Some statistics questions

Moderator: Doctor MJ

LesGrossman
Head Coach
Posts: 6,158
And1: 4,114
Joined: Mar 24, 2014

Some statistics questions 

Post#1 » by LesGrossman » Sun Nov 13, 2022 10:45 am

* where would I look up the numbers of threes taken of, let’s say, the playoff teams in 2008 vs. 2021?
* are there numbers for open vs. contested threes in 2008?
* does the point difference matter at any point within the tournament officially?
* is there an analysis of 3‘s taken over the 48 minutes on average? Meaning, is there an even distribution of threes taken per minute and team or does it rise or fall towards the end of the game?
Pray for Israel
Peace in Jerusalem

Fan of the game of Basketball, no matter the team, league or players. Opposed to all sorts of person cult and show/entertainment/marketing over substance.
User avatar
ijspeelman
Forum Mod - Cavs
Forum Mod - Cavs
Posts: 1,729
And1: 859
Joined: Feb 17, 2022
Contact:
   

Re: Some statistics questions 

Post#2 » by ijspeelman » Sat Nov 26, 2022 8:34 pm

LesGrossman wrote:* where would I look up the numbers of threes taken of, let’s say, the playoff teams in 2008 vs. 2021?
* are there numbers for open vs. contested threes in 2008?
* does the point difference matter at any point within the tournament officially?
* is there an analysis of 3‘s taken over the 48 minutes on average? Meaning, is there an even distribution of threes taken per minute and team or does it rise or fall towards the end of the game?


* where would I look up the numbers of threes taken of, let’s say, the playoff teams in 2008 vs. 2021?
https://www.basketball-reference.com/playoffs/NBA_2008.html
https://www.basketball-reference.com/playoffs/NBA_2021.html

* are there numbers for open vs. contested threes in 2008?
public tracking data only goes back to 2013-14. https://www.nba.com/stats/players/shots-closest-defender-10

* does the point difference matter at any point within the tournament officially?
what do you mean by this?

* is there an analysis of 3‘s taken over the 48 minutes on average? Meaning, is there an even distribution of threes taken per minute and team or does it rise or fall towards the end of the game?

I have never seen one done. Basketball reference has shot charts like the below for each individual player which can show when they take threes, but none for entire teams or the league.
https://www.basketball-reference.com/players/h/hausesa01/shooting/2023
LesGrossman
Head Coach
Posts: 6,158
And1: 4,114
Joined: Mar 24, 2014

Re: Some statistics questions 

Post#3 » by LesGrossman » Thu Dec 1, 2022 10:50 pm

Thank you for replying, i didnt expect anything here any more.

ijspeelman wrote:
LesGrossman wrote:* where would I look up the numbers of threes taken of, let’s say, the playoff teams in 2008 vs. 2021?
* are there numbers for open vs. contested threes in 2008?
* does the point difference matter at any point within the tournament officially?
* is there an analysis of 3‘s taken over the 48 minutes on average? Meaning, is there an even distribution of threes taken per minute and team or does it rise or fall towards the end of the game?


* where would I look up the numbers of threes taken of, let’s say, the playoff teams in 2008 vs. 2021?
https://www.basketball-reference.com/playoffs/NBA_2008.html
https://www.basketball-reference.com/playoffs/NBA_2021.html

Yes, thank you.
ijspeelman wrote:* are there numbers for open vs. contested threes in 2008?
public tracking data only goes back to 2013-14. https://www.nba.com/stats/players/shots-closest-defender-10

Excellent, thank you. I can use that in my thesis.
ijspeelman wrote:* does the point difference matter at any point within the tournament officially?
what do you mean by this?

Its specific to the NBA compared to other, mostly european sports and leagues, that the point difference does not matter in the placement taht leads to the playoffs. In football for example, if you have even points, the tie breaker is usually the goal difference.

That is the reason why teams have their rookies play garbage time - it doesnt matter wether you lose by 1 or 50. At least thats what i think - adn what the question was about. Does it matter by how many points teams won, at any point int he NBA season?
ijspeelman wrote:* is there an analysis of 3‘s taken over the 48 minutes on average? Meaning, is there an even distribution of threes taken per minute and team or does it rise or fall towards the end of the game?

I have never seen one done. Basketball reference has shot charts like the below for each individual player which can show when they take threes, but none for entire teams or the league.
https://www.basketball-reference.com/players/h/hausesa01/shooting/2023

Yes, i was afraid thats the case.

Also, am i right to assume that a missed shot with foul is not counted at any point? If you pull game data from NBA.com, those do not appear as misses, right?
Pray for Israel
Peace in Jerusalem

Fan of the game of Basketball, no matter the team, league or players. Opposed to all sorts of person cult and show/entertainment/marketing over substance.
User avatar
ijspeelman
Forum Mod - Cavs
Forum Mod - Cavs
Posts: 1,729
And1: 859
Joined: Feb 17, 2022
Contact:
   

Re: Some statistics questions 

Post#4 » by ijspeelman » Fri Dec 2, 2022 3:25 am

LesGrossman wrote:Its specific to the NBA compared to other, mostly european sports and leagues, that the point difference does not matter in the placement taht leads to the playoffs. In football for example, if you have even points, the tie breaker is usually the goal difference.

That is the reason why teams have their rookies play garbage time - it doesnt matter wether you lose by 1 or 50. At least thats what i think - adn what the question was about. Does it matter by how many points teams won, at any point int he NBA season?


TLDR: No, it does not matter.

According to https://ak-static-int.nba.com/wp-content/uploads/sites/2/2017/06/NBA_Tiebreaker_Procedures.pdf,
"Tiebreaker Basis – 2 Teams Tied
(-) Tie breaker not needed (better overall winning percentage)
(1) Better winning percentage in games against each other
(2) Division leader wins a tie over a team not leading a division
(3) Division won-lost percentage (only if teams are in same division)
(4) Conference won-lost percentage
(5) Better winning percentage against teams eligible for the playoffs in own
conference (including teams that finished the regular season tied for a playoff
position)
(6) Better winning percentage against teams eligible for the playoffs in other
conference (including teams that finished the regular season tied for a playoff
position)
(7) Better net result of total points scored less total points allowed against all
opponents (“point differential”)

Tiebreaker Basis – Three or More Teams Tied
(-) Tie breaker not needed (better overall winning percentage)
(1) Division leader wins tie from team not leading a division (this criterion is
applied regardless of whether the tied teams are in the same division)
(2) Better winning percentage in all games among the tied teams
(3) Division won-lost percentage (only if all teams are in same division)
(4) Conference won-lost percentage
(5) Better winning percentage against teams eligible for the playoffs in own
conference (including teams that finished the regular season tied for a playoff
position)
(6) Better net result of total points scored less total points allowed against all
opponents (“point differential”)

Playoff Tie-Break Procedures
(1) (a) Ties to determine the division winners must be broken before any other
ties.
(b) When a tie must be broken to determine a division winner, the results of
the tie-break shall be used to determine only the division winner, and not
for any other purpose.
(c) When three or more teams are tied, teams will be assigned a seeding
based on the multi-team tiebreaker basis above. If three or more teams
have equivalent records, the first tiebreaker criteria that creates
differentiation applies. If two or more teams remain tied after applying
that criteria, the team(s) that are not tied are assigned their seed and the
tiebreaker criteria restarts with the remaining teams.
Example: Teams A, B, and C are all tied at the end of the season. None of
the teams are division winners, and Team A had a 4-0 record in common
games, and Teams B and C had a 1-3 record in common games. Team A
would receive the highest seed, and the 2 Team tiebreaker would then
apply to Teams B and C to determine their relative seed.
(2) If application of the criteria in subparagraph a. or b. does not result in the
breaking of a tie, the playoff positions of the tied teams will be
determined by a random drawing."

LesGrossman wrote:Also, am i right to assume that a missed shot with foul is not counted at any point? If you pull game data from NBA.com, those do not appear as misses, right?


Shots where fouls occur are only counted as a field goal attempt when the attempt is successful (it will then also count as a made field goal). Otherwise, the field goal attempt is not counted.

If you just look at data alone, its like the foul was never taken. For example, in the game that happened today, Davis Bertans attempted two field goal attempts, but on one he got fouled so the shot attempt did not count. If you wanted to see that attempt based on nba.com's internal stat video viewer you would need to look at the foul of the player who fouled on the shot (see here: https://www.nba.com/stats/events?CFID=&CFPARAMS=&GameEventID=205&GameID=0022200325&Season=2022-23&dir=A&flag=1&sort=dsc&title=Bey%20S.FOUL%20(P1.T4)%20(P.Fraher)). These are specified as shooting fouls, but only by a player and not on a specific player.
LesGrossman
Head Coach
Posts: 6,158
And1: 4,114
Joined: Mar 24, 2014

Re: Some statistics questions 

Post#5 » by LesGrossman » Fri Dec 2, 2022 9:18 am

Excellent, i will use this. THank you!
Pray for Israel
Peace in Jerusalem

Fan of the game of Basketball, no matter the team, league or players. Opposed to all sorts of person cult and show/entertainment/marketing over substance.
LesGrossman
Head Coach
Posts: 6,158
And1: 4,114
Joined: Mar 24, 2014

Re: Some statistics questions 

Post#6 » by LesGrossman » Sun Dec 4, 2022 4:01 am

By the way, i made the mistake to only take a quick look at first, since you summarized it in the TLDR. As a matter of fact though, thats not quite accurate from a scientific standpoint - the point difference does matter as the least relevant tie breaker, see below. I wonder though wether this rule was ever, i mean EVER applied in the history of the league.
ijspeelman wrote:
LesGrossman wrote:Its specific to the NBA compared to other, mostly european sports and leagues, that the point difference does not matter in the placement taht leads to the playoffs. In football for example, if you have even points, the tie breaker is usually the goal difference.

That is the reason why teams have their rookies play garbage time - it doesnt matter wether you lose by 1 or 50. At least thats what i think - adn what the question was about. Does it matter by how many points teams won, at any point int he NBA season?


TLDR: No, it does not matter.


(7) Better net result of total points scored less total points allowed against all
opponents (“point differential”)

Tiebreaker Basis – Three or More Teams Tied

(6) Better net result of total points scored less total points allowed against all
opponents (“point differential”)

Pray for Israel
Peace in Jerusalem

Fan of the game of Basketball, no matter the team, league or players. Opposed to all sorts of person cult and show/entertainment/marketing over substance.
User avatar
ijspeelman
Forum Mod - Cavs
Forum Mod - Cavs
Posts: 1,729
And1: 859
Joined: Feb 17, 2022
Contact:
   

Re: Some statistics questions 

Post#7 » by ijspeelman » Sun Dec 4, 2022 4:28 am

LesGrossman wrote:By the way, i made the mistake to only take a quick look at first, since you summarized it in the TLDR. As a matter of fact though, thats not quite accurate from a scientific standpoint - the point difference does matter as the least relevant tie breaker, see below. I wonder though wether this rule was ever, i mean EVER applied in the history of the league.
ijspeelman wrote:
LesGrossman wrote:Its specific to the NBA compared to other, mostly european sports and leagues, that the point difference does not matter in the placement taht leads to the playoffs. In football for example, if you have even points, the tie breaker is usually the goal difference.

That is the reason why teams have their rookies play garbage time - it doesnt matter wether you lose by 1 or 50. At least thats what i think - adn what the question was about. Does it matter by how many points teams won, at any point int he NBA season?


TLDR: No, it does not matter.


(7) Better net result of total points scored less total points allowed against all
opponents (“point differential”)

Tiebreaker Basis – Three or More Teams Tied

(6) Better net result of total points scored less total points allowed against all
opponents (“point differential”)



Point differential matters but not three point differential. Yeah, three pointers add to the point differential, but more threes or less threes does not directly correlate with a higher or lower point differential.
LesGrossman
Head Coach
Posts: 6,158
And1: 4,114
Joined: Mar 24, 2014

Re: Some statistics questions 

Post#8 » by LesGrossman » Sun Dec 4, 2022 2:28 pm

ijspeelman wrote:
LesGrossman wrote:By the way, i made the mistake to only take a quick look at first, since you summarized it in the TLDR. As a matter of fact though, thats not quite accurate from a scientific standpoint - the point difference does matter as the least relevant tie breaker, see below. I wonder though wether this rule was ever, i mean EVER applied in the history of the league.
ijspeelman wrote:
TLDR: No, it does not matter.


(7) Better net result of total points scored less total points allowed against all
opponents (“point differential”)

Tiebreaker Basis – Three or More Teams Tied

(6) Better net result of total points scored less total points allowed against all
opponents (“point differential”)



Point differential matters but not three point differential. Yeah, three pointers add to the point differential, but more threes or less threes does not directly correlate with a higher or lower point differential.

Of course. The thesis is that teams whcih lie behind in a late stage of the game rise their risk by taking more threes, because the loss is limited to losing the game (based on the premise that the goal difference does not matter). With this tie breaker rule this isnt entirely true. However we all know that teams play garbage time with their bench warmers so they seem to ignore this rule. I wonder if i could look up wether this rule was ever applied at all...
Pray for Israel
Peace in Jerusalem

Fan of the game of Basketball, no matter the team, league or players. Opposed to all sorts of person cult and show/entertainment/marketing over substance.
User avatar
ijspeelman
Forum Mod - Cavs
Forum Mod - Cavs
Posts: 1,729
And1: 859
Joined: Feb 17, 2022
Contact:
   

Re: Some statistics questions 

Post#9 » by ijspeelman » Sun Dec 4, 2022 2:54 pm

LesGrossman wrote:
ijspeelman wrote:
LesGrossman wrote:By the way, i made the mistake to only take a quick look at first, since you summarized it in the TLDR. As a matter of fact though, thats not quite accurate from a scientific standpoint - the point difference does matter as the least relevant tie breaker, see below. I wonder though wether this rule was ever, i mean EVER applied in the history of the league.


Point differential matters but not three point differential. Yeah, three pointers add to the point differential, but more threes or less threes does not directly correlate with a higher or lower point differential.

Of course. The thesis is that teams whcih lie behind in a late stage of the game rise their risk by taking more threes, because the loss is limited to losing the game (based on the premise that the goal difference does not matter). With this tie breaker rule this isnt entirely true. However we all know that teams play garbage time with their bench warmers so they seem to ignore this rule. I wonder if i could look up wether this rule was ever applied at all...


I will say, in my experience, that finding sources for old NBA rules is very hard. It seems a lot of the old rules have been lightly forgotten to time (even the illegal defense rule established in 1974ish that went to 2003. It’s basically impossible to find specific details on it).

Return to Statistical Analysis