Living and Dying by the Three-Pointer

[RSN Diaspora]

As the C's put away game two, I made the following note in the game thread:

Solid win. They need to figure out how to shoot in the paint—these 3s aren’t sustainable.

But this wasn't a universally held opinion:

If they get the driving, passing, and spacing that they had in the first half, it’s sustainable

(assuming White or Theis don’t take any)

They’re going to get a shitload of wide open threes every single game and I want them to take all of them. I liked them working in some midrange in the first half but I don’t foresee a whole lot of success at the rim with Milwaukee packing the paint.

So can we continue to live and die by the three, effectively ceding the paint for all but occasional mismatches and openings? The Celtics made 20 three-pointers last night, the Bucks only three. Beyond the arc, the Celtics shot 46.5% (20 out of 43 attempts) while the Bucks shot 16.7% (3 out of 18 attempts). We were even on 2nd chance points (nine each) and fast break points (six each). But we got outscored by 30 in the paint: 54-24. If we can continue to shoot 46.5% from three, it won't matter: obviously, a bucket that is worth 1.5 times one sank from inside the arc is vastly preferable. But we shot 36.4% in game one against the Nets, 35.5% in game two, 41.4% in game three, and 37.1% in game four. In every one of those games, we scored more in the paint than Brooklyn. In game one against Milwaukee, we shot 36% from downtown, barely eclipsing the Bucks' 35.3% from three while also getting dominated everywhere else on the floor.

Each opponent is different and each game is different, but can we realistically expect to put as many eggs in the proverbial three-point basket in the remainder of this series? I'm skeptical, but I'd love to be convinced.

 

[tims4wins]

As the C's put away game two, I made the following note in the game thread:



But this wasn't a universally held opinion:




So can we continue to live and die by the three, effectively ceding the paint for all but occasional mismatches and openings? The Celtics made 20 three-pointers last night, the Bucks only three. Beyond the arc, the Celtics shot 46.5% (20 out of 43 attempts) while the Bucks shot 16.7% (3 out of 18 attempts). We were even on 2nd chance points (nine each) and fast break points (six each). But we got outscored by 30 in the paint: 54-24. If we can continue to shoot 46.5% from three, it won't matter: obviously, a bucket that is worth 1.5 times one sank from inside the arc is vastly preferable. But we shot 36.4% in game one against the Nets, 35.5% in game two, 41.4% in game three, and 37.1% in game four. In every one of those games, we scored more in the paint than Brooklyn. In game one against Milwaukee, we shot 36% from downtown, barely eclipsing the Bucks' 35.3% from three while also getting dominated everywhere else on the floor.

Each opponent is different and each game is different, but can we realistically expect to put as many eggs in the proverbial three-point basket in the remainder of this series? I'm skeptical, but I'd love to be convinced.

I posted this in the other thread, but if the Celts shot their season average (35.6%) on 3s last night, they'd still have won by like 8 or 10.

The issue in game 1 was 10-34 (29.4%) from 2. If they shoot 16-34 (47.1%) which still isn't great (they were at 54.7% for the year), then that's a tie game.

Given how it seems Milwaukee is ok in conceding the 3, so long as the Celts are taking good 3s, I think we have a reasonable expectation that the Celts will outperform their season long (35.6%) 3 point average. In which case, over the course of 7 games, I think that will be enough to prevail.

 

[BigJimEd]

I'll quote these stats from @Cellar-Door in the other thread.

I would say, one thing about MIL is they give up a lot of 3s that would be classified as "wide open" in fact they led the league in wide open 3PA allowed in the regular season.
In 2022 the Celtics shot over 40% from 3 on wide open looks on pretty high volume. So shooting in the low to mid 40s this series is not otherworldly, it's just above average.

So decided to look it up....

Celtics have gotten 45 3PA through 2 games that NBA.com classifies as "wide open" and hit 18 of them, so they are actually right at what they shot in 2022.

I do expect the Bucks to adjust though.

 

[dhellers]

Unlike the Nets, your are NOT beating the Bucks simply by imposing your will.
Which means you have to take advantage of your opportunities, and limit theirs. And then hope for the best.

While seasonal averages might inform decisions, the real choice is: if a 3pt opportunity will yield the highest expected points in a possession, than that 3 is the one you take.
That's not necessarily the first available 3, so that doesn't mean you cut short ball movement. But it does means that if the Bucks err on the side of giving you great looks, you take them.

What am I really saying: there is no shortcut for good performances from as many players as possible. GW and PP hitting open 3 makes will make a difference.
Al hitting a few will help, also Marcus (and and White?) -- mixed in with interior scoring when the opportunities are there.
And of course: JT and JB attacking and finishing/passing are crucial -- with a special attention to "no stupid turnovers".

If that isn't enough, if the Bucks shoot the lights out or get just enough pressure on the 3 shooters to render them mediocre, you tip your hat and move on.

As stated elsewhere, this is a heavyweight fight!

 

[joe dokes]

So can we continue to live and die by the three, effectively ceding the paint for all but occasional mismatches and openings? The Celtics made 20 three-pointers last night, the Bucks only three. Beyond the arc, the Celtics shot 46.5% (20 out of 43 attempts) while the Bucks shot 16.7% (3 out of 18 attempts). We were even on 2nd chance points (nine each) and fast break points (six each). But we got outscored by 30 in the paint: 54-24. If we can continue to shoot 46.5% from three, it won't matter: obviously, a bucket that is worth 1.5 times one sank from inside the arc is vastly preferable. But we shot 36.4% in game one against the Nets, 35.5% in game two, 41.4% in game three, and 37.1% in game four. In every one of those games, we scored more in the paint than Brooklyn. In game one against Milwaukee, we shot 36% from downtown, barely eclipsing the Bucks' 35.3% from three while also getting dominated everywhere else on the floor.

Each opponent is different and each game is different, but can we realistically expect to put as many eggs in the proverbial three-point basket in the remainder of this series? I'm skeptical, but I'd love to be convinced.

Chances are, if the Celtics' 3-pt shooting craters, it will be at least in part to changes in the Bucks' defense, which should open up other ways to score on open shots, instead of the open 3's they are getting now.

 

[JakeRae]

Threes are a poor category choice. The NBA divides shots for tracking purposes into restricted area, paint, mid-range, each corner, above the break, and backcourt. I think you can simplify that to rim, lane, mid-range, corner, above the break as the five principle areas you can shoot from. Getting to the rim is the best shot in the game still. Corner threes are the second best shot in the game. If we’re getting either of those, it’s a very good possession.

One issue with the Bucks defense is they are really really good at preventing shots at the rim and turning them into shots in the lane. Those are not particularly good shots. So, if we decide we want to attack the rim, we will likely get forced to settle for shots from 5-8 feet in traffic and it will look like what happened in Game 1. The Bucks concede above the break threes, which are actually the third most efficient category of shot, so mid-range and lane shots are particularly bad against the Bucks because there is always a good look above the break. The key is to generate corner or rim looks. It’s not about getting either one specifically. If we are getting a reasonable percentage of shots from either location we are winning the offensive battle.

To illustrate, in the regular season we shot 35% above the break, 40% from the corners, 40% in the mid-range, 42% in the lane, and 70% in the restricted area. In the playoffs we’ve been at 35%, 45%, 46%, 36%, 65% by category. Corner threes are a .600+ TS% option. We should take as many as we can get and we will consistently win if we are getting a lot of looks from the corner like we did in Game 2.

 

[joe dokes]

To illustrate, in the regular season we shot 35% above the break, 40% from the corners, 40% in the mid-range, 42% in the lane, and 70% in the restricted area. In the playoffs we’ve been at 35%, 45%, 46%, 36%, 65% by category. Corner threes are a .600+ TS% option. We should take as many as we can get and we will consistently win if we are getting a lot of looks from the corner like we did in Game 2.

The corner point reminds me of something that van Gundy highlighted. After Tatum drove and dished to the corner a couple of times, he ran the baseline out to the corner he just passed to and was wide open to receive a pass from the pass receiver after the latter took a few dribbles toward the lane.

 

[benhogan]

1. The Bucks have 3 shot blockers around the rim (Giannis, Lopez, Portis)
2. Jrue is strong with stripping the ball in the lane
3. The Bucks desperately need turnovers and transition points since their top option in the half-court offense is out for the series
4. Middleton is one of their better perimeter defenders

Why on earth would anyone find it necessary to play into the Bucks strength, THE RIM, when if you rotate the ball you can find an Open3 from the Corner or above the break

Slow the pace down, be strong with the ball, shoot 35-40% on 3s and the Bucks will struggle to score over 100 in the halfcourt. The C's will win

If the C's shoot better than 40% from 3, then they will blow out the Bucks. This will take them away from the rim. Pump fakes will be even more effective.

If they turn the ball over and shoot below 35% from 3 the C's will lose.

 

[the moops]

MIL was the only team in the league to allow more than 40 three pointers per game by their opponent. Their 40.6 allowed was 5.5 more than league average. So yes, live and die by the three because that is where their weakness is

 

[Cellar-Door]

So I can get deeper into it later but the basic breakdown on this series when BOS is on offense is.... The Bucks are very good at defending the paint(season numbers aren't as good as they are because Lopez was out) but not particularly good at defending 3s, the Celtics are a good 3pt shooting team and not great between the 3pt line and restricted. So this is a case where the Bucks are giving the Celtics the shots that the Celtics want so long as they work for good ones like game 2 by collapsing and kicking and avoid forcing shots inside against the help and over penetrating like game 1.

 

[benhogan]

So I can get deeper into it later but the basic breakdown on this series when BOS is on offense is.... The Bucks are very good at defending the paint(season numbers aren't as good as they are because Lopez was out) but not particularly good at defending 3s, the Celtics are a good 3pt shooting team and not great between the 3pt line and restricted. So this is a case where the Bucks are giving the Celtics the shots that the Celtics want so long as they work for good ones like game 2 by collapsing and kicking and avoid forcing shots inside against the help and over penetrating like game 1.

and it's even more exaggerated when you lose Kris Middleton and replace him with Portis

 

[SoxJox]

Chances are, if the Celtics' 3-pt shooting craters, it will be at least in part to changes in the Bucks' defense, which should open up other ways to score on open shots, instead of the open 3's they are getting now.

I think this is exactly right. It is an interactive dynamic, or, as I like to say, the "squeezed-tubular-balloon" effect. You squeeze one section of the balloon to address an issue, and another issue (read opportunity) pops out or is created in another. Of course, it depends on how Ime can adjust and the players can execute. I think there is sufficient diversity in the Celtics' game plans to make those adjustments.

 

[SoxJox]

and it's even more exaggerated when you lose Kris Middleton and replace him with Portis

This comment reminded me of a thought that occurred to me watching Game 2. What I found interesting, after losing Smart, was how Grant W. sucked up that vacuum. Of course, there's plenty of discussions elsewhere about the deserved love being showered on Grant, but I thought there were at least 3-4 clear instances where Giannis backed off and passed simply because Grant was guarding him.

 

[Cellar-Door]

Ok, so a bit more detail now that I'm not on my phone....

MIL wants to play a certain way on defense.
They play big, they protect the rim and paint at all costs, with a great defensive C in Lopez, and Giannis playing the same role TL does... put him on the worst player and let him roam. This means they are a monster team in the paint... especially since they usually play with Portis who is big a lot, and they have Jrue a rangy all-defense PG and Middleton a solid rangy wing.
The weakness of this is you give up 3s.

The strategy for Bud is that you make teams beat you with 3s, while using your size to bully them on the boards on offense, and getting enough points to beat them between 2 good scorers and lots of chances. Also they took a lot of 3s outside Giannis. Also to use the length and shotblocking and hard traps/press to create turnovers to get fastbreak points to supplement your good not great halfcourt offense.

Until last year.... this never worked in the playoffs....
In 2020 the Heat took 40 3PA a game hit 37% and coasted to a 5 game win, in 2019 after getting past a bad shooting C's team (36 3PA but only shot 30%) they ran into the Raps (38 3PA his 37% bye bye in 6).

This year in round 1 they got the perfect matchup... the Bulls take the least 3s and suck at it, and the guys who were best at it were hurt.

Against the Celtics the problem they have is this:
1. The Celtics are a lot more like those Heat and Raps teams on offense, they are happy to take close to 40 a game and likely to hit 35-40%.
2. The Celtics have an excellent defense that can slow them down
3. The Celtics are big without losing much shooting, no real size advantages in the backcourt/wings.
4. Middleton their best all-around scorer probably is out, so they have to go with a smaller guy who can't close out as well making the 3s easier, and making them easier to guard/keep off the glass.

That means they need to win turnovers and transition.... and hope for poor or lower volume shooting.

Comparing to BRK isn't a great strategy because.... BRK can't defend the rim or rebound, the two things the Bucks are best at, and they live in the midrange (KD and Kyrie) and behind the arc (everyone else).

So in this series there are 3 areas of the court....

At the rim... on offense the Celtics are good not great here... MIL is excellent defensively.
midrange... the Celtics aren't particularly good here, MIL is good defensively
Outside the arc... the Celtics ARE good here, MIL is weak.

So the Celtics are going to take a lot of 3s... BUT... one key is how you manipulate to get them. MIL is great at the rim because they collapse and sell out to protect it, with the hope that either you challenge the shot blockers, or you get stuck, miss the pass and turn it over so they can run for easy buckets. What the Celtics need, and did in game 2, is to be patient, probe with drives, make the right pass, then if needed a 2nd pass or 2nd drive while the driver relocates, get the right looks for the right guys with MIL in motion.

So where last series was about using size and at the rim finishing to punish a small team, this is about getting a good/big team in motion to get open rhythm 3s, and force them to either outscore you (against your top D) or eventually go small, possibly opening up the rim some.

 

[slamminsammya]

One small point about maximizing per possession efficiency - in general it is true you want to more or less get whatever shot will give you the highest number of points in expectation. When a game is close and you are protecting a small lead or trailing by a bit this stops being true and the variance actually matters a lot more, since the marginal value of each point becomes much different.

This only matters for maybe a handful of possessions in a series, but they will be high leverage. If the Celtics are down 1 in the final minute, you will obviously mind that they have trouble getting to the rim.

 

[lars10]

One small point about maximizing per possession efficiency - in general it is true you want to more or less get whatever shot will give you the highest number of points in expectation. When a game is close and you are protecting a small lead or trailing by a bit this stops being true and the variance actually matters a lot more, since the marginal value of each point becomes much different.

This only matters for maybe a handful of possessions in a series, but they will be high leverage. If the Celtics are down 1 in the final minute, you will obviously mind that they have trouble getting to the rim.

When it mattered yesterday the Celtics went to the rim a few times.. they also hit two threes to seal it. I don't understand why a team should force it to the rim when their opponent is giving up wide open threes and your team is actually pretty good at hitting wide open threes. Why does it make more sense to take a contested layup against two or three defenders in the paint than it does to kick the ball out to another player that has nobody on him? How much does the percentage variance change with how the Bucks defend? Because twos are a lot harder than normal and threes are a lot easier than normal.

 

[slamminsammya]

When it mattered yesterday the Celtics went to the rim a few times.. they also hit two threes to seal it. I don't understand why a team should force it to the rim when their opponent is giving up wide open threes and your team is actually pretty good at hitting wide open threes. Why does it make more sense to take a contested layup against two or three defenders in the paint than it does to kick the ball out to another player that has nobody on him? How much does the percentage variance change with how the Bucks defend? Because twos are a lot harder than normal and threes are a lot easier than normal.

Yeah, I am mostly talking about situations like you are down 1, final possession - the difference between a 3 and a 2 is none, you just take the higher percentage shot. And similar types of situations.

 

[benhogan]

I don't understand why a team should force it to the rim when their opponent is giving up wide open threes and your team is actually pretty good at hitting wide open threes. Why does it make more sense to take a contested layup against two or three defenders in the paint than it does to kick the ball out to another player that has nobody on him?

NBA media intelligentsia is propping the 1 on 3 rim attempts as a highly efficient shot. Morons

 

[lovegtm]

Ime has adjusted well to get more Grant and PP out there, since they're elite shooters who make this strategy a lot worse for the Bucks.

If the Celtics play tight rotations and don't put DWhite and TL together too much, they can keep a lot of shooting on the floor.

Conceding lots of open 3s to Tatum/JB/Grant/PP/Horford is likely not a winning strategy over 7 games.

 

[wilked]

I’ve always felt that a 3 being “only” one more point than a 2 masks it’s power. The better way to view it is that it is 50% more points, and if instead the NBA scored it 4 within the arc and 6 outside the arc people would see it more clearly.

there’s a reason NBA teams focus on not giving up uncontested threes, and it’s because your average NBA player hits those more often than not. That’s easy enough to see during a shoot around at practice. That the Bucks seemingly are either willing or unable to prevent these looks consistently for the Celts is something to exploit again and again until the Bucks are either forced to adjust or are on a plane home

 

[Red Averages]

I’ve always felt that a 3 being “only” one more point than a 2 masks it’s power. The better way to view it is that it is 50% more points, and if instead the NBA scored it 4 within the arc and 6 outside the arc people would see it more clearly.

there’s a reason NBA teams focus on not giving up uncontested threes, and it’s because your average NBA player hits those more often than not. That’s easy enough to see during a shoot around at practice. That the Bucks seemingly are either willing or unable to prevent these looks consistently for the Celts is something to exploit again and again until the Bucks are either forced to adjust or are on a plane home

I put it in the other thread, but you can shoot 33% from 3 or 50% from 2 and have the same points outcome (on offense, and shooting more often can lead to more transition for the opponent).

Generally 33% is seen as a poor shooting night from 3.
Generally 50% shooting from 2 is viewed as a good performance.
Therefore if you can get quality shots from 2 or 3 it’s not even close, as a decision. Amazing how fast the game can swing when you hit three 3s in as many possessions.

 

[Eddie Jurak]

I put it in the other thread, but you can shoot 33% from 3 or 50% from 2 and have the same points outcome (on offense, and shooting more often can lead to more transition for the opponent).

Generally 33% is seen as a poor shooting night from 3.
Generally 50% shooting from 2 is viewed as a good performance.
Therefore if you can get quality shots from 2 or 3 it’s not even close, as a decision. Amazing how fast the game can swing when you hit three 3s in as many possessions.

There are some other factors that come into play. More likely to be fouled, to get "and 1s" on 2s. Paul Pierce was a master of this who probably had a mid range game that was as efficient as his 3-point game.

 

[jose melendez]

I put it in the other thread, but you can shoot 33% from 3 or 50% from 2 and have the same points outcome (on offense, and shooting more often can lead to more transition for the opponent).

Generally 33% is seen as a poor shooting night from 3.
Generally 50% shooting from 2 is viewed as a good performance.
Therefore if you can get quality shots from 2 or 3 it’s not even close, as a decision. Amazing how fast the game can swing when you hit three 3s in as many possessions.

That you Jimmy O'Brien?

 

[wade boggs chicken dinner]

So where last series was about using size and at the rim finishing to punish a small team, this is about getting a good/big team in motion to get open rhythm 3s, and force them to either outscore you (against your top D) or eventually go small, possibly opening up the rim some.

Good post and this is a good point. MIL may be strong and physical but they are not particularly fleet of foot so when BOS gets them in rotation, the more the ball swings, the more open shots will be IMO. Getting MIL to rotate will also make them spend energy.

To echo BenHogan, if the Cs don't turn the ball over and limit run outs on missed shots,they should be ok.

 

[BaseballJones]

I’ve always felt that a 3 being “only” one more point than a 2 masks it’s power. The better way to view it is that it is 50% more points, and if instead the NBA scored it 4 within the arc and 6 outside the arc people would see it more clearly.

there’s a reason NBA teams focus on not giving up uncontested threes, and it’s because your average NBA player hits those more often than not. That’s easy enough to see during a shoot around at practice. That the Bucks seemingly are either willing or unable to prevent these looks consistently for the Celts is something to exploit again and again until the Bucks are either forced to adjust or are on a plane home

The bolded is wildly untrue. Go here: https://www.nba.com/stats/teams/shots-closest-defender/?Season=2021-22&SeasonType=Regular Season&CloseDefDistRange=6+ Feet - Wide Open

Sort through the data (regular season, defender 6+ feet away...which is what they classify as wide open), and look at the shooting percentages.

Only 5 teams in the NBA hit 40% or better on uncontested threes:

Atlanta: 42.7%
LAC: 41.8%
Brooklyn: 40.9%
Miami: 40.8%
New York: 40.1%

There wasn't a single team that shot even close to 50% on wide open threes. The NBA average on uncontested threes this season was 38.1%. Not even close to "hitting those more often than not".

Your point is that mathematically it's still a good bet, and of course that's correct. But the average player doesn't come anywhere near hitting uncontested threes (in real games...practice is irrelevant) more often than not.

 

[wilked]

The bolded is wildly untrue. Go here: https://www.nba.com/stats/teams/shots-closest-defender/?Season=2021-22&SeasonType=Regular Season&CloseDefDistRange=6+ Feet - Wide Open

Sort through the data (regular season, defender 6+ feet away...which is what they classify as wide open), and look at the shooting percentages.

Only 5 teams in the NBA hit 40% or better on uncontested threes:

Atlanta: 42.7%
LAC: 41.8%
Brooklyn: 40.9%
Miami: 40.8%
New York: 40.1%

There wasn't a single team that shot even close to 50% on wide open threes. The NBA average on uncontested threes this season was 38.1%. Not even close to "hitting those more often than not".

Your point is that mathematically it's still a good bet, and of course that's correct. But the average player doesn't come anywhere near hitting uncontested threes (in real games...practice is irrelevant) more often than not.

I alluded to it later in the post, but what I meant there is truly uncontested, ie practice shootaround. Truly uncontested in a game is rare, obviously there is a gradient and it’s not binary. You are correct and it’s a good clarification.

 

[wilked]

View: https://youtu.be/nXCol8YeqxQ


This is an extreme of course, one of the best shooters ever and an atypically good session, but when I’ve shown up early or had a chance to watch NBA players just shoot the ball warming up you see any player that is reasonably defined as a “shooter” sinking most of his threes. It’s a shot you’re expected to make with no one on you, and to counter it NBA defenses are built and designed to contest that shot with a hand in the face, not allow the shooter to set himself, etc.

All of which makes it somewhat astounding that the Bucks would set up their defense to allow teams that advantage. I have to think they adjust in Game 3, otherwise the Celts will just rinse and repeat.

 

[lovegtm]

View: https://youtu.be/nXCol8YeqxQ


This is an extreme of course, one of the best shooters ever and an atypically good session, but when I’ve shown up early or had a chance to watch NBA players just shoot the ball warming up you see any player that is reasonably defined as a “shooter” sinking most of his threes. It’s a shot you’re expected to make with no one on you, and to counter it NBA defenses are built and designed to contest that shot with a hand in the face, not allow the shooter to set himself, etc.

All of which makes it somewhat astounding that the Bucks would set up their defense to allow teams that advantage. I have to think they adjust in Game 3, otherwise the Celts will just rinse and repeat.

The bolded is too simplistic. Their defense is set up to allow 3s, including from the corner, but with some amount of contest and out of rhythm for the offense. That lowers the success rate considerably, and also takes away the possession equity of sometimes finding guys on drives or getting fouled.

By extending offensive possessions, relocating, and moving the ball (and the Bucks' D) more, the Celtics were able to:
- get more uncontested 3s, including from the corner
- get those 3s for better shooters
- take those 3s in rhythm (intangible, but seems to help)
- find guys for layups when the Bucks slipped up inside
- get some follows at the rim when layups were missed

These are all things that don't happen when you take the early above-the-break 3 that the Bucks are set up to concede.

The Celtics are probably going to just miss quality 3s at some point in the series, but getting all of the above makes it more likely they'll win 4 and advance, especially if they can keep limiting the Bucks offense as they have been.

 

[Red Averages]

There are some other factors that come into play. More likely to be fouled, to get "and 1s" on 2s. Paul Pierce was a master of this who probably had a mid range game that was as efficient as his 3-point game.

Yes. Also more likely to commit turnovers, draw offensive fouls etc. there are trade-offs outside of just the shooting percentage clearly.

 

[HomeRunBaker]

I’ve always felt that a 3 being “only” one more point than a 2 masks it’s power. The better way to view it is that it is 50% more points, and if instead the NBA scored it 4 within the arc and 6 outside the arc people would see it more clearly.

there’s a reason NBA teams focus on not giving up uncontested threes, and it’s because your average NBA player hits those more often than not. That’s easy enough to see during a shoot around at practice. That the Bucks seemingly are either willing or unable to prevent these looks consistently for the Celts is something to exploit again and again until the Bucks are either forced to adjust or are on a plane home

Bud’s defensive system shuts down the paint and dribble penetration. Of course it can be exploited by making a ton of three’s but it also prevents us from getting into their bench due to foul trouble, keeps us from getting into the penalty early in a quarter and overall keeps us off the line. This is the same system Bud used to win an NBA title two years ago and how he led the Hawks to a 60-win season. This isn’t Doug Moe’s defense from the ‘80s here.

 

[CreightonGubanich]

Right, Bud's defensive system is not math-ignorant. NBA teams average 1.3 points per possession on shots at the rim, and about 1.05 points per possession on non-corner three point attempts (ht: Dan Devine @ The Ringer). They're not trapped in the 80's, they are trying to take away the most efficient scoring chances in basketball. And if they want to play Brook Lopez, they don't have a choice. Lopez is Kanter-level bad at moving his feet on the perimeter. The only place he's an effective defender is planted in the lane. Luckily, he's bigger and longer than Kanter, and a better shot deterrent at the rim. And Giannis is an interior wrecking ball. Putting Portis on the floor in Middleton's place is a double-down on this defensive philosophy, but it certainly leaves them short on wings to stick with Tatum and Brown.

The key for the Celtics is taking the shots the defense is giving them - which means a lot of three pointers - but not necessarily just firing away at the top of the key the first time Brook Lopez drops into the paint on a pick and roll. Those offensively-stagnant, off-the-dribble threes can be decent shots, if you're Jayson Tatum left alone behind the screen. The Bucks are willing to gamble that teams will take those shots - the first decent look shown by the defense - and miss enough of them so that the Bucks can dominate the boards and get Giannis out in transition. Instead, the Celtics need to embrace what Tatum was saying about passing up the good shot to get the great shot. That's not always a great philosophy, but it is against this specific defense. Drive into the paint, but move the ball instead of trying to over-penetrate. Attack the closeout, and do it again. It helps the Celtics that most of their perimeter shooters are threats when they attack those closeouts - Grant Williams in particular has made incredible strides in that area. Otherwise, teams end up just passing the ball around the perimeter while the Bucks wall off the paint. In game 2, the Celtics showed elite offensive execution (for stretches of the game, anyway) in being able to break down that defense.

 

[RSN Diaspora]

So the Celtics are going to take a lot of 3s... BUT... one key is how you manipulate to get them. MIL is great at the rim because they collapse and sell out to protect it, with the hope that either you challenge the shot blockers, or you get stuck, miss the pass and turn it over so they can run for easy buckets. What the Celtics need, and did in game 2, is to be patient, probe with drives, make the right pass, then if needed a 2nd pass or 2nd drive while the driver relocates, get the right looks for the right guys with MIL in motion.

So where last series was about using size and at the rim finishing to punish a small team, this is about getting a good/big team in motion to get open rhythm 3s, and force them to either outscore you (against your top D) or eventually go small, possibly opening up the rim some.

This is a really good counter-point. Adore Marcus though I do, he has the occasional Rondo-esque capacity to make a brilliant play on the defensive end and then follow it up with a bone-headed three-point attempt from way beyond the arc. I don't think we were better by not having him in game two, but not settling for either a three with ample time left on the clock or a last-second heave as the shot clock gets to zero is probably more important than I considered in the opening post. It'll be interesting to see if they stay three-heavy but smart with the passing.

 

[benhogan]

Good article in the Athletic on Celtic shot selection, a lot of it has to do with SCREENS! (get Baynes on hotline)

https://theathletic.com/3294629/2022/05/05/celtics-simple-adjustment-how-setting-better-screens-helped-sink-bucks-film-breakdown/?source=dailyemail&campaign=601983

Amid all the playoff adjustments, sometimes the simplest ones go unnoticed. As much as the Celtics changed before Tuesday night’s Game 2, one of the major improvements they harped on was setting better screens.

“That was another point,” Udoka said. “They obviously want to protect the paint. We shoot over 30 percent of our shots from midrange on the season and only took 17 percent in Game 1. We knew those were open. You don’t have to over-penetrate and drive into the bigs or take the 3 only. We have two elite midrange shooters and we welcome those shots, especially when they are back in that drop. I talked to Jayson, Jaylen specifically about that.

“And we wanted to set better screens and free them up. And we knew we would have those looks. And they knocked them down.”

The Celtics still shot 43 3-point attempts. There, too, the value of a good screen jumped out on the game film

Though the Bucks have a big-time defense, they don’t have a long list of options with both the length and athleticism to bother Tatum and Brown, especially with Khris Middleton out of the lineup. Milwaukee would much rather avoid switches in certain situations that would lead to mismatches. The Celtics know that. And when they get a good screen, they can take full advantage of it,

 

[slamminsammya]

The bolded is wildly untrue. Go here: https://www.nba.com/stats/teams/shots-closest-defender/?Season=2021-22&SeasonType=Regular Season&CloseDefDistRange=6+ Feet - Wide Open

Sort through the data (regular season, defender 6+ feet away...which is what they classify as wide open), and look at the shooting percentages.

Only 5 teams in the NBA hit 40% or better on uncontested threes:

Atlanta: 42.7%
LAC: 41.8%
Brooklyn: 40.9%
Miami: 40.8%
New York: 40.1%

There wasn't a single team that shot even close to 50% on wide open threes. The NBA average on uncontested threes this season was 38.1%. Not even close to "hitting those more often than not".

Your point is that mathematically it's still a good bet, and of course that's correct. But the average player doesn't come anywhere near hitting uncontested threes (in real games...practice is irrelevant) more often than not.

This is also a very flawed way to consider the question of how NBA players would generally fare on wide open looks, because there is a massive selection bias for who is taking these threes. Teams will not help off strong 3 point shooters, meanwhile they are happy to let Daniel Theis or Draymond put up wide open looks from deep.

This points to a much more general constraint on our ability to make conclusions from game data, that we can't learn certain types of common knowledge precisely because if they are commonly understood they will impact the strategy in such a way that we can't observe a lot of counterfactuals, like "what would happen if you started 5 centers?" or "what would Steph Curry's 3p% be if he was wide open for half of them?"

This source of bias is really hard to deal with, it's partly why some of the earlier metrics like WS/48 seemed to imply you should just stack your team with guys who can get offensive rebounds.

 

[BaseballJones]

This is also a very flawed way to consider the question of how NBA players would generally fare on wide open looks, because there is a massive selection bias for who is taking these threes. Teams will not help off strong 3 point shooters, meanwhile they are happy to let Daniel Theis or Draymond put up wide open looks from deep.

This points to a much more general constraint on our ability to make conclusions from game data, that we can't learn certain types of common knowledge precisely because if they are commonly understood they will impact the strategy in such a way that we can't observe a lot of counterfactuals, like "what would happen if you started 5 centers?" or "what would Steph Curry's 3p% be if he was wide open for half of them?"

This source of bias is really hard to deal with, it's partly why some of the earlier metrics like WS/48 seemed to imply you should just stack your team with guys who can get offensive rebounds.

The poster I responded to said that "the average NBA player" hits uncontested threes more often than not. The NBA average on uncontested threes is 38.1%. I understand what you're saying, but how else are we supposed to look at it?

 

[ManicCompression]

I don't understand why a team should force it to the rim when their opponent is giving up wide open threes and your team is actually pretty good at hitting wide open threes

One really good reason is that you want to get your opponent into foul trouble and you're not going to do that without being aggressive. Milwaukee is not a particularly deep team, so getting Lopez or even Portis into foul trouble on drives to the hoop forces them to rely on a weak bench unit. This impacts both the offensive and defensive end of the floor - if Lopez isn't playing as much, you can then find an easier path to buckets near the rim (which we saw was helpful in game two). With Giannis the Destroyer, Blinder of Referees on the other side, it's essential to push the envelope on the foul game to make that war of attrition more equal.

Basketball isn't just a math problem - if it was, why wouldn't every shot be a three pointer? Even the best shooting offenses (and Boston isn't one of those) need to at least threaten the paint in order to open up better shots on the perimeter. It's like running the ball in the football - yes, it's more efficient to pass, but the most efficient thing is play action. if the other team doesn't believe you're going to run the ball effectively, it won't commit resources to stopping it, which mutes the effectiveness of this highly efficient play. One is necessary to maximize the value of the other.

 

[joe dokes]

One really good reason is that you want to get your opponent into foul trouble and you're not going to do that without being aggressive. Milwaukee is not a particularly deep team, so getting Lopez or even Portis into foul trouble on drives to the hoop forces them to rely on a weak bench unit. This impacts both the offensive and defensive end of the floor - if Lopez isn't playing as much, you can then find an easier path to buckets near the rim (which we saw was helpful in game two). With Giannis the Destroyer, Blinder of Referees on the other side, it's essential to push the envelope on the foul game to make that war of attrition more equal.

Basketball isn't just a math problem - if it was, why wouldn't every shot be a three pointer? Even the best shooting offenses (and Boston isn't one of those) need to at least threaten the paint in order to open up better shots on the perimeter. It's like running the ball in the football - yes, it's more efficient to pass, but the most efficient thing is play action. if the other team doesn't believe you're going to run the ball effectively, it won't commit resources to stopping it, which mutes the effectiveness of this highly efficient play. One is necessary to maximize the value of the other.

The Celtics did do quite a bit of driving (and dishing) in Game 2. At some point, I suppose the possible MIL adjustment is to play for the dish, not the driver's shot.

 

[slamminsammya]

The poster I responded to said that "the average NBA player" hits uncontested threes more often than not. The NBA average on uncontested threes is 38.1%. I understand what you're saying, but how else are we supposed to look at it?

Slightly better would be average of averages, with some attempts threshold. I think better than that would be modelling per player wide open 3p% as a function of their overall 3p% and then inferring from that.

 

[BaseballJones]

Slightly better would be average of averages, with some attempts threshold. I think better than that would be modelling per player wide open 3p% as a function of their overall 3p% and then inferring from that.

Ok, feel like doing that and letting us know what you come up with? (that's not snark...that's a serious request)

 

[slamminsammya]

Ok, feel like doing that and letting us know what you come up with? (that's not snark...that's a serious request)

It's on my list. I have just been putting all my NBA ideas off because I don't want to have to figure out a good free NBA data API. I am spending my non work coding time doing important things like using ML to reverse engineer a football management simulation but Ill get around to it.

 

[Cellar-Door]

No player on the Celtics made 50% of his wide open 3s this year, during the post Jan 1 period PP shot 50.6% no other Celtic was close at all.

Wide open NBA game shots still aren't like practice, the reasonable range for a very good NBA shooter is probably 43—47, top shooters might touch 50

Edit I think people overestimate the impact of everything other than very tight strong contests, anything else is a couple percentage points different

 

[ManicCompression]

There's also something about being wide open on a three that puts players off rhythm. In Game 2, Grant got a pass in the corner and had enough time to make a club sandwich, eat it, let it digest, take a shit, and then shoot - seriously, Milwaukee didn't even bother contesting - and it clanged off the rim.

Like, you obviously want to be more open than not, but it should ideally be a pass to the shooting pocket and then immediately out on schedule. When guys get time to marinate, I don't think it has a positive impact on the shot.

 

[Cesar Crespo]

For the Celtics to be ocnsidered an average 3 point shooting team (and for the numbers to suggest this) actually suggest they are better than that. What!?!

Think about how bad the team started from 3 this year. I'd have to look, but I'm guessing from January 1st on, they were in the top 10.

 

[Cellar-Door]

For the Celtics to be ocnsidered an average 3 point shooting team (and for the numbers to suggest this) actually suggest they are better than that. What!?!

Think about how bad the team started from 3 this year. I'd have to look, but I'm guessing from January 1st on, they were in the top 10.

I posted it in another thread... 3rd best 3pt% over that time... tied for 7th highest volume too

 

[lars10]

One really good reason is that you want to get your opponent into foul trouble and you're not going to do that without being aggressive. Milwaukee is not a particularly deep team, so getting Lopez or even Portis into foul trouble on drives to the hoop forces them to rely on a weak bench unit. This impacts both the offensive and defensive end of the floor - if Lopez isn't playing as much, you can then find an easier path to buckets near the rim (which we saw was helpful in game two). With Giannis the Destroyer, Blinder of Referees on the other side, it's essential to push the envelope on the foul game to make that war of attrition more equal.

Basketball isn't just a math problem - if it was, why wouldn't every shot be a three pointer? Even the best shooting offenses (and Boston isn't one of those) need to at least threaten the paint in order to open up better shots on the perimeter. It's like running the ball in the football - yes, it's more efficient to pass, but the most efficient thing is play action. if the other team doesn't believe you're going to run the ball effectively, it won't commit resources to stopping it, which mutes the effectiveness of this highly efficient play. One is necessary to maximize the value of the other.

Problem was... they weren't really getting those calls down low were they? That was the problem.. they were basically blocks and wasted trips. Lopez hasn't been getting called all that much on drives/blocks.. his fouls have been more in going for rebounds ...at least that's what I remember. My frustration with all the drives is that the C's weren't getting calls on drives hardly at all.

Edit: I want the C's to make more outside shots so it forces Lopez to leave the lane a lot more instead of just clogging up the middle.

 

[Cesar Crespo]

I posted it in another thread... 3rd best 3pt% over that time... tied for 7th highest volume too

So using mathemagics, lets say they were 5th in 3 point shooting since January 1st and that's their true talent level. That's 46 games. Considering the team is probably still reverting back to their career norms, I'm perfectly happy with them raining 3s. Looking at the personal, I don't buy that it's an average 3 point shooting team either. They were an average 3 point shooting team in a year where Tatum and Brown were largely off, but are now back on.

 

[Cellar-Door]

BTW basketball is definitely a math problem, and the answer to why every shot isn't a 3 is that a lot of situations give you a high enough percentage chance of making a 2 and/or getting foul that it is worth more than a 3.

A good look at a 3 is usually worth somewhere between 1.02 and 1.2 points per possession depending on shooter.
A good look at the rim is worth more than 1.2 points per possession.
Getting good looks is a product of offensive variety and what the defense does. If a good defense sells out to prevent you from getting good rim attempts, and can do it without fouling then your shot selection will lean towards 3s. If the defense can't protect the rim well (see BKN) or fouls a lot (also BKN) then your shot selection will lean more heavily towards the paint.

 

[RSN Diaspora]

If we assume a hungrier and more aggressive Miami squad tomorrow night (and less friendly officials), then we are going to have to improve beyond the arc. Last night, it was only inside the paint where we were above the league average in shots made. When you get to the line as often as we did and stifled Miami in two-point range the way we did, it doesn't much matter. But I suspect Miami will get more calls on their home court and will adjust their post and mid-range game and be more competitive. Assuming that happens, we're gonna need to be a lot better than this from three:

 

[Eddie Jurak]

Not too worried about White Tatum Grant and Brown shooting a collective 4-26 again.