Basketball Analytics TImeline

[BenAlamar]

I am collecting a timeline of important moments in basketball analytics. What I have compiled so far is here: https://docs.google.com/spreadsheets/d/ ... sp=sharing

Please add any event related to hires, innovations, advances in uses - whatever comes to mind as important for how basketball analytics has grown over the last 20 years - with this form: https://forms.gle/YwoXkAV4HScguZ9Z8.

[Crow]

I not immediately sure on dates, something the form makes mandatory with no open field option.

Maybe I'll use the form later but right now I just want to try to remember events:

Call here for metrics to have individual and global +/- components. ProTrade example.

Earliest uses of raw +/-.
First APM.
Wayne Winston comments on Kevin Durant.
First RAPM.
ESPN publishing RPM.

Debates on RAPM.
Debates on Wins Produced.
First Prior informed RAPM.
First Prior informed RAPM with tracking data.
RAPM factors. Nbashotcharts.com.
Joe Sill and later publications of RAPM methodology.

Lineup and pair RAPM.
Other possible RAPM splits.

A Starting Point on NBA analytics.

Other books.

Rosenbaum hired by Cavs.
And the other PhDs.

Lashro hired by Nuggets.
Zarin as Assistant GM.
Chaikin with Heat. Other game simulators.

BRef starts.
Lineup data.
The accessible stat finder.

Warriors listening to Gefland suggestion.

Small ball centers.
3 pt centric offenses.
More general Moreyball. The corner 3.

Player value calculations- Impact / salary.

2 PG lineups.

Concept and measurement of creation
shot quality
rim protection
spacing

JQAS.

Coaching RAPM.

Admitting that overall defensive quality measurements have been stone age.

My insistence that NBA lineup management practices are stone age. Too chaotic, not concentrated enough testing / usage.

Leaguewide lineup big data analysis and lineup clustering.

7 seconds or less.

BPM. EPM, Darko, LEBRON. Others.

Draft Pick value charts.

Pelton to ESPN.

CBAfaq.

Dunc'd On podcast. Explosion of podcasts.

Nylon Calculus.

BBall-Index.com

NBA analysis in academic journals.

Player types via cluster analysis.

"Position-less basketball"
The modern 4. Dirk, etc.
3&D wings.

College basketball stat websites: KenPom, barttorvik, etc.

International basketball stat websites. Youth basketball stat websites.

PER and other first gen metrics.

TS%.

Rates other than per game.

Nba.com.

Webscraping.

82games article Blueprint(s) for an NBA Championship Team
And viewtopic.php?t=107
And viewtopic.php?t=109

Usage and beyond basic usage.

The forum hack/ destruction of 2011.

Late game tactics.

Network analysis of plays.

Oliver's Journal of Basketball Studies. Hollinger's Alley Oop.

Annual guides.

Real GM and many fan boards and the explosion of analysis developed at fan sites.

Basketball Intelligence.

The 10ish person analytic shop of Washington Wizards and lack of team success and visible impact to date.

The "nerd" hostility / backlash. Attacks from Charles Barkley, Stan Van Gundy, etc.

Moneyball book and subsequent articles about "analytics" applied to basketball. Earliest and common use of term?

Extended boxscores.

The Rodman case / argument.

Analytics based injury management.

Analytics based commercial and collegiate sports management programs.

Fantasy basketball. Spread and diversification of betting and betting analysis.

Broader / easier use of R and Python.

The coming AI app revolution.

Application of machine learning.

Shot charts. Awareness of mid-range data.
The rise of visualizations. With little or no accompanying explanation / analysis.

StatMuse and other history search efforts.

Trade machines.

The evolution of analytics job titles / functions.

Players using analytics. KD hiring his own guy for at least awhile. Player agents using analytics.

Metric retrodiction studies.

Attention to error estimates.
The padding method.

The wisdom in the crowd. Superforecasting.

Attention to communication of analytics.

Diversification in analytics staff by demographics, skills. (How about more by age?)

APBRmetrics fadeaway with hires and other outlets. The lack of a stronger ongoing SABRmetrics like hub.

viewtopic.php?t=8871

assisted and unassisted fg%s.
Antoine Walker case.
What is an assist really worth?
Rebounding diminishing returns.

4 factor importance viewtopic.php?t=8851

Moving beyond linear models

Ed Kupfer.

viewtopic.php?t=8733#p21234

Game theory in general and specifically for play sequencing.

Basketball tracking data from broadcast video.

Editors who accepted, fought for and refined analytic content at top mainstream media outlets.

Sloan research papers. Hackathon.

Collegiate sports analytics clubs.

Aging curves.

Pbpstats.com

2k use of nba "dna"

The long 3.

BBallBreakdown play from behind defense.

Adjustment of data from garbage time.

Player similarity algorithms.

Scoring "skill curves"

Impact of home / away scorekeeping.

Awareness of impact of altitude

The Russell Westbrook case.

Ownership reaction to analytics / "arms race"

Explanation of analytics to casuals (and kids)

Referee stats

Understanding linkages of offense and defense (offensive rebounding / transition defense, by shot type, etc.)

Simple adjusted lineup data.

Limitations of raw lineup data.

Evaluation of the Triangle Offense.

NESSIS and other conferences.

Replacement player level and implications.

Pot calling kettle black

Alternate WinScore / revised Wins Produced

The "Coaching Analytics" position in the ecosystem.

Rich Cho and the Microsoft staffer creation of a list of all possible legal trades.

The arrival of financial quant owners.

Lacob's "light years ahead" statement.

Whoever got Popovich to accept the 3pt revolution despite his hatred of it.

Some analyst departures as important as arrivals.

The rise of a League Office analytic group.

Analysts directly advising owners.

The lesser known influence of analytics on Presti decisionmaking.

The rise of for hire sports anaytics groups, some with unique data products.

Confidentiality agreements.

[DSMok1]

Ben--would you consider expanding this to 25 years?

I feel the founding of the original APBR analysis group on Yahoo groups by Dean Oliver in 2001 (https://web.archive.org/web/20130314110 ... /message/1) was a pretty important moment.

Then the founding of 82games.com by Roland Beech in 2002.

[BenAlamar]

All great suggestions - thank you - will try to track down the relevant dates.

I not immediately sure on dates, something the form makes mandatory with no open field option.

Maybe I'll use the form later but right now I just want to try to remember events:

My call here for metrics to have individual and global +/- components. ProTrade example.

Earliest uses of raw +/-.
First APM.
Wayne Winston comments on Kevin Durant.
First RAPM.
ESPN publishing RPM.

Debates on RAPM.
Debates on Wins Produced.
First Prior informed RAPM.
First Prior informed RAPM with tracking data.
RAPM factors. Nbashotcharts.com.
Joe Sill and later publication of RAPM methodology.

Lineup and pair RAPM
Other possible RAPM splits.

A Starting Point on NBA analytics.

Other books.

Rosenbaum hired by Cavs.
Lashro hired by Nuggets.
Zarin as Assistant GM.
Chaikin with Heat. Other game simulators

BRef starts.
Lineup data.

Warriors listening to Gefland suggestion. (Celtics not listening to me in Finals last season.)

Small ball centers.
3 pt centric offenses.
More general Moreyball. The corner 3.

Player value calculations- Impact / salary.

2 PG lineups.

Concept and measurement of creation
shot quality
rim protection
spacing

JQAS.

Coaching RAPM.

Admitting that overall defensive quality measurements have been stone age.

My insistence that NBA lineup management practices are stone age. Too chaotic, not concentrated enough testing / usage.

7 seconds or less.

BPM. EPM, Darko, LEBRON. Others.

Draft Pick value charts.

Pelton to ESPN.

CBAfaq.

Dunc'd On podcast. Explosion of podcasts.

Nylon Calculus.

BBall-Index.com

NBA analysis in academic journals.

Player types via cluster analysis.

"Position-less basketball"
The modern 4. Dirk, etc.
3&D wings.

College basketball stat websites: KenPom, barttorvik, etc.

International basketball stat

websites. Youth basketball stat websites.

PER and other first gen metrics.

TS%.

Rates other than per game.

Nba.com.

Webscraping.

82games article Blueprint(s) for an NBA Championship Team

Usage and beyond basic usage.

The forum hack/ destruction of 2011.

Late game tactics.

Network analysis of plays.

Oliver's Journal of Basketball Studies. Hollinger's Alley Oop.

Annual guides.

Real GM and many fan boards and the explosion of analysis developed at fan sites.

Basketball Intelligence.

The 10ish analytic shop of Washington Wizards and lack of team success and visible impact to date.

The "nerd" hostility / backlash. Attacks from Charles Barkley, Stan Van Gundy, etc.

Moneyball book and subsequent articles about "analytics" applied to basketball. Earliest and common use of term?

Extended boxscores.

The Rodman case / argument.

Analytics based injury management.

Analytics based commercial and collegiate sports management programs.

Fantasy basketball. Spread and diversification of betting.

Broader / easier use of R and Python. The coming AI app revolution.

Application of machine learning.

Shot charts.
The rise of visualizations. With little or no accompanying explanation / analysis.

StatMuse.

Trade machines.

The evolution of analytics job titles / functions.

Players using analytics. KD hiring his own guy for at least awhile. Player agents using B analytics.

Metric reproduction studies.

Attention to error estimates.
The padding method.

The wisdom in the crowd. Superforecasting.

Attention to communication of analytics.

Diversification in analytics staff by demographics, skills. (How about more by age?)

APBRmetrics fadeaway with hires and other outlets. The lack of a stronger obgoing SABRmetrics like hub.

viewtopic.php?t=8871

assisted and unassisted fg%s.
Antoine Walker case.
What is an assist really worth?
Rebounding diminishing returns.

4 factor importance viewtopic.php?t=8851

Moving beyond linear models

Ed Kupfer.

viewtopic.php?t=8733#p21234

Game theory in general and specifically for play sequencing.

Basketball tracking data from broadcast video.

Editors who accepted, fought for and refined analytic content at top mainstream media outlets.

[BenAlamar]

Agree entirely on the importance of those events. The reason for the 20 years is to tie it to the publication of Basketball on Paper - there is a definitely a prologue that includes these.

Ben--would you consider expanding this to 25 years?

I feel the founding of the original APBR analysis group on Yahoo groups by Dean Oliver in 2001 (https://web.archive.org/web/20130314110 ... /message/1) was a pretty important moment.

Then the founding of 82games.com by Roland Beech in 2002.

[Crow]

Glad that you find items on the list useful.

I know it jumps around and you might not touch on all.

[Crow]

I mentioned a few items not really adequately pursued yet. That could be a large addition to the list or a new list.

When will NBA lineup management reach a high degree of optimization? 2030? 2040? Very little obvious recent change/ improvement. Very little public mention of topic by teams and very little attention to the topic by media (outside my efforts and a few others).

[Mike G]

Code: Select all

2004	Publication of Basketball on Paper
2004	APBRMetrics Yahoo Group Created

The APBR-Analysis Yahoo group was around for a few years before BOP, and many of Dean's topics were debated there.
That Yahoo group was a spin-off from the original APBR Yahoo group. Enough stat-heads there got on the nerves of a few old-timers, and DeanO started the Analysis group.

2004 may be the year this APBRMetrics forum was created by Kevin Pelton.

[BenAlamar]

Yes - 2004 was when KP created the Yahoo group

[DSMok1]

The Yahoo groups were from 2001 to 2004. This APBRMetrics group moved to the precursor of this forum in 2004:

http://godismyjudgeok.com/DStats/APBRme ... 7&amp.html

[Mike G]

Actually APBR Yahoo group was around for a few years before Dean's Analysis spinoff. 1997 or '98 I'd guess.

[glwall3]

Here is an interesting article on a computer scientist that developed an algorithm to rank players in the late 50s:

https://keithlyons.me/blog/2018/09/25/d ... -in-sport/

[Crow]

That was an early start.

[bchaikin]

Chaikin with Heat. Other game simulators.

wait a minute...

i graduated from case institute of technology (case western reserve university) in 1981, same school as donald knuth, actually played on the school basketball team (spartans, under coach bill sudeck), created an nba computer game simulation, and have worked in the nba for 3+ decades (since the early 1990s) doing player evaluation using said simulation, have two championship rings...

is alamar writing a book?...

[Crow]

Alamar could well be writing or considering another book of some sort, for someday; but the immediate reason for the call for help with basketball analytics timeline, as I understand it, was a Sloan Conference panel on the topic, especially for his introduction of it as moderator. It is available online. https://youtu.be/6zBmxq4L6nc?feature=shared
A modest number of items from this list (or similar) were mentioned.