out behind the 3 point line and asked to create something out of nothing. This is usually what leads to his bad shot selection in my opinion. I really believe he is a better player when not dominating the ball.
TMac needs to do whatever the team needs that particular game, he needs to be a chamellion....sometimes scorer, sometimes distrubtor, sometimes rebounder, sometimes a bit of everything. He just needs to play his role, just like everyone else, and that role will change from game to game. Play as a team...that is the goal....from everyone. DD
I don't know too much what to think of your correlations. (I have a lot of thoughts, but nothing conclusive.) So, purely from observations: I think McGrady is better when he plays off the ball. Among players that are required to handle the ball a lot, I think McGrady is the best at catching-and-shooting, and pretty good at cutting and finishing. Unfortunately, because our second-best facilitator, Rafer, has trouble against good defensive point guards (destroying Baron Davis notwithstanding), McGrady generally helps the team more against good teams when he has the ball. If the Rockets could switch it up, where Rafer is the primary ball-handler against mediocre and bad teams, and McGrady is the primary ball-handler against good defensive teams, I think it would be best for the Rockets. If not, I think he is forced into the primary ball-handler role until the Rockets upgrade the point guard position.
I don't know much about multiple linked spreadsheets or correaltionships of variable factors...or how to create an equation making Chuck Hayes the MVP of the NBA.... But I do know it's easier to get into a good shooting mode...and it increases your desire to play D...when you actually TOUCH the basketball. So regardless of the PhD-level mathematics, I'd say anything that gets McGrady's teammates to move without the ball and to get them the ball when they are in motion ought to have a good chance for success.
T-Mac should play the way Kobe is playing right now or the way Ginobili usually plays for the Spurs, spot up shooter when he is OUT of the 3pt line, and then play around screens to take shots off the ball and not off the dribble or penetrate after going around a screen. All off them are quite easy plays (see Richard Hamilton for the shooting and penetrating coming off the screen) and should end in rising T-Mac shooting % for around .460. At the end of the fourth i would like isolation plays for him to start at the top of the key the way the Spurs and the Lakers do at the end of the games trying to get a layup or to get to the line.
Game Score ... just a composite rating that adds up all the good stats and subtracts the bad stats, with appropriate weights.
How do you think the offense, and specifically the way we use T-Mac, should change in the 4th quarter? The offensive balance seems to go out the window when it really counts ... a bad thing? And I've heard that being more involved on offense will increase your desire to play defense before. I haven't seen any kind of evidence to support that, though. I think it could just as easily be said that if a player is less involved in offense then he'll try to do more on defense to compensate (not to mention he may have more energy to expend on that end, and he could be in better position for transition D ). I do agree that more movement from other players when McGrady or Yao has the ball can only be a good thing. I'd like McGrady to look to pass more to players going towards the basket. On PnRs, he has a tendency to both dribble and pass away from the basket. It's the safer play in terms of limiting turnovers, but it often doesn't get us the best quality shots.
