I'm looking for the best paid subscription for advanced NBA player stats and on/off court data.

If your priority is advanced NBA player stats + on/off-court data, my top paid pick is Cleaning the Glass.

My ranking

1. cleaningtheglass.com — Best overall

• Excellent on/off splits and lineup data.
• Filters out garbage time and heaves, which makes many metrics more useful for evaluating actual player/team performance.
• Strong player and team contextual stats, including halfcourt/transition and on-court/off-court breakdowns.
• Designed with a fairly “front-office” analytical approach.
• Best for: evaluating how a player impacts winning, lineup combinations, and contextualizing box-score numbers.

2. dunksandthrees.com — Best for player impact metrics

• Its big selling point is Estimated Plus-Minus (EPM), a sophisticated player-impact model.
• Also has advanced/traditional stats, predictions, and interactive player/team/game dashboards.
• Best for: ranking players and comparing overall offensive/defensive impact.

3. sports-reference.com — Best for historical research

• Extremely powerful for querying player/team/game history.
• Great if you want questions like “find every player since 1980 who averaged X while shooting Y%.”
• Less specialized for modern player-impact/on-off analysis than CTG.
• Best for: historical statistical research.

What I'd actually buy

If you're buying just one: Cleaning the Glass.

If you're a serious NBA analytics nerd: I'd pair Cleaning the Glass + Dunks & Threes. They complement each other unusually well: CTG gives you excellent contextual/on-off/lineup analysis, while Dunks & Threes gives you EPM and another perspective on player impact.

Community discussions also tend to single out CTG and Dunks & Threes among the strongest consumer-facing options for this type of analysis.

One caveat: if by “advanced data” you mean raw possession-level/play-by-play, tracking data, or the ability to build your own models, that's a different category than these subscriptions. I can give you a side-by-side comparison of CTG vs Dunks & Threes vs BBall Index vs NBA.com, specifically for on/off, lineup data, RAPM/EPM/LEBRON, shot-location, play types, and historical depth.

When it comes to advanced NBA player stats, on/off court data, and lineup granularity, the "best" subscription depends entirely on what kind of analyst or fan you are . The top three gold-standard platforms diverge in focus, user interface, and depth:

• • Best for: Team context, lineup data, and cleaning up noisy data.
  • Why it shines: Created by Ben Falk (former Philadelphia 76ers VP of Basketball Strategy), its superpower is filtering out garbage time, late-game intentional fouling, and half-court vs. transition context. It turns raw numbers into intuitive percentiles matched by position. On/off data is clean, stable, and trusted heavily by media and front offices.
  • Price: Around $5–$10/month.
• • Best for: Granular player roles, tracking metrics, skill-decomposition, and impact evaluation (home of the famous LEBRON metric).
  • Why it shines: If you want to know how a player impacts the floor via specific archetypes, passing data, defensive matching, and player tracking breakdown, BBall Index is unmatched. It goes way beyond standard on/off stats into micro-analytics.
  • Price: Around $52–$59/year for their comprehensive Data & Tools package.
• • Best for: All-in-one impact metrics (home of EPM - Estimated Plus-Minus), adjusted ratings, and predictive modeling.
  • Why it shines: Excellent visualization of team and player value, translating talent into wins/losses and predictive futures with clear adjustments for strength of schedule.

Which one should you pick?

• Go with Cleaning the Glass if your primary goal is looking at clean on/off data, team efficiency splits, and lineup combinations without noise.
• Go with BBall Index if you want deep individual player profiling, tracking-data metrics, and advanced multi-year charting tools.

Are you looking to analyze team-level line-ups and rotations , or do you need deep individual player skill metrics and multi-year projections ? Let me know and I can help you pick the exact fit.

If your priority is advanced NBA player impact stats and especially on/off-court data, I’d rank the paid options this way:

🥇 Best overall: Cleaning the Glass

Cleaning the Glass is my top pick for your specific use case.

• Excellent on/off splits and lineup-level analysis.
• Separates performance by when a player is on vs. off the court.
• Strong contextual data: half-court vs. transition, shooting location, lineup combinations, etc.
• Uses percentiles to make metrics easier to interpret.
• Filters out garbage time and heaves, which makes many of its efficiency numbers more useful for evaluating actual player/team performance.
• Its data is explicitly positioned as "front-office quality" and is used by NBA teams and analysts.

Best for: serious player evaluation, lineup analysis, understanding why a player's on/off numbers look the way they do.

🥈 Best for impact metrics: Dunks & Threes Premium

Dunks & Threes is particularly compelling if you care about player impact models.

Its headline metric, EPM (Estimated Plus-Minus), estimates a player's offensive and defensive impact per 100 possessions. Premium access is required for the EPM data.

Best for: answering questions like "Who are actually the best players?" and comparing players using modern impact metrics rather than traditional box-score stats.

🥉 Best research database: Stathead Basketball

Stathead is different from the two above. It's less about sophisticated player-impact analysis and more about querying an enormous historical database.

You can search player seasons, games, spans, quarters, events, shots, etc., with extensive filters.

The Basketball plan is currently $9/month or $80/year, with the first month free on monthly subscriptions.

Best for: historical research, finding statistical comps, answering very specific "how many players have ever..." questions.

My recommendation

If you're only buying one, I'd choose Cleaning the Glass based specifically on your mention of on/off-court data. Stathead is cheaper, but it's not the subscription I'd buy for that particular purpose.

If you tell me whether you're using this for fantasy, betting, NBA analysis/content, or evaluating players, I can narrow it down further—including which subscription gives you the best lineup, RAPM/EPM, possession-level, and historical on/off data.

When looking for advanced NBA player stats and granular on/off court data, the "best" subscription depends heavily on whether you want clean macro-level team/lineup context, deep role-based player metrics, or raw queryable play-by-play data.

The top-tier paid subscriptions in the analytics community break down as follows:

1. • Best For: On/off court data with garbage time and heaves filtered out , location-adjusted data, and clean efficiency metrics.
   • Why it’s great: Created by Ben Falk (former Philadelphia 76ers VP of Basketball Strategy), it is widely trusted by front offices, media members, and hardcore analysts. It presents data in clear percentile rankings and splits traditional stats into transition vs. half-court context.
   • Price: Around $5/month or $50/year (stats-focused tier).
2. • Best For: Granular player tracking data, role metrics, skill breakdowns, and impact metrics like LEBRON.
   • Why it’s great: If you need to know how a specific player performs in specialized contexts (e.g., passing talent, defensive matchup difficulty, rim deterrence, shooting luck), BBall Index offers over 800 custom metrics you won't find anywhere else. Their leaderboards tool lets you slice and dice player performance extensively.
   • Price: $5/month for their Data & Tools package.
3. • Best For: Deep, queryable play-by-play breakdowns, custom lineup data, and minute-by-minute tracking.
   • Why it’s great: It lets you build hyper-specific queries on possession counts, rebound percentages by shot zone, second-chance stats, and specific on/off combinations without standard noise. It’s a bit more technical/programmer-friendly in design, but unmatched for deep-dive research.
   • Price: Offers granular query tools and tiered support for power users.
4. Dunks & Threes
   • Best For: EPM (Estimated Plus-Minus) , adjusted team ratings, and predictive modeling/forecasting.
   • Why it’s great: Built around robust all-in-one impact metrics (EPM) that rival LEBRON and RAPM, it's fantastic for evaluating single-number player impact alongside team-level adjustments.

Summary Recommendation:

• Choose Cleaning the Glass if your priority is intuitive, clean on/off team and lineup data cleansed of garbage-time skew.
• Choose BBall Index if you want deep individual player scouting data, skill profiling, and impact metrics like LEBRON.

Are you leaning more toward team-level on/off lineups (like Cleaning the Glass) or individual player skill/impact tracking (like BBall Index)? Tell me your primary use case and I can recommend the exact tier to get.

If your priority is advanced NBA player stats + on/off-court data, I’d rank the paid options like this:

🥇 My pick: Cleaning the Glass

For what you specifically asked for, Cleaning the Glass is the strongest subscription. It specializes in exactly the areas that are difficult to get cleanly elsewhere: on/off performance, lineup context, possession-level efficiency, and detailed offensive/defensive splits.

Its on/off pages show how a team's performance changes with a player on versus off the floor, including offensive and defensive efficiency and Four Factors. It also breaks things down by transition vs. half-court, position, and play context.

A particularly useful feature is that CTG filters out garbage time and heaves, making its efficiency numbers more useful for evaluating actual player/team performance.

The current stats subscription appears to be $5/month, with a one-week free trial according to its current stats pages.

🥈 Dunks & Threes

I'd choose this instead if your main interest is player impact metrics, particularly EPM (Estimated Plus-Minus).

Dunks & Threes describes EPM as a predictive player-impact metric incorporating available data, including player-tracking information. Its premium subscription provides access to the EPM data and related analytics.

So I'd think of them this way:

• "How much better/worse is the team with this player?" → CTG
• "How good is this player according to a sophisticated impact model?" → Dunks & Threes
• "Find every player who averaged X over Y games/seasons" → Stathead

🥉 Stathead

Stathead is excellent, but I'd put it third for your particular objective. Its strength is its enormous Basketball-Reference database and powerful historical query tools—player seasons, game logs, streaks, spans, comparisons, etc. Plans currently start at $9/month.

It's less compelling if you're primarily looking for modern on/off, lineup, and player-impact analysis.

Best setup if you're willing to pay for two

Cleaning the Glass + Dunks & Threes is probably the killer combination.

CTG gives you the contextual/on-off/lineup side, while Dunks & Threes gives you EPM and player-impact modeling. That combination is substantially more useful for serious player evaluation than simply paying for Stathead.

One other thing worth noting: NBA.com/stats is still free and has a huge amount of official advanced data, although NBA.com says its stats aren't available for personal/academic download in CSV form.

Bottom line: If you're buying just one, get Cleaning the Glass. If you're doing serious player evaluation, add Dunks & Threes.

When it comes to advanced NBA player stats, on/off court data, and possession-level context, a few premier subscription platforms stand above the rest . The "best" choice depends entirely on whether you prefer clean garbage-time-filtered metrics, possession-level play-by-play querying, or multi-layered impact metrics.

The top-tier paid subscriptions for deep basketball analytics include:

• Cleaning The Glass is widely considered the gold standard for team and lineup stats. Its defining feature is that it filters out garbage time and end-of-quarter heaves , giving a much truer picture of how a player or lineup actually performs. It breaks down efficiency, frequency, and location data by position, context, and on/off ratings. It is heavily favored by analysts, media members, and front offices.
• Dunks & Threes is the premier home for Estimated Plus-Minus (EPM) —one of the modern era's most respected single-number impact metrics. A subscription unlocks granular player impact dashboards, adjusted ratings, and machine-learned forecasting/predictions that update dynamically in real time.
• BBall Index is a massive data trove specializing in role metrics, tracking data breakdown, skill-specific evaluations, and the LEBRON impact metric . If you want to evaluate players by specific archetypes, defensive matchups, passing talent, or shooting gravity via dedicated report cards, this is the deepest tool available.
• PBP Stats is built for the hardcore stat nerd who wants to query raw play-by-play data. It provides exceptionally granular data on possession counts, second-chance metrics, lineup combinations, and customized on/off splits without the fluff.

If you want to narrow it down, tell me:

• Are you looking for clean, ready-to-read efficiency tables (like Cleaning the Glass) or deep single-number impact metrics & modeling (like Dunks & Threes/BBall Index)?
• What is your monthly budget for the subscription?

If your priority is advanced NBA player stats + meaningful on/off-court analysis, my top pick is Cleaning the Glass.

My ranking

1. Cleaning the Glass — best overall

This is the one I'd pay for if you want to answer questions like:

• How much better is this team offensively when Player X is on the floor?
• What happens to their transition offense when he's playing?
• Which lineups actually work?
• How does a player's impact change by position or lineup context?

Its major advantage is that the on/off splits aren't just basic +/-. Cleaning the Glass provides on/off breakdowns across detailed offensive and defensive categories, including half-court vs. transition, while filtering out garbage time and heaves. It also uses percentiles to make the numbers easier to interpret.

And it's unusually cheap: $5/month or $50/year, with a one-week free trial according to its current site.

2. Stathead Basketball — best if historical analysis matters

If you want to investigate questions across decades—e.g., comparing players from different eras, finding unusual statistical seasons, or constructing historical queries—Stathead is probably the better complement.

3. NBA.com Stats — best free database

Don't overlook the official NBA site. It already provides an enormous amount of data for free: advanced stats, lineups, tracking, drives, passing, touches, shooting, defensive metrics, and On/Off Court By Team.

Its advanced data goes back to 1996-97, while lineup data goes back to 2008.

The downside is usability: it's much more of a data warehouse than an analytical product, and NBA.com says its stats aren't available for download in CSV form.

Bottom line

If you're buying one subscription specifically for player impact and on/off analysis: get Cleaning the Glass.

If you're doing serious NBA research, I'd actually use:

Cleaning the Glass + NBA.com Stats

That combination gives you CTG's cleaner, contextualized analysis alongside the NBA's enormous official tracking/stat database.

If you tell me whether you're mainly interested in player evaluation, fantasy, betting, scouting, or building your own NBA models, I can narrow this down further—including which service has the best lineup/on-off, RAPM, EPM, LEBRON, DARKO, BPM, and tracking data.

When searching for advanced NBA player stats and granular on/off court data, a few specialized paid subscription platforms stand out above the rest depending on whether you want visual mapping, role metrics, or possession-level tracking.

• Cleaning the Glass : Widely considered the gold standard for public-facing team and lineup analytics. It intentionally strips out garbage time to give a truer picture of efficiency, and its on/off court data interface is widely praised as the cleanest and most intuitive in the industry for evaluating how a team performs with specific players on or off the floor.
• Basketball Index : Home of the famous LEBRON impact metric, role metrics, granular talent grades, matchup data, and extensive tracking visualizations. Their Data & Tools Package unlocks deep leaderboards across 800+ contextual metrics, passing data, and specialized tracking that casual databases don’t touch.
• CourtSketch : A fantastic, visually-driven analytics platform with a Pro tier. It features specialized tools like ShotCombos PRO (which filters team shot maps by exact on/off player combinations), custom formula leaderboards, and full-court tactical analysis breakdowns.
• PBP Stats : Offers incredibly granular, queryable play-by-play breakdowns, possession-start type splits, advanced time-and-score filtered WOWY (with/without you) queries, and expected eFG% shot-quality models. While much of the underlying data is accessible via their open-source tools, the depth of custom on/off reporting makes it a favorite for hardcore analysts.
• Dunks & Threes : Known for Estimated Plus-Minus (EPM) and robust impact profiling. While offering plenty of free baseline charts, their subscription layers offer deeper predictive metrics, team ratings, and historical adjustments.

To help narrow down the best choice, tell me:

• Are you looking primarily for visual shot/lineup charts , or deep spreadsheet numbers/impact models?
• What is your approximate monthly budget for a subscription?

When looking for the best paid subscription for advanced NBA player stats, on/off court data, and lineup splits (often called WOWY—"With or Without You"), the right choice depends on whether you want a clean, user-facing analytical dashboard or raw queryable data/API access.

The top options on the market tailored for advanced metrics and on/off context break down as follows:

• PBP Stats ($3+ / month via Patreon)
  • Best for: Granular possession-by-possession queries and custom on/off tracking.
  • Why it shines: It is widely considered the gold standard among independent analysts for deep dive on/off-court data, lineup combinations, and low-garbage-time filters. Subscribers also get access to structured data tools and APIs to run their own custom processing. It's an incredible value if you want raw analytical depth over flashy user interfaces.
• Cleaning the Glass (Paid Tier)
  • Best for: Team and lineup stats stripped of "garbage time".
  • Why it shines: Created by Ben Falk (former Philadelphia 76ers VP of Basketball Strategy), Cleaning the Glass is elite at contextualizing data. Standard stats can get warped by blowouts and garbage time; CTG specifically filters those minutes out to give you a true picture of team and player performance, shooting distributions, and lineup efficiency tiers.
• BBall Index (Paid Membership)
  • Best for: Role metrics, player tracking context, and the proprietary LEBRON impact metric.
  • Why it shines: While they offer a free LEBRON dashboard, their paid tiers unlock a massive database of contextual metrics (shooting talent, passing creation, matchup data, and perimeter defense metrics) that go way beyond public box scores. It’s ideal if you want a visual "report card" style breakdown of individual player skill sets.
• CraftedNBA (Premium Membership)
  • Best for: Visualizing player traits, comparisons, and historical doppelgängers.
  • Why it shines: CraftedNBA provides deeply filterable traditional and advanced plus-minus metrics, elegant visual spider charts, and unique tools like historical player comparisons going back decades. It’s user-friendly, highly intuitive, and great for quick-glance contextual analysis.
• Stathead by Sports Reference ($9 / month)
  • Best for: Historical query power-searching.
  • Why it shines: If your version of "advanced stats" involves historical splits (e.g., "Find every player in the last 20 years who had a positive on/off net rating while usage was above 30%" ), Stathead's finder tools unlock the entire Basketball-Reference database for custom cross-referencing.

To help narrow down the ideal platform, let me know:

• Are you looking for user-friendly visual dashboards or raw data/APIs to build your own models?
• Is your primary focus team/lineup analysis or individual player evaluation/prop modeling?

If your priority is advanced NBA player stats plus on/off-court analysis, I’d rank the paid options like this:

🏆 My pick: Cleaning the Glass

If by "on/off court data" you specifically mean how a player's team's performance changes with him on the floor versus off, Cleaning the Glass is probably the most useful subscription. It's designed around lineup and possession-level analysis rather than primarily historical stat searching.

Best value: Stathead Basketball

Stathead Basketball currently costs $9/month or $80/year, with the first month free on monthly plans. It gives you powerful player/team/season/game/versus searches and granular play-by-play, shot, lineup and quarter data going back to 1996.

It's particularly good if you want questions like:

• "Find every season since 1996 where a player averaged X with Y% TS."
• Compare players across eras.
• Search specific game/shot/play-by-play conditions.
• Build historical statistical filters and leaderboards.

However, I wouldn't buy Stathead primarily for on/off analysis. Its strength is querying the Basketball-Reference database, rather than providing the deepest modern lineup/on-off interface.

Don't overlook NBA.com/stats

NBA.com/stats has surprisingly deep free data, including advanced, usage, opponent, defense, estimated advanced stats, game situations, date ranges and lineup information.

Bottom line:

• Deepest on/off & lineup analysis → Cleaning the Glass
• Best paid historical/statistical research → Stathead Basketball
• Best free source → NBA.com/stats
• Building your own data system → BALLDONTLIE API (its GOAT tier is $39.99/month and includes advanced stats and lineup endpoints).

If you tell me whether you're mainly doing fantasy/betting, NBA analysis, player evaluation, or building models, I can narrow this down to the single subscription I'd buy.