Predictive analytics depends on the inputs. I suspect the Rockets have something similar that factors in future risk of injury, current injuries and recovery times, player values, and more.
Well, likely Hollinger had talked to the people in the TNT before releasing his data, they told him that the rockets would be at 8th regardless. But seriously, this list do not make any sense to me. The rockets just defeated the no.1 team last night without four starters.
It is Morey's shop so you can bet on it. But I would guess that even Morey did not have a plan for three starters out with injuries at the same time.
I'm just seeing it thrown about in the HR world and it disgusts me. We seem to forget that human nature and action are not tied directly to formulas.
The explanation of the Hollinger ratings even says it tends to not be as accurate in the first quarter of the season(first 20.5 games) of which most teams are 3 or 4 games short of right now : I assume regression to the mean includes regression of our overall play relative to the talent we have trotted out so far this year due to injuries. Our strength of schedule has been relatively weak and we haven't blown those teams out(thinking that close game with 0-14 Philly has something to do with it). GS or Dallas both have weaker SOS but they at least blew those teams out. Basically we are a statistical outlier - the kind that makes you rethink your rankings. Playing a historically bad Philly team twice early in the season and not blowing them out in either game hurts us. Meanwhile Toronto & Dallas feasted on them and the hollinger rankings which heavily weight margin of victory overvalue them as a result. .....I don't think predictive analytics can account for the human element of a team's rotations when there is a rash of injuries, or how one team could be coached to lose.
Unless they're projecting the Rockets to be without these injured players most of the season, those are ridiculous odds. Sactown is given a better shot than us to make the postseason.
Whatever. Look at OKC. Expected to go 28-36 the rest of the way and finish 12th in the west. Shows that statistical analysis is not everything.
I'm assuming that's what it has to be, since OKC is still slated to have a terrible year. These things don't really matter, if we recall during Parsons rookie year we had like a 90 percent chance to make the playoffs.
Our schedule hasn't been super hard yet, but we've had far from the easiest schedule so far. The average winning% of our opponents is currently .470. That ranks 20th in the league. Dallas and Phoenix have both had easier schedules to this point. But yeah like everyone else said, this projection can't account for injuries. It also can't account for garbage time. For instance, last night we blew out the Grizzlies, but their 3rd stringers closed the lead to 9 on our 3rd stringers. That "hurts" us in these rankings.
Actually it was 96% after that 4-0 road trip. And really, the Rockets should've made the playoffs. They choked down the stretch that season.
Memphis has been on a roll for the last year and a half. Hollinger is still there as VP of Basketball Operations. A lot of his roster moves helped shape the current team. Not sure why you call him a bozo.
Honestly, these things and mean absolutely nothing. Its pretty much clickbait. Power rankings are fairly meaningless too. The only number that matters is your record at the end of the season. Power rankings won't get you in the playoffs or home court advantage. The Rockets are not well liked by the media and we need to get used to it and not let it upset us.
It is stats based on margin and SOS. All it is saying is that we have had an ewsier draw so far, and havent been as dominant in individual games as our overall record suggests. Now, we have had a couple of huuuge wins end up as 10pt or less margins because we played our bench the 4th. So it affects the stats. we win a few more hard games and it will bump us. it is actually quite a useful tool from the halfway point of the season
As I noted in the game thread when this was brought up, and as many people have said, it is essentially a program that ranks teams based primarily on point differential and strength of schedule and then applies a predetermined amount of randomness in simulating the remainder of the season before averaging the results. First of all, this early in the season the randomness can run away with the model (observe the 50-win Orlando and Boston teams at the extreme points of the standings, or the one where Miami finishes with 20 wins) Secondly, the model is strongly slanted toward teams that are setting the world on fire in the previous 10 games. The Rockets, in the last ten, have a point differential of 0. Compare to, say, the Clippers, who have a worse record but in the last ten are winning by an average of 12.3 points per game. The ranking sees this and counts them as a team on the rise, while the Rockets would be considered to be worse in reality than their record indicates due to having a point differential resembling a .500 team. The model has no method of accounting for injuries. If a healthier Rockets team is to score better in the Hollinger stats, it will only be by improving their point differential as the strength of schedule also increases on its own. I have zero idea why the playoff odds calculator is out already - usually they wait until February or so - but I'd ignore it entirely. The Hollinger rankings, on the other hand, usually does do a very good job of noting who's hot and who's not - as long as you have in mind the appropriate mitigating factors not captured in the rankings such as "oh hey, that team has been missing 3 starters the last month". Hope this helps.
Formulas are guesses at how things will turn out. Not 100% accurate. Measuring the Rockets playoff odds without taking into account current injuries is not going to be very accurate. Spurs, Memphis odds probably are. Knowing the weaknesses in a formula is usually critical to whether it can be trusted for a particular issue. Usually, the Hollinger playoff odds are good late in season for teams that are basically how they were for most of the season.
