[Simple Stats] Three Point Shooting of the Core Five

Simple ? Where the pictures . My head hurts

 

Where is @JCDenton when you need a graph.

 

I'm too dumb to understand what this means

 

But but i thought asians are good with math, or is that another stereotype.
Arent all Clutchfans members mostly asians?

 

So can you extrapolate what percentage of games they should win by the percentage of threes they should make?

I would but have to leave my PC for a bit.

 

Too many Maury wannaBEs and fake statisticians here

If it aint graphed or tabulated, it's grain of salt

 

Thanks for doing this. The matrix is fun to see. Did you plot each box score 3s per player made to create that?

You mean Anderson, not Z, right? I have noticed that. Ryno is either the only hot shooter in a game, and conversely the only cold one.

And this folks is why you don't let math geeks tell coaches who to play. I admit that -12 correlation is extremely interesting, but it doesn't mean don't play them together. Lulz

For those following along at home, a negative correlation means when Anderson is hot, Gordon is not.

What it can also mean is the defense is picking its poison, which can't happen unless they both play together, assuming they are playing together all the time, which of course they aren't. Or because Gordon doesn't start, it can also mean the opponents are either better at defending our starter #1 option (Ryno) or our bench #1 option, but not both.

And I think the high correlation of Gordon and Bev being hot together is equally interesting. But I'll leave others to suggest reasons.

 

Cool to look at.

Doesn't mean much at all. You can't draw a single conclusion from that aside from Anderson being hot when others are not as @heypartner mentioned.

All of those correlations, both negative and positive, are fairly close to 0. Hence, no real correlation at all.

 

I woke up black today, sorry to disappoint

 

Meanwhile D-MO shot 7.7% since Feb 1st somewhere else. 1 made, 13 attempts. That's much worse than Brew. Who was sad Morey got rid of him?

 

Sorry I didn't want to bump a dmo post

 

What you should do for fun is create a multivariate model to see whose threes best predict winning. Initially it would imply setting cutoffs for each guy, maybe their median threes per game, so you can generate dichotomous variables for each. If you wanna go further, make a linear regression model for each player to see what each incremental three does to our winning pct.

Aside from James, my guess is Ryan Anderson might also have a good correlation to winning. When his shot is on it's hard to beat us.

 

Do some stats for Lou, Gordon, Bev, Harden & Ariza & Anderson about their best spots on the floor from 3pt land, how they get their shots (off of 1 pass, 2 or straight chucking) and their distance. Then maybe we can send it to the coaches and up our %. We should be way better than we are, which is scary, but I'm sure mainly Harden, Ariza & Gordon can be more effective. Harden less off the dribble, Ariza more from the corner and his best spots & Gordon closer to the line. Anderson can hit from far out, miss as well, but he's still around 41% while Gordon is 38%. Would also be interesting to know if Lou can hit a shot from further beyond the line.

 

imagine if we replaced Ariza with Paul George.

 

Or Eric Gordon with Lebron James!

 

Simply put:

1) The Rockets have 6 capable shooters in their rotation.

2) if any two of them go "off" in a game, then they have a chance to win.

If three of them shoot well, then they will likely win

If 4+ shoot well, they can compete with golden state

The frequency of who goes off depends on how good of a shooter they are.

Ex: Gordon 1 of every 2.5 games
Anderson 1 of 3
Williams 1 of 3
Ariza 1 of 5
Bev 1 of 4.5
Harden 1 of 4

Combine those percentages and you have the basis of today's Rockets offensive philosophy

 

I worded my question poorly. By matrix you mean one axis is the player, the other axis is list of games and the column per game is 3s made?