I think we all know that Daryl has a model for everything. We also know that Morey's models tell him the most efficient way to score is 3's, FTs, and close shots. Does Morey have a model that predicts likelihood of winning a game with a certain lead and future shot distribution? You'd have to think that if the team is sitting with a 22 pt lead a halftime, a model would tell you that if you took more 2s than 3s in the second half, the odds of winning the game are greater. If this isn't something the front office has thought about, hopefully this thread spurs that discussion. There's something to be said about ideal shot distribution to maximize points throughout the course of the season but ultimately, we should be implementing models and shot distributions that give us the greatest chance of winning every game. Another case in point. When you are down 1 with 30 seconds to go, a 3pt is probably the worst percentage shot in terms of likelihood of winning and that's exactly what James took last night. Shouldn't we have analytics that tell players not to do exactly what James did last night?
The odds of winning are greater if u take more twos than the threes in the 2nd half based on what exactly?
The theory would be that taking more 2s probably has less variability. If you have a large lead, your goal changes from maximizing your chances of outscoring the opponent to wanting to minimize your chances of getting outscored by a lot. I have no idea what the data would show, but i imagine more 3s increases your chance of both outscoring your opponents by a lot (unnecessary if you're already ahead by a lot) and also being outscored by your opponents by a lot (the one scenario that loses you the game). Shooting lots of 3s , for example, might give you a 60% chance of outscoring your opponent, a 20% chance of being outscored by a little, and a 20% chance of being outscored by 20 (ie, losing). Shooting 2nds might result in 40% chance outscoring your opponent, 50% of being outscored by a little, and a 10% chance of being outscored by 20. If that were the case, then your best bet to win after you have a big lead is the latter strategy. It would be similar to being ahead by 10 with a minute to go - a shotclock violation is better in that scenario than taking a 3 a few seconds into the shot clock even though you'll score less points on average. That said, changing a strategy for 1 minute of a game is easier than changing a whole style of play for a half. So while it might a good strategy in theory, the Rockets may not know how to execute it effectively and thus it works out more poorly than the numbers might suggest.
only guys who can get to the rim are Harden and EG...EG is always making bad decisions, and Harden had an injured shoulder...Harden was 5-12 from 2 last night only guys who are competent from midrange are CP3 and Harden...they both took a few Rockets let OKC drop 69 points in the 2nd half, and the focus is on our offense
I don't think analytics have to tell a player that chucking up a 3 after burning 20 seconds doing nothing while being down 1 in the last possession of a game is a bad decision, that's just common sense.
I didn’t like what harden did there at the end of the game. 30 seconds left take it to the hoop early if you don’t get the basket or foul u still have plenty of time to foul and get the ball back. I especially didn’t like it because the possession before down 1 he took it to the hoop and scored to put us up 1. Why not do that again? None the less this was a bad loss. Allstar game and Clint back cant get here fast enough. Major blown opportunity to be 1.5 back of okc last night and be in the 4th spot with Portland. Just bad this one hurt.
In the final two minutes of a game where you are down 4 pts, statistically, the 2 pt shot is more effective. Let's say the average 4th quarter 3 pt shooting percentage of the Rockets is 36% and the average 2 pt shot percentage is 52%. To reach 4 pts through two pters would have a (.52*.52) 27% chance while taking two threes will have a (.36*.36) 13% chance. Even if you include a third 2 pt attempt(.52^3)14%, the two is more efficient in this scenario . The smaller the sample size, the more efficient 2 pt shots are.
Don't know if there is a stats for this. It seems whenever we have a good 3 point shooting half say above 40%, the second half 3 point shooting is usually subpar, like in the 20-30%. Vice versa. In another words, we rarely have 2 hot shooting halves.
Of course, top teams adjust and extend the defense. This where the layups parts of our offense can thrive. Problem is CP3 is still hit or miss and not having the Capela lob option hurts.
Pretty much every single human besides Rocket fans believe that Harden gets the benefit of most calls.
Not talking about Harden. Talking about the game in general with all the phantom calls that OKC got 2nd half. It grinded the game to halt and let them back in the game. When I see refs disrupting the flow of a game like that it's obvious to me they're not liking the flow so far and they've got to do something about it. At the end they don't care who wins or loses but gotta cover the spread and hit the over if you know what I mean.
i think part of the reason harden shot the 3 at the end was the precedent during that entire game by zach zarba, who was swallowing his whistle when he got hit on drives
YES THANK YOU. For the love of god this is the first time someone has actually understood this on this site. We have a bunch of wanabee statisticians that stumbled onto "3s are more efficient!" What they fail to recognize is that the strategy is situational because the 3pt shot is subject to more variance. A fadeaway 3 at the end of the game while you are fatigued is going to have a lower % than your normal 3 pt shot coupled with the fact that you have less total possessions with a shot that has more variance (a bad bad mix). I actually think we should still take open 3s and what not regardless but this team takes so many bad 3pt shots.
