Case Study 1: The Pace-and-Space Revolution of the 2015-18 Golden State Warriors
Executive Summary
The Golden State Warriors' dominance from 2014-15 through 2018-19 reshaped how NBA teams approach pace and offensive efficiency. This case study examines how the Warriors' statistical profile illustrated the power of pace-adjusted analysis, why traditional counting statistics failed to capture their revolutionary style, and how rate-based metrics helped explain their historic success.
Background: A Statistical Outlier
The Traditional View
In 2014-15, the Warriors won 67 games and the NBA championship. By traditional counting statistics, they appeared strong but not historically dominant:
| Category | Warriors | League Leader | Rank |
|---|---|---|---|
| Points Per Game | 110.0 | DAL (110.6) | 2nd |
| Rebounds Per Game | 44.7 | DET (45.2) | 6th |
| Assists Per Game | 27.4 | GSW | 1st |
| Field Goals Made | 41.2 | GSW | 1st |
Solid numbers, but 110 PPG hardly seemed revolutionary. Many teams had scored more in previous seasons.
The Rate-Based Reality
However, rate statistics revealed something extraordinary:
| Metric | Warriors | League Avg | Rank |
|---|---|---|---|
| Offensive Rating | 111.6 | 105.6 | 1st |
| Defensive Rating | 101.4 | 105.6 | 1st |
| Net Rating | +10.1 | 0.0 | 1st |
| Pace | 98.3 | 96.0 | 8th |
The Warriors' +10.1 net rating was historically elite. They outscored opponents by 10.1 points per 100 possessions while playing at above-average pace.
Pace Analysis: Playing Faster with Purpose
Pace Trends (2014-2019)
| Season | Warriors Pace | League Avg | Differential |
|---|---|---|---|
| 2014-15 | 98.3 | 96.0 | +2.3 |
| 2015-16 | 99.3 | 95.8 | +3.5 |
| 2016-17 | 100.9 | 96.4 | +4.5 |
| 2017-18 | 99.4 | 97.3 | +2.1 |
| 2018-19 | 100.4 | 100.0 | +0.4 |
The Warriors consistently played faster than league average, but their pace advantage diminished as other teams adopted similar strategies.
Why Pace Mattered
1. Transition Opportunities
The Warriors generated an unusually high percentage of transition possessions:
| Season | Transition Freq | League Avg | Points per Trans Poss |
|---|---|---|---|
| 2015-16 | 18.2% | 15.1% | 1.28 |
| 2016-17 | 17.8% | 15.4% | 1.31 |
Transition plays yielded approximately 1.28-1.31 points per possession versus 1.00-1.05 in halfcourt offense.
2. Defensive Rebounds to Fast Breaks
The Warriors' defensive rebounding strategy prioritized quick outlets: - DREB% was not elite (74.2%, 14th) - But their time from rebound to halfcourt crossing averaged 3.2 seconds (2nd fastest)
This trade-off—slightly worse defensive rebounding for faster transition—only made sense in pace-adjusted context.
3. Shooting Efficiency Maintained at High Pace
The Warriors defied the conventional wisdom that faster pace reduces efficiency:
| Team | Pace | eFG% | TS% |
|---|---|---|---|
| GSW 2015-16 | 99.3 | 55.4% | 58.5% |
| League Avg | 95.8 | 50.0% | 54.0% |
They played 3.5 possessions faster per game while maintaining elite efficiency—a combination that generated their historic +10.7 net rating that season.
The 2015-16 Season: 73 Wins Through the Lens of Rate Stats
Record-Breaking Production
The 73-9 Warriors set numerous records. Traditional statistics captured some: - 110.0 PPG (2nd in league) - 24.1 APG (2nd) - 9.4 3PM per game (1st by wide margin)
But rate statistics revealed the full picture:
| Metric | 2015-16 Warriors | All-Time Rank |
|---|---|---|
| Net Rating | +10.76 | 3rd (since 1970) |
| Offensive Rating | 114.5 | 1st (since 1970) |
| eFG% | 55.4% | 1st (since 1970) |
Points Per Game vs. Offensive Rating
Why did the Warriors rank only 2nd in PPG despite having the best offense ever by ORtg?
Answer: Pace Effects on Counting Stats
Let's compare: - Warriors: 110.0 PPG at 99.3 pace = 110.8 ORtg - Dallas: 110.6 PPG at 100.2 pace = 110.4 ORtg
Dallas scored 0.6 more points per game but was less efficient because they used 0.9 more possessions per game to do so.
Offensive Rating Decomposition
The Warriors' 114.5 ORtg came from:
| Component | Warriors | League Avg | Advantage |
|---|---|---|---|
| eFG% | 55.4% | 50.0% | +5.4% |
| TOV% | 13.0% | 14.0% | -1.0% |
| OREB% | 24.4% | 23.1% | +1.3% |
| FT Rate | 0.209 | 0.220 | -0.011 |
Their massive eFG% advantage (55.4% vs 50.0%) drove the historic offense. Low turnover rate and decent offensive rebounding helped. They drew fewer free throws than average—but didn't need them given their shooting efficiency.
The Three-Point Dimension
Volume in Context
| Season | Warriors 3PA | League Avg | Differential |
|---|---|---|---|
| 2014-15 | 27.0 | 22.4 | +4.6 |
| 2015-16 | 31.6 | 24.1 | +7.5 |
| 2016-17 | 30.5 | 27.0 | +3.5 |
| 2017-18 | 29.3 | 29.0 | +0.3 |
The Warriors pioneered high-volume three-point shooting, but the league caught up. By 2017-18, they were merely average in attempts.
Efficiency Maintained at Volume
What made the Warriors special wasn't just volume—it was maintaining elite efficiency at high volume:
| Season | Warriors 3P% | League Avg | Rank |
|---|---|---|---|
| 2014-15 | 39.8% | 35.0% | 1st |
| 2015-16 | 41.6% | 35.4% | 1st |
| 2016-17 | 38.3% | 35.8% | 1st |
| 2017-18 | 39.1% | 36.2% | 1st |
Expected Value Analysis
Warriors 3-point expected value (2015-16): - 31.6 3PA * 41.6% * 3 = 39.4 expected points from threes per game
League average 3-point expected value: - 24.1 3PA * 35.4% * 3 = 25.6 expected points from threes per game
Differential: +13.8 expected points per game just from three-point shooting.
Per-Possession Player Analysis
Stephen Curry: Per-100 Possession Profile
Curry's 2015-16 MVP season in rate terms:
| Metric | Raw Per Game | Per 36 | Per 100 Poss |
|---|---|---|---|
| Points | 30.1 | 31.7 | 38.4 |
| Assists | 6.7 | 7.1 | 8.5 |
| 3PM | 5.1 | 5.4 | 6.5 |
His 38.4 points per 100 possessions was historically extraordinary. Only a handful of seasons in NBA history exceeded this rate.
Efficiency at Elite Usage:
| Metric | Curry 2015-16 | League Avg |
|---|---|---|
| Usage Rate | 32.6% | 20.0% |
| TS% | 66.9% | 54.0% |
| eFG% | 63.0% | 50.0% |
A player maintaining 66.9% TS% at 32.6% usage was unprecedented. The expected efficiency decline at high usage simply didn't apply to Curry that season.
Draymond Green: Rate Stats Reveal Hidden Value
Green's 2015-16 traditional stats seemed merely good: - 14.0 PPG, 9.5 RPG, 7.4 APG
But rate-based analysis revealed elite impact:
| Metric | Green On | Green Off | Differential |
|---|---|---|---|
| Offensive Rating | 117.6 | 107.8 | +9.8 |
| Defensive Rating | 100.2 | 106.5 | -6.3 |
| Net Rating | +17.4 | +1.3 | +16.1 |
The +16.1 on/off differential suggested Green was as impactful as players with far superior counting statistics.
Pace-Adjusted Historical Comparison
The 73-Win Warriors vs. 72-Win Bulls
| Category | 1995-96 Bulls | 2015-16 Warriors |
|---|---|---|
| Wins | 72 | 73 |
| PPG | 105.2 | 114.9 |
| Opp PPG | 92.9 | 104.1 |
| Pace | 91.1 | 99.3 |
| Net Rating | +12.2 | +10.76 |
Pace-Adjusted Analysis:
The Bulls played 8.2 fewer possessions per game. Adjusting Warriors' stats to Bulls' pace: - Projected PPG: 114.9 * (91.1/99.3) = 105.4 - Projected Opp PPG: 104.1 * (91.1/99.3) = 95.5
At Bulls' pace, the Warriors would have averaged 105-96 scores—very similar to Chicago's 105-93 profile.
Net Rating Comparison:
Despite lower net rating (+10.76 vs +12.2), the Warriors were remarkably close to the Bulls' dominance level. The difference came from: - Bulls played slower, inflating point differential in counting terms - Different era dynamics (pace, three-point emphasis, defensive rules)
Lessons for Rate-Based Analysis
Lesson 1: Counting Stats Can Deceive
The Warriors averaged "only" 110 PPG despite having the best offense in modern history. Without ORtg, this historic efficiency would have been obscured by pace effects.
Lesson 2: Pace-Adjusted Comparisons Enable Cross-Era Analysis
Comparing the 73-win Warriors to the 72-win Bulls required pace adjustment. Raw scoring differentials were meaningless across different pace environments.
Lesson 3: On/Off Differentials Reveal Hidden Impact
Draymond Green's counting stats didn't suggest top-20 player impact. Rate-based on/off analysis revealed he was among the most impactful players in the league.
Lesson 4: Efficiency at Volume Matters
The Warriors maintained elite efficiency at high pace and high volume. Rate statistics (ORtg, eFG%, TS%) captured this better than raw counting statistics.
Lesson 5: Context Requires Multiple Metrics
No single metric captured the Warriors' excellence: - PPG missed the pace context - ORtg missed the defensive excellence - Net Rating captured both offense and defense efficiency - On/Off revealed individual contributions within the team framework
Analytical Framework Applied
Step 1: Raw Statistics
- Established baseline production: 110 PPG, 24 APG, etc.
Step 2: Pace Context
- Identified 99.3 pace vs. 95.8 league average
- Calculated possessions differential: +3.5 per game
Step 3: Rate Conversion
- Converted to ORtg: 114.5 (historically elite)
- Converted to DRtg: 103.8 (excellent)
- Calculated Net Rating: +10.7
Step 4: Per-Possession Player Analysis
- Evaluated Curry's 38.4 points per 100 possessions
- Identified Green's +16.1 on/off differential
Step 5: Historical Comparison
- Pace-adjusted to 1995-96 for Bulls comparison
- Found comparable dominance across eras
Step 6: Synthesis
- Combined efficiency, pace, and individual metrics
- Concluded historic offensive efficiency at sustainable pace
Conclusions
The Warriors dynasty demonstrated why rate statistics are essential for modern basketball analysis:
-
Pace-adjusted metrics revealed historic efficiency that raw scoring totals obscured.
-
Per-possession statistics enabled fair cross-era comparisons between fundamentally different playing environments.
-
On/off differentials identified impact that counting statistics missed entirely.
-
Three-point efficiency at volume could only be properly evaluated through expected value and eFG% frameworks.
-
Net rating combined offensive and defensive excellence into a single comprehensive measure.
The Warriors didn't just win games—they transformed how analysts understand what winning basketball looks like in statistical terms.
Discussion Questions
-
If the Warriors had played at league-average pace, how would their historical ranking change?
-
How should we weight pace-adjusted statistics vs. raw statistics when comparing teams across eras?
-
Did the Warriors' success cause other teams to increase pace, or did improved analytics reveal pace's value?
-
What do Draymond Green's on/off numbers suggest about the limitations of box score statistics?
-
How does the Warriors' three-point volume affect our interpretation of their efficiency metrics?
Data Sources
- NBA.com official statistics
- Basketball-Reference.com
- Cleaning the Glass
- NBA tracking data
- Historical pace data from Basketball-Reference