Who's to blame for Celtics' ice-cold finish in loss to Pistons?

I'll go with this one. The team getting behind that much to a nobody team. I know down 17 doesn't mean jack in an NBA game, but come one guys.

I stated just as the season started that the top teams are going to have to bring it EVERY night. No nights off in the East. Boston immediately puts themselves in the 2nd tier which right now looks like 6-7th place in the East when they pull crap like that. Now lop off 10 more games from the schedule. 6-7th place becomes 9-10th place in actuality. Still early, though.

I can assure you, you don't see the Lakers taking a night off mentally.

db

Yeah, bingbang, the Celtics tend to have a ram it down your (our) throat approach to breaking out of slumps. An extension, perhaps of the shooters mentality from 3-ville on crack.

That's Brad's fault. He's proven over time it's his system or the highway. And part of that is Bombs Away! It may prove to be his demise this season.

db

Vankisa,

Now you're talkin'.  Haven't heard the term selection bias since Master's Stats in grad school!  Loosely defined or corollary as using certain parts of a data set (3-point shots) to support your actions and ignoring the rest (2-point shots are made up of dunks, layups, mid-range and near 3-point shots not just 2-point shots).  3-point shots are clearly defined.  Stand behind the line, hoist, make or miss.  2-point shots are far more complicated to define since it is any shot inside the the 3-point arc.      

Perhaps something they should be looking at is the EV of taking a 3 vs. driving to the basket, scoring and getting fouled and making the free throw.  I'm going to guess by the NBA's actions to date that it is a little less fruitful than chucking up a 3 since there are multiple steps to the process.  But I'm assuming that they have processed that data.  That's what makes a hypothesis.    

The whole reason the analytics started was because the EV for 3's is slightly better than making a two or jump shot.  That's because the 3-point line is defined and most players shooting a three put their feet just behind the line and let it go.  The problem is, where are you taking the 2?  Making a layup or dunk is very high percentage-wise.  Hoisting mid-range jumper makes the average make percentage of a two drop considerably.  Taking a shot just inside the three point line is almost like shooting a 3.  

We can tell from Brad's behavior that he has not broken down the ranges of the 2's.  If he did, he would be driving to the basket like a mad-man when the first three bricks off.  That's why we've seen the complete disappearance of the mid-range shot, ala Kevin Garnett side of the key shot he used to hit very regularly.  You don't hear anymore of players "getting to their spots" anymore.  There are no spots.  

Since there are so many games on LP I'll never see, you may have inspired me to do some data collecting of my own on this and breakdown the 2-point shot and see for ourselves.  This definitely needs to be explored.  Thanks for the inspiration!

db

OK, someone has done this already with EV, of course, but it does show what we already see has happened.  It will just be a question of determining what the percentage "foul" rate comes in at when one drives to the basket, scores, is fouled and then completes the 3-point play.  I'd call it the "And 1 with Free Throw Made" rate.  There is not much difference in EV for shooting a three and making a layup.  The question is how much more EV is there in the 2-point sequence described above.  Actually the layup/dunk is worth slightly more statistically.  But I'm going to guess the rate at which you make a two and are fouled is low as it should be.  That's what the defense is trying to do is get you to not make the basket by fouling you.    

https://fivethirtyeight.com/features/how-mapping-shots-in-the-nba-changed-it-forever/

db

P.S. I believe selection bias is a step or two further down the road from initial data mining, which is going into a large data base (all the NBA shots, location of, etc.) and "finding" what you are looking for to support your hypothesis.

EV used to stand for Expected Value.  That last chart in the article he calls it something different but it's the point value you can expect by hoisting the ball from anywhere on the floor.  EV = actual points for the shot (2 or 3 or 1 for free throw) X probability of making that shot.  Both are readily available and easy to calculate given the NBA's long history of data gathering.  

All the analyses does is spell out the obvious.  Lay-up/dunks and 3-point shots (given they are 50% more valuable than 2 point shots but at a price...length/distance) are the more efficient ways to score.  There are two ways to get to the finish line.  Us old schoolers know that you open your game by going inside to get the defense to collapse in the paint.  You do this enough times and the 3-point becomes wide open.  

That's the way it was for years.  Even in the Larry Bird era the 3-point shot was used at the end of games to win it or lose it.  But never as part of a full game strategy.

Today's mentality is to establish your 3-game first which opens up the lane.  The problem with the 3-ball mentality is if your guys ain't hittin' right out of the chute, you're going to be behind and chucking 3's at the end of the game to catch up.  Sound familiar?  You can even dial in the microscope on a game and see that has been our problem at certain points of the game resulting in games we lose that we should have won.  

Mike Ditka and Joe Gibbs were two great NFL coaches during the 80's.  Two totally different ways they coached their teams.  One coach was calm, cool and collected all the time.  The other was lam blasting everyone in sight on every play.  Both won Super Bowls.  Think of 3-ball and old school inside-out in this context.  The latest example of it has been GSW.  

The answer for the perfect solution that produces the greatest points on the board looks more like what I remember for modern portfolio investment theory in grad school.  It is how you mix the scoring options that will determine who has the best (maximum) "return" on their investment.  That is, how does a team mix their shot selection between 3-point shots, driving for And 1's, pounding it in the paint, fast break, etc.  that produces the maximum number of points during a game, given a load of variables.  As we have said on this site numerous times after we pier up out of the weeds, it's the team with the most points when the final horn goes off that wins every time.  

So someone needs to develop that allocation theory of shot selection that maximizes the points on the board during a game based upon a load of variables from both teams, where the game is being played, how hot it is inside, who's starting, who's coming off the bench....you get the picture.  It's been done in other disciplines, especially in the investment world.  It just hasn't been completed in the basketball world yet.  Right now, the NBA is just exploring the other side of the coin with 3-ball.

Now baseball is the ultimate extreme.  It has a stat for EVERY situation but it still does not have a unifying theory and application on how to maximize runs given what/who you've got to work with.  That's just the way they've done it all these years and it works.  They don't need one.  They've covered all the bases!  

I don't know if basketball will eventually wind up this way but there are tons of variables to consider in this sport as well.  Most of us are probably not baseball fanatics on this site, but I can spell out any situation and you can tell me with great accuracy what the skipper is going to do.  

The key as I see it is to make the game look like you are the only one on the court like you are practicing by yourself by putting yourself in a position to be at your spot where you are the most productive and be wide open.  The good (and the bad) news is you have to have the team help you (and you help them) get to those spots.  Plus there's the other guys trying to stop you.  They read from the same book.

db

Unfortunately, I can't type!! My wife gets on me for not pursing a career that matched my degrees! So you guys and gal get to receive the brunt!

db

NYCelt,

And it works in baseball because that game naturally breaks like football. Basketball is more continuous and therefore harder to pin down in the scenario world.

Worcester,

Your smarts saves lives. Far more important. My smarts is based on slight of hand! Like I've said before, I can convince you beyond a shadow of a doubt the sun will rise in the West tomorrow with my tools!

dboss,

Your wish is Brad's command!!  Shocker!!!

db

P.S. After last night's game, someone has tipped Brad off to this site!!!

Watching that game last night vs. TOR reminded me of those old 1920's videos of NY City intersections without traffic lights!

But no one was in an accident. Organized chaos. Everyone moving but with purpose. Now to TOR it must have looked like a train wreck!!

db