Behind the numbers.

Player in Focus compares NBA regular-season production across eleven saved seasons, 2015–16 through 2025–26. Data was retrieved from NBA.com on September 22–23, 2026. These are historical snapshots, not a live game feed.

Current-season updates

2026–27 production and game trends use the validated NBA completed-game cache. Skill cards use separately verified BBall Index captures when the provider offers the season. Missing ratings are never copied from a prior year. Each capture keeps its own date and season-specific comparison pools. Current NBA ranks and comparisons update from the same accepted final-game snapshot. Qualification is per statistic; a player’s actual production remains visible below a ranking threshold. No DNPs or missing games are filled with zeros. The existing qualification and low-sample rules still apply.

Activity: steals and blocks

The Activity index averages the steals and blocks per-game percentiles when both comparisons are available. These are recorded box-score results. The defensive skill profile is a separate comparison using BBall Index ratings.

What the page shows

Points, rebounds, assists, minutes, steals, blocks, fouls and turnovers are per game. Shooting percentages use season makes divided by attempts, rather than averaging individual game percentages. Estimated true shooting is points ÷ [2 × (field-goal attempts + 0.44 × free-throw attempts)]. It is labeled estimated because the free-throw factor is an approximation.

All figures cover the full season across every team the player appeared for. Team selection finds players; it does not restrict their stats to a team stint. Historical team membership comes from NBA player game logs. Biography age, height, weight, college and country use the NBA season bio response. Position and original NBA headshots describe the NBA profile; the photograph’s uniform may differ from the selected season. Missing photographs fall back to initials.

Who gets a rank?

• All categories require at least 500 season minutes.
• Field-goal percentage and estimated true shooting also require 200 field-goal attempts.
• Three-point percentage also requires 100 three-point attempts.
• Free-throw percentage also requires 100 free-throw attempts.

These are Mavs in Focus comparison rules, not official NBA leaderboard qualifications. Players below a threshold still have their raw numbers shown, but no rank, percentile or top/bottom badge. A dash is not zero. Fixed minimums apply even in shortened seasons, which changes the eligible population.

Ranks, percentiles and color

Competition ranks use unrounded values: ties share a rank and leave a gap in the next rank. For most categories, percentile = 100 × the count of eligible players with a strictly lower value ÷ (eligible players − 1). Fouls and turnovers reverse that comparison so fewer ranks higher. This does not adjust for role, playing time, usage or possessions.

Red marks the lowest third, yellow the middle third and green the highest third. A gold Top 10% badge requires an unrounded percentile of at least 90; a red Bot 10% tab requires a percentile below 10. Rounded dial numbers may therefore sit on a badge boundary. Ties can prevent any player from reaching 100. Percentiles and colors describe one statistic, not overall player quality. Minutes reflect workload; steals and blocks alone do not establish defensive quality.

Group indices

Volume score

The Volume index is the equal-weight average of the Points, Rebounds, Assists and Minutes per-game percentiles. All four comparisons must be available. This describes the size of a player's statistical role, not overall player quality.

Efficiency score

The Efficiency index equally averages estimated True Shooting, Field Goal, Three-Point and Free Throw percentage percentiles. Each statistic uses its own qualified comparison pool. All four comparisons must be available; any provisional component makes the index provisional.

Discipline score

Combines foul and turnover rates into a score out of 100. Fewer fouls and turnovers per game rank higher. Read it alongside playing time and ball-handling responsibility.

NBA defensive results

The detailed defense view retains six NBA result cards for every listed season, 2015–16 through 2025–26: overall defended shots, shots under six feet, isolation, pick-and-roll ball-handler results, deflections and contested shots. These support the BBall Index profile; they do not contribute additional weights to it. Team selection finds the player rather than restricting the data to a team stint.

NBA detail percentiles use100 × (number of worse values + half tied values) ÷ eligible players. Lower opponent-normal shooting differential and lower points allowed per possession rank higher. Deflections and contested shots are per game. Minimums:200 overall defended attempts,100 attempts under six feet,50 possessions per play type and500 hustle minutes. Samples below the provisional floors show a dash. These describe particular outcomes or activity, not complete defensive quality.

Sources: NBA defended shots, NBA defensive play types, NBA hustle, retrieved September24–25,2026.

Provisional comparisons: low sample, actual data

A dashed violet border, clock icon and “Provisional · Low sample” badge distinguish low-volume comparisons. Gauge colors still describe performance. Actual minutes, attempts or possessions remain visible. No totals are inflated and no modeled skill is extrapolated. The lower floors are editorial display safeguards, not statistically validated confidence thresholds.

• Traditional per-game stats and modeled defensive skills: at least 100 season minutes.
• Traditional shooting percentages: at least 100 minutes and 25 relevant attempts (FGA for FG/estimated TS, three-point attempts for 3P, free-throw attempts for FT).
• NBA shot-defense results: at least 25 defended attempts.
• NBA play-type results: at least 20 recorded possessions.
• Hustle activity: at least 100 source minutes.

Every comparison uses the original qualified pool; provisional players never enter that pool, receive a leaderboard rank, or earn Top/Bot 10% awards. Standard-card provisional comparisons use the existing below-value formula, clamped to 0–100. Defensive provisional comparisons use the same midrank formula as qualified defensive cards. At least 20 qualified comparison values are required. Qualification is per metric: a player can have a qualified play-type result and provisional skill cards.

A group index is provisional if any contributing comparison is provisional; all required inputs must still exist. Missing modeled core inputs are never imputed. A profile with too few comparison peers remains unscored. Historical provisional entries carry the same clock marker. The editorial Take continues to rely on qualified statistics.

Season trends and Mavs In Focus Take

Trends show the selected season first, followed by up to four earlier calendar seasons in the saved range. Missing seasons and insufficient samples remain dashes; a clock symbol marks provisional comparisons. Bars start at zero and scale within that stat’s visible five-season window. Changes use unrounded numbers and compare only the immediately preceding season; rounding can make displayed subtraction differ by a tenth.

Each season has its own qualifying comparison group. A percentile change is not proof of skill growth or decline. Mavs In Focus Take is a concise statistical interpretation of the selected player-season, with rule-based summaries for the wider directory and reviewed commentary for the Kyrie Irving 2024–25 example. It does not claim film review or reporting access.

Sources: NBA traditional stats, NBA player bios, NBA player index and regular-season game logs, retrieved using nba_api. Independent fan coverage; not affiliated with the NBA or Dallas Mavericks.

Back to Player in Focus

Offense profile: role-weighted skills

Our offense score combines the skill percentiles shown on the cards, giving more influence to the skills that matter most for the player’s recorded role. It brings together shooting, finishing, playmaking, shot creation, screening and offensive rebounding into one score out of 100.

Choose similar roles to compare the player with peers doing a similar job, or the league for a broader comparison. The overall score summarizes the skills; it is not an overall league percentile.

A qualified score requires 500 season minutes and enough eligible peers. Players with 100–499 minutes may receive a low-sample score. Missing required skills remain unscored.

Post scoring contributes when the player has enough recorded post-up usage and an available skill rating. Otherwise, the profile uses the remaining offensive areas.

Defense profile: role-weighted skills

Our defense score combines the skill percentiles shown on the cards, giving more influence to the skills that matter most for the player’s recorded role. It brings together one-on-one defense, navigating and defending screens, chasing shooters, help defense, rim protection and defensive rebounding into one score out of 100.

Choose similar roles to compare the player with peers doing a similar job, or the league for a broader comparison. The overall score summarizes the skills; it is not an overall league percentile.

A qualified score requires 500 season minutes and enough eligible peers. Players with 100–499 minutes may receive a low-sample score. Missing required skills remain unscored.

Assignment difficulty provides context about the opponents he guarded; it adds no separate bonus or penalty.

Player comparisons

Overall places the four category indices side by side; it does not create one combined player grade. Volume, Efficiency and Discipline average their component percentiles, requiring every component. Defense uses the profile blend described above. Comparisons default to the league defensive benchmark and can switch to each player’s own role. Different seasons use their own qualified peer groups; they are not a shared cross-season ranking. Violet low-sample markers remain attached to provisional indices.

Defensive cards: what each number means

Assignment difficulty

How demanding were the offensive players he was assigned to guard? A higher number means tougher assignments, not better defense. This is context rather than a separate bonus or penalty in our profile score.

Isolation skill

Can he make a scorer work one-on-one? This rating looks at reducing shot quality and suppressing attempts, with the offensive ability of the matchup taken into account.

Screen navigation

Getting through a screen is only part of the job. This rating looks at how well the defender limits shot quality and attempts around ball screens, accounting for the opponent’s skill.

Chasing shooters

Staying attached to shooters matters before they catch the ball. This rating looks at limiting attempts and shot quality around off-ball screens, with opponent skill taken into account.

Rim protection

Protecting the basket is more than blocking shots. This rating brings together deterrence, activity and disruption around the rim, including help situations.

Help defense

What happens when a teammate needs help? This estimate combines activity and effectiveness in help situations, adding another side to the direct-matchup picture.

Mobile-big defense

This is BBall Index’s estimate of mobile screener defense. Read it alongside rim protection and overall impact to see how that part of the player’s role fits the larger picture.

Screener rim defense

This is BBall Index’s screener rim-defense estimate. Pair it with rim protection to explore the player’s work around screening actions and the basket.

Defensive rebounding

Finishing a defensive possession matters. This is a modeled rebounding skill rating, rather than a count of rebounds per game.

These provider ratings are compared with qualified players in the selected season and comparison group. Higher skill percentiles indicate stronger estimates; higher assignment difficulty means tougher assignments. Each percentile counts worse values plus half of tied values. The provider models are not calculated by multiplying minutes or extrapolating small samples. Eligibility and low-sample rules are described above.

Overall Shot Defense

Compares opponent shooting allowed with those opponents’ usual shooting percentage. Subtract the opponent-normal percentage from the defended percentage; a lower difference rates better.

At the Rim · Under 6 FT

Uses the same opponent-normal shooting comparison for defended attempts under six feet. A lower difference rates better. This distance category is not the same as the restricted area.

Isolation Scoring Results

Points allowed divided by recorded isolation possessions. Lower points per possession rates better. These play-ending results answer a different question from the modeled Isolation skill rating.

Pick-and-Roll Scoring Results

Points allowed divided by recorded pick-and-roll ball-handler possessions. Lower points per possession rates better. A pass that leads to another action may be classified elsewhere.

Deflections

Total deflections divided by source games. A higher per-game rate places a player higher in this activity comparison. This measures disruption activity, not overall defensive quality.

Contested Shots

Total contested shots divided by source games. A higher rate indicates more contest activity. The shot-defense results help show what happened on defended attempts.

NBA result percentiles compare all positions within the selected season, counting worse values plus half of ties. Qualified minimums are 200 overall defended attempts, 100 attempts under six feet, 50 play-type possessions, or 500 source minutes for hustle activity. Provisional floors are 25 defended shots, 20 possessions, or 100 hustle minutes, with at least 20 qualified peers. NBA result cards provide context and are not additional inputs to the Defensive Profile.

Player Overview and Explore Offense/Defense use similar-role comparisons by default. When fewer than 20 role peers qualify, an available league comparison is shown automatically and labeled. The player’s actual role remains visible. Missing inputs and insufficient playing time remain unscored; no scores or sample limits are changed.

Screen coverage

BBall Index’s overall screener-defense skill estimate, compared with the selected eligible peers. It is distinct from a ball handler’s screen navigation.

Post scoring

The gauge compares BBall Index Post Up Shooting Talent against the selected peers. The card back shows NBA post-up results as supporting evidence; those result rates are not a decomposition of the provider’s model.