Jump to content

Shot Quality Analytics


Norm Peterson

Recommended Posts

Someone posted a tweet by the Shot Quality account and I started poking around on their website (https://shotquality.com). It's a little different than Kenpom. It analyzes more the individual than the team and adopts the belief that many of us have stated in the past: a bad shot is still a bad shot even if it goes in. But he also adopts the converse: a good shot is still a good shot even if it doesn't go in.

 

And then he breaks down players and teams that are doing the best job of taking the best shots. And he gives players credit for passing not based on whether their teammate hit the shot or not, but whether they should have hit it, at what rate they should have hit it, and what value the shot had (e.g., not all 2-pointers are created equally: some offer a higher chance at an offensive rebound or an and-1.)

 

Now, he's marketing this to both college and NBA teams, so he's trying to make money, but it's an interesting approach.

 

He's ranked, among many other things, all players nationally for the number of points their passing should have generated per game (taking into account the probability the shot had for going in whether the teammate hit the shot or not, and the value of the shot, 2 vs. 3, etc.)

 

Dalano Banton is an intriguing prospect not simply for his length at his position. He's also among the top 40 players nationally (among 350 D1 teams) in terms of points created per pass that leads to a shot, assuming his teammates hit the shots from those passes at the rate they would be expected to hit them (i.e., a 40% look is a 40% look, whether the ball goes in or not).

 

This takes the assist stat to another level and factors out of the equation whether the teammate came through on their end or not.

 

Very interesting. You stat guys might want to take a deeper dive and let us all know what you think.

Link to comment
Share on other sites

The premise is that you take every shot attempt and convert it into a percentage and then you compile them.

The math works like this.

 

Shot Quality is calculated by multiplying the percent chance a shot has of going in, i.e. the probability that the basket is made and the type of shot, denoted by the no. of points that shot entails - either a 2-pointer or a 3-pointer.

So, if I shoot a 3-pointer that has a 25% chance of going in,
Then, Shot Quality = 3 (type of shot) * .25 (% chance of going in) = 0.75

Or, I have a mid range shot (2 pointer) that has a 35% change of going in .
Then: Shot Quality= 2 (type of shot) * .35 (% chance of going in) = 0.70

 

The % chance of going in number isn't some static number.  The secret sauce is how well they combine the various data and factors they're using to compute this per shot. No idea of the accuracy.

 

Link to comment
Share on other sites

Numbers for Nebraska thus far:

 

image.png

 

Most people are aware that Yvan is having shooting issues and some of that you could attribute to him taking low percentage shots.

What stands out to me is McGowens being noted for taking a lot of bad shots.  Is that something that people have been picking up by watching him play? If you just looked at shooting % you'd assume that McGowens is taking good shots and Thorir is taking poor shots. 

 

It would appear that both McGowens and Yvan are tagged as 40% of their possessions as being poor. I think turnovers and assists might be factored in here but I don't know for sure.

Link to comment
Share on other sites

5 minutes ago, hhcmatt said:

The premise is that you take every shot attempt and convert it into a percentage and then you compile them.

The math works like this.

 

Shot Quality is calculated by multiplying the percent chance a shot has of going in, i.e. the probability that the basket is made and the type of shot, denoted by the no. of points that shot entails - either a 2-pointer or a 3-pointer.

So, if I shoot a 3-pointer that has a 25% chance of going in,
Then, Shot Quality = 3 (type of shot) * .25 (% chance of going in) = 0.75

Or, I have a mid range shot (2 pointer) that has a 35% change of going in .
Then: Shot Quality= 2 (type of shot) * .35 (% chance of going in) = 0.70

 

The % chance of going in number isn't some static number.  The secret sauce is how well they combine the various data and factors they're using to compute this per shot. No idea of the accuracy.

 

 

Also, a "good" shot is anything that is .75 or better. So, Thor taking a three, even though he's only hitting at about a 25% clip, would still be a good shot compared to someone driving to the elbow for a mid-range jumper.

Link to comment
Share on other sites

Nebraska's numbers from last year

 

Thor seems pretty consistent in his shot selection

Yvan seems to have improved since last year.

After Cheatham and Kavas these statistics indicate we had guys generally hurling the ball at the rim which seemingly is backed up by the eye test and record of the team.

 

image.png

Link to comment
Share on other sites

If Thor took nothing but threes and only hit 25% of them, his points per shot is still .75.

 

If someone else took a bunch of mid-range shots and hit 37.5% of them, the points per shot would be the same.

 

That probably has something to do with why you still want Thor taking 3s. That, plus you expect his numbers will improve as he settles in.

Link to comment
Share on other sites

1 hour ago, Norm Peterson said:

This takes the assist stat to another level and factors out of the equation whether the teammate came through on their end or not.

 

I like this!  As you guys have heard me say in the past, I think the assist/turnover ratio is kind of dumb (the two aren't connected really at all). 

 

Pass the ball to a guy who shoots a "good shot" equals a successful pass? Yup, sounds good to me.

Link to comment
Share on other sites

2 minutes ago, Norm Peterson said:

 

Also, a "good" shot is anything that is .75 or better. So, Thor taking a three, even though he's only hitting at about a 25% clip, would still be a good shot compared to someone driving to the elbow for a mid-range jumper.

 

I wonder how much of this is built on the type/location of the shot vs calculated ability.  We have a full season of Thor shooting 3s.  Is Eduardo Andre shooting a wide open 3 a good or bad shot? 

Link to comment
Share on other sites

1 minute ago, hhcmatt said:

 

I wonder how much of this is built on the type/location of the shot vs calculated ability.  We have a full season of Thor shooting 3s.  Is Eduardo Andre shooting a wide open 3 a good or bad shot? 

 

According to their website, the algorithm they developed factors in ability. Along with 89 other factors, which are apparently objectively measured (I'm not sure how -- it's proprietary, I'm sure, but even a simple explanation of how you determine how "open" a 3-point attempt was  doesn't appear to be offered.)

Link to comment
Share on other sites

One thing is clear: it's expectation-based analysis, not outcome-based. The data are derived from past outcomes but the determination of whether a shot is a good one or not is based on expectations rather than outcomes.

 

That makes sense to me. The one time you make an improbable shot doesn't justify having taken it if the likelihood was that you'd miss it 90% of the time. So, let's develop a system that puts our players in position to make high-percentage shots that factor in the shot's value. Hoiberg's system does that. I don't think our last few coaches' systems did. In fact (and I remember disagreeing with this at the time) Doc Sadler used to say that if a defender was close enough to foul you on the 3-point attempt, then "you wasn't open." Actually, that's bad math, Doc. A contested three has more projected value than an uncontested jumper from the elbow off the dribble.

 

And, intuitively, I think we knew this, especially after watching our teams get beat by players hitting treys with hands in their faces. You can guard a three. You can contest a three. But if you're a 3-point shooter and you can get the shot off without it getting blocked, it's probably still a shot worth taking.

 

See, for example, that dagger three at the end of the Nevada game.

Edited by Norm Peterson
Link to comment
Share on other sites

1 hour ago, hhcmatt said:

Numbers for Nebraska thus far:

 

image.png

 

Most people are aware that Yvan is having shooting issues and some of that you could attribute to him taking low percentage shots.

What stands out to me is McGowens being noted for taking a lot of bad shots.  Is that something that people have been picking up by watching him play? If you just looked at shooting % you'd assume that McGowens is taking good shots and Thorir is taking poor shots. 

 

It would appear that both McGowens and Yvan are tagged as 40% of their possessions as being poor. I think turnovers and assists might be factored in here but I don't know for sure.

 

I think Trey has been pressing a bit and has forced some shots going to the basket. Some of this may be attributed to him playing too fast at times. Otherwise, I think he has played within himself thus far. Thor looks to be in a mini slump since the first game.

Link to comment
Share on other sites

39 minutes ago, hhcmatt said:

 

Excelling in everything including shooting his way into the KC section of the 3pt shooting matrix

 

image.png

I watched Brandon Ubel in an AAU tournament back in ‘08. He stood in the corner and shot (mostly missed) 3s the entire tourney. He was also skinny and slow. He probably ended up being one of the hardest working players we’ve ever had at Nebraska.

Link to comment
Share on other sites

While interesting, I see this as being far more useful in nba than college.  Way more games and data sets.  There is simply no way this cannot be extremely noisy with such little data at this point.  I don’t know if they are using shooting stats from prior years where applicable, but in any case there are lots of players on each team that are shooting far better or worse in a few game sample.  It obviously will become more and more accurate as it goes along.  

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...