The NBA.com advanced stats tool allows us to look at statistical breakdowns of players, by quarter. Given the recent debate on how the coaches have chosen to allot minutes, particularly in the 4th quarter, I thought it might be interesting to see if there's any evidence to support what they're doing (you know, something beyond the obvious vendetta the coaches have against certain players ). For quarters where there isn't enough minutes for a given player, I put in blanks. I also included results for the entire team in the final row. The Rockets appear to play progressively better from one quarter to the next. Code: [SIZE="2"][B] 1st Qtr 2nd Qtr 3rd Qtr 4th Qtr MIN +/- MIN +/- MIN +/- MIN +/- ----------- ---------- ----------- ----------[/B] Lin 9.1 -1.3 7.6 -0.5 9.5 +1.6 7.4 -0.8 Harden 11.3 +0.6 8.5 +0.0 11.4 +0.4 8.6 +0.5 Parsons 9.4 -2.4 10 +0.2 9 +1.7 9.6 +0.0 Patterson 8.6 -2.6 5.9 +1.3 9.7 +1.1 6.9 -0.8 Asik 9 -0.3 7.8 +0.0 7.7 +0.0 7.5 +1.3 Douglas - - 5.3 -0.4 - - 7.6 +1.3 Delfino - - 7.1 +1.6 5.2 -0.2 8.4 +1.4 Morris - - 5.9 -0.6 - - 7.3 -0.9 Smith - - - - - - 5.2 +3.2 [B][I]TEAM 12.0 -0.1 12.0 +0.2 12.0 +0.4 12.0 +0.5[/I][/B] [/SIZE] The Rockets seem to play well with Douglas and Delfino in the fourth quarter, on average. I think that's why we tend to see a bit more of the Douglas-Harden-Delfino combination than some fans here may like. Greg Smith, in a relatively small sample size, looks to really give the team a boost in the 4th quarter on average.
I don't buy into these things you call advanced stats.......:grin: In all seriousness it is good info, but I still have a tough time accepting these types of stats. There are so many variables when it comes to these stats it makes it difficult to differentiate the good usable stats from the ambiguous waste.
First thing that stands out to me.... Increasing +/- each quarter...This is the reason the Rockets are trying to run. They want to wear other teams out.
A player can suck for the majority of the quarter but drains 2 3s at the end and have a net positive.
OK, but what do you mean by "accept"? I'm not suggesting that these stats tell the whole story. I think when a coach like McHale keeps players in the game a little longer that some may like, or pulls some players out faster, he's doing it based on a "feel" for the flow of the game (which, to me, is mostly related to how the scoring margin changes over time as the player combinations on the floor change). This is pretty much what +/- stats are attempting to capture, so I find it useful to review them to understand what's maybe motivating the coach's substitution patterns. But, again, its just one piece of the story.
This to me is what has been the staple of Morey's time as GM. He seems to have a disproportionate ability to get players at little cost (late first, 2nd rounders, free agents) that actually end up being substantial contributors. Smith is just the next in the line of Brooks, Parsons, Landry, et. al.
I know you're meant to be condescending here. But actually, if a player drains 2 3s each quarter for 4 quarters, he'd average 24ppg and be a superstar. That's utterly ridiculously good production that not even KMart can accomplish. As for the spirit of your post, 22 games is a large enough sample size that these random variations tend to even out.
Let me play the devil's advocate (the devil being LOFs in this case). The offense tends to run through Harden in crunch time because coaches trust Harden in that role more than anybody else. Lin is very ineffective when he plays off the ball. That's why you see him not a good 4th quarter performer. In other words, the 4th quarter offense is just not playing to Lin's strengths. What's interesting though is the 1st quarter numbers. I have not been keeping track on the substitution pattern, so I may be wrong. I think the starters play at least half of the quarter together before any substitution happens, unless someone is in early foul trouble. So the starters' numbers should be fairly uniform in the first quarter. Yet, Parsons and Patterson are substantially worse than Harden. Can someone tell me who are the players first get subbed out? Or is there even a regular pattern? edit: Harden's average minutes in the 1st is 11.3. He's the only starter who almost never sits in the opening quarter. So his higher +/- means most of the time we are down right out of the gate. Then we begin to catch up when the bench comes in playing with Harden.
As far as I'm aware. Lin and Asik are usually the first to get subbed out in the 1st quarter. Patterson too depending on how he plays. Harden and Parsons usually plays the entire 1st quarter (barring any foul troubles). Lin comes back in the second quarter whiles Harden sits out (McHale usually wants one of them on the court at all times).
Oh, dear. This is incredibly stupid for so many reasons. No one, not a single person in the world, thinks unadjusted +/- is useful for evaluating players. That's why there is adjusted +/-, and even better, RAPM. And that's not even to mention sample size. Or developing your best players. Or figuring out how to get your best players working well on the court together.
I'm aware of the dangers of pinning one's evaluation of players solely on +/- stats, whether they be unadjusted or adjusted. And I'd argue that adjusted +/-, broken down by quarter, over a mere 20 game period, would be far less informative than the numbers presented in the first post. The argument that certain players should be playing more because they are to be given more priority for development is of course a valid point to make, but it is a separate argument.
Some formido-ble complaining there, interbro. Since the idea is to figure out which players on the same team are doing their best or their worst in which quarters across relatively standardized lineups, and they've all played the same opponents (i.e. since they are on the same team) then what exactly are you whining about? The variables that are usually "adjusted" for are all constants here. If somebody were to say "James Harden posted a +3 against the Bucks in the second quarter, therefore he is better than LeBron James, who posted a -4 against the Cavs in the second quarter" then all sorts of red flags should go up. But here? Not so much at all. Have you even checked out the advanced stats tool? An internet statistician like yourself might appreciate it. It is totes baller. You can +/- down to the lineup level with ease.
but pure numbers dont account for garbage time where there is no defense and 2nd/3rd stringers on the court... however, with smith, simply watching the game and you can see the man has potential...
This definitely shows to some degree that substitutions need to be adjusted, such as bringing Delfino in for the last 2 minutes of the 1st quarter and possibly (ducking here) starting Douglas in the 3rd or subbing him midway through the 3rd quarter and bringing Lin in with fresh legs for the 4th.
