"Statistics are always right... unfortunately its the human interpretation that is 90% wrong." -- unknown Sometime object appear closer than what they are. In other words and as previously stated, Minnesota appears to have a better offense than Atlanta but we know that Atlanta has a better offense because they are more efficient at it. So don't be fooled by the numbers... These power rankings attempts to quantify how close you are to a perfect team. An efficient offense and an efficient defense are required to be the cream-of-the-crop. Winning percentage does not really quantify how good a team is. The prime example that most would remember is the 2006-2007 season where Dallas lost to the Golden State Warriors and San Antonio took it all. San Antonio was right there at the top and Dallas was right in the middle. If you go by winning percentage Dallas should have been crowned the champion but we all know what happened. We all know what happens in the playoffs: defenses tighten which means it is harder to score. If you are not efficient at scoring then you will not be able to score and you will be beaten. Therefore, the 3rd quadrant are pretenders. If you are in the 1st quadrant you will be a pretender as well because the tightened defense is not enough. Also another thing to consider is that this is only 13 games in. Therefore, the variance is still quite high. As the season progresses, the values will converge to the true value. The rankings at this point only have 15% of the games. The value in the data is valid but it will improve with time.
Yes - this would be helpful (although I suppose anyone could tabulate it?) My gut feeling is that this data is not especially effective until 30+ games into the season. Too much in the early games depends upon who you've played - it needs time to 'iron out' and read effectively. If you've tagged a couple of games against the Warriors, then you're likely to have an unusually high OE. Or if you've played the Bucks twice, then your OE will be abnormally low. It's the kind of thing that would benefit from interdependent comparison - so checking off OE vs the DE of a team and scaling each and every result. That starts to get ugly I suppose (so many calculations) - but I'm pretty sure that's the kind of data the inputs into Hollingers system.
This is wrong! We are not pretenders, we are lottery-bound! No seriously, great work on the graph! Now I know whose games I will need to watch (the contenders like NOH and DAL; I just hate SA's guts.)
It shows that you aren't including SOS. Miami has one win against above .500 team. Actually, all three good teams in that division should be overrated were it not for Atalanta playing terrible against some bad teams, I think they're a combined 0-11 or something against good teams outside their division.
this is where all this sort of analysis loses me. the "it could have gone a different way" argument. the argument that seeks to tell me how a good a team is by something other than how many games they won.
You have evaluate the team based on what is on the court, not "what ifs". Do you know when Yao Ming and Brooks would be back? Even if they do come back, do you know when Brooks and Yao will perform at mid-season form? if we don't even know those two things you end up waiting until the end of season to do your power rankings. as to the OP, i'd like to give my interpretation to the results of the data: 1. The under performance is due to RA's poor coaching decisions. Thank you for backing up with data something that i have been pointing out game after game: Rick adelman is dragging this team down. 2. The "contender" status always came with the caveat that: 1. Yao Ming regains is old self 2. Ariza was still on the team The Ariza trade made the team play Battier-c-bud combo more, and unfortunately our sfs are playing like they don't even belong in the D-League, much less the NBA. Great thread, great rankings. 5 stars.
You have to remember luck plays a part whenever there is human activity, and in the NBA there is quite a lot of human activity.On games that are won by 1-5 points, a series of lucky shots could end up winning or losing the game, totally coloring your perception of the team. It would be great if luck is indeed a skill, however no one is lucky on a consistent basis so just looking at wins and losses in black and white doesn't give you accurate info. This doesn't even take into account the strength of schedule, upon which wins and losses are dependent as well.
1. Great teams sure find a way of being "lucky", huh? Wonder why that is. I'm guessing it's because they're able to keep games close enough to have luck matter. Because they have guys who respond well under pressure. Because they have players who aren't as mistake-prone, otherwise. "Luck" tends to even out. The cream rises to the top. 2. Of course you have to consider context, particularly regarding schedule. But the best teams usually find a way to win...that's why they're the best teams. We call great teams, "great" not because of games they should have won...but because they converted opportunities to victories. It doesn't make me feel better when the Rockets lose for someone to post a formula telling me how they're really better than the team that just spanked them. Because wins and losses are ultimately ALL that matters. They are how we judge the success of a season. That's what makes a season memorable. Did you get enough wins to qualify for the playoffs? Did you win enough games in that playoff series to advance to the next? That's it. That's what matters.
Yeah. Kyle Lowry said something like "we are a much better team than 4-9" (I think it was that at the time). I remember that Francis and Mobley kept saying similar things ("this is a team we should have beaten, we are better than they are", etc.). JVG said it best: You are what your record says you are.
Bill Parcells said that, too. And he's only talking about a 16 game schedule. I remember Francis/Mobley saying that crap year in and year out. Drove me nuts. Great teams are great BECAUSE they win. Not because a formula said they should have won more.
I agree...not because a formula says it and not because some players who overestimate themselves say it. The results prove where you stand.
Dually impressed mate. Love the graphs. You rock. So, would it be fair to say that the team ranked the highest is the one with the greatest positive spread between their offensive efficiency and their defensive efficiency?
Another subject I would like to see analysts like you and Durvasa study is the pace factor and teams ability to dictate pace. We all know that every team has a pace factor, some play slow, some play it fast. The great contrast in pace from recent history that I can think of was our Van Gundy led teams playing against D'Antoni's Suns teams. I used to watch those games and in my mind would keep a running track of how fast PHX was getting off their shots and how many of our possessions were they getting us to take quicker shots vs. how many times we forced them to take a shot outside of 7 seconds and we forced them to defend for over 18 seconds. It always seemed to me that no matter how the stats turned out, if we forced them to play our pace, we stayed in the games much better, but if they got us sucked in to taking quick shots, then they always seemed to beat the living daylights out of us. On the other hand, in my opinion, a true championship caliber team like the Lakers, it seems to me, they can play efficiently regardless what pace. In fact, it seems to me that they have a knack for slowing it down or speeding it up in direct contrast to what the opponent is trying to do. Sure wish there was some quantitative analysis we could see on this. I am convinced beyond the shadow of a doubt that championship teams have to be able to both run an up tempo fast break offense efficiently and also run the half-court set offense efficiently. I cannot think of a championship team that did not have the ability to do both. They have to be efficient at any pace.
remember the 94 Rockets throwing up tons of points with the Suns...then grinding it out with the Knicks...and winning both series?
Certainly you'd agree that strength-of-schedule is a relevant factor to consider? Also, while I think winning is what teams strive for, to really gauge the strength of a team you need to look at how they won/lost. Not all wins are equal in conveying how good a team is, and similarly not all losses are equal in conveying how bad a team is. There are some things that pure point differential doesn't capture very well, granted. The Rockets have a particular knack for folding in close game situations, and that's not simply a luck thing. Its indicative of a larger problem -- namely, lacking a closer.
