Stats Analysis: Offensive Efficiency (PGs)

The Bootleg's Basketball Stats Analyst Cameron Tana steps up with another of his somewhat complex, but fascinating columns, this time focusing on the historic individual "offensive efficiency" of Stanford point guards. Cameron has broadcast Stanford MBB games, both as Dave Flemming's color man in 1996-97 & as play-by-play announcer in 1997-98. He has also done NBA game charting for www.82games.com

Stats Analysis: Offensive  Efficiency (Individuals) - Point Guards

Previous columns covered team offensive and defensive efficiency ratings. This column applies the same concept of "possession-based" statistics to individual players. We will use these player offensive efficiency ratings to compare current Cardinal Mitch Johnson and his play to the careers of other Stanford point guards. 

Like team offensive efficiency ratings, individual offensive efficiency ratings are "points produced per 100 possessions". In his book Basketball on Paper, Dean Oliver, now with the Denver Nuggets front office, introduced formulas for estimating the number of points and possessions for which a specific player is responsible. These formulas were complicated enough that Oliver included them in an appendix to a book that already had plenty of numbers. This column introduces the basic parts of the calculations without getting into the detailed formulas. A team can score by field goals (including three-pointers) and free throws. There are two other offensive statistics: assists and offensive rebounds. Assists directly lead to field goals, while offensive rebounds extend possessions and sometimes lead to points. Points produced and scoring possessions have the following basic structure: 

Points Produced = (Field Goal Part + Free Throw Part + Assist Part) x (1-Team Offensive Rebound Part) + Offensive Rebound Part 

The basic principle behind these formulas is to "distribute credit" for a team's points amongst teammates. For example, credit for a basket's two points would be divvied up between the scorer, the assister, and the offensive rebounder. Therefore, the sum of individual points produced equals the team's total points. The media will commonly refer to dominant point guard being responsible for a total number of points by adding points to assists multiplied by a factor of 2.0 or so. If this was done for every player, the sum would be much greater than the actual points scored. In Oliver's ratings formulas, if credit is given for assists, credit must be taken away from the scorer. 

Since this column is focused on offensive point guard play, we will focus on the division of credit between field goals and assists. Credit for assists is proportional to teammates' effective field goal percentage. If a player has good-shooting teammates, passing is more valuable. Higher teammates' effective field goal percentage may also indicate that the assists are generally leading to easier shots. If a player has poor-shooting teammates, passing is less valuable and the team might be better off if the guard shot more himself. Conversely, credit is removed from the scorer for assists on his scores proportionally to his effective field goal percentage. If the scorer is a good shooter or getting the ball for easy scores, more credit should be given to his teammate for getting him the ball and in the right place. Here's what these parts look like: 

Assist Part ~ 0.5 x Teammates' Field Goal Percentage x Assists 

Field Goal Part ~ Field Goals x ( 1 - 0.5 x Player's Effective Field Goal Percentage x Percentage of Player's Field Goals That are Assisted) 

The other parts of the points produced formula are the free throw part and two offensive rebound parts. Free throw part is just free throws made, because passes that lead directly to free throws are not recorded as statistics like assists, although perhaps they should be. The offensive rebound parts subtract credit for the teams' offensive rebounds and add credit for the player's offensive rebounds that extend possessions. Our next column will discuss Brook Lopez's return, so we will go over the impact of offensive rebounds when we assess his and other frontcourt players' individual offensive efficiency ratings. 

Total offensive possessions for the player include scoring possessions and non-scoring possessions. Scoring possessions have the same parts as the points produced with the parts weighted differently to estimate the possessions where at least one point is scored. The non-scoring possessions include a missed field goal part, a missed free throw part, and a turnover part. The missed field goal part estimates missed field goals that are rebounded by the defense. The missed free throw part estimates the trips to the line where no free throws are made. Turnovers always end possessions so the turnover part is simply the number of turnovers. 

Total Possessions = Scoring Possessions + Missed Field Goal Part + Missed Free Throw Part + Turnover Part 

These total possessions lead to a key statistic in evaluating players, which is the percent possessions used while on the floor. Since the player shares possessions with his four teammates on the floor, 20% is average. Generally, there will be a decrease in a player's offensive efficiency rating as this percentage increases. The typical Stanford point guard has not been responsible for much more than 20% of possessions. This reflects both the ability level of Stanford point guards relative to his teammates and the design of the Stanford offense, which does not rely on the point guard to create off the dribble. 

"The Creators" (High-Possession Point Guards) 

The atypical Stanford point guard is embodied by Brevin Knight. Knight used over 27% of possessions (Pct POSS) in each of his last three years on the Farm. Despite the heavy burden he carried for the offense, he maintained a high offensive rating (OFF Rtg) in each of those years. His "assist rate" (AST Rate), the percentage of his teammates' field goals he assisted upon, was well over 40%. No Stanford point guard since has had an assist rate of above 33%. 

Brevin Knight's Career 

Year Min Pct POSS OFF Rtg EFG AST Rate TO Rate FT Rate
1994 916 22.7% 90.4 37.6% 30.5% 25.7% 30.8
1995 912 27.1% 109.7 48.4% 40.1% 20.7% 33.3
1996 915 28.8% 108.4 46.1% 48.5% 22.5% 32.3
1997 960 27.3% 114.8 47.4% 45.4% 21.3% 34.3

Knight could have had a higher offensive efficiency rating if he had used fewer possessions. However, this would not necessarily been optimal for the team. By using a higher number of possessions, Knight allowed his teammates to use fewer possessions and have a higher rating. For example, two-guard Kris Weems only had to use 16% of possessions during his year starting alongside Knight and posted an offensive rating of 118. In his junior and senior years, despite his additional experience, Weems's ratings fell to 117 and 110 as his possession usage increased to above 18%. 

The other teammate that benefited greatly from playing with Knight was Tim Young. With Knight, Young got enough touches to account for 23% and 24.5% of possessions, while posting ratings of 109 and 111. After Knight graduated, Young's possession percentage fell to 20% and 21%, but his rating improved only slightly to 112.5 and 112. 

Knight had a brilliant career, but even he was not particularly "efficient" during his freshman year. It was phenomenal for a freshman point guard to play so many minutes and control so many possessions for Stanford, but Knight would not reach a highly efficient level of play until after playing about 1,000 minutes. The other point guard that falls into a similar category as Knight in his high use of possession is Julius Barnes. Barnes was somewhat of an "accidental" point guard. He started at shooting guard his junior year alongside Tony Giovacchini at the point. He would have played the same position his senior year, except for the fact that Chris Hernandez got hurt. As starting point guard, Barnes used nearly 25% of the possessions while on the floor. Barnes was more of a shoot-first point guard, as his assist rate was the lowest of any starting point guard from 1994-2006. On the positive side, he managed to get up his shots without turning the ball over. His low turnover rate (TO Rate or "percentage of possessions turning the ball over") still allowed him to post a decent offensive rating despite being responsible for a high number of possessions. 

Julius Barnes' Career 

Year Min Pct POSS OFF Rtg EFG AST Rate TO Rate FT Rate
2000 207 20.3% 96.4 44.0% 29.0% 27.4% 19.2
2001 466 17.2% 96.7 46.0% 11.5% 21.5% 17.2
2002 806 19.8% 110.1 50.0% 20.3% 18.2% 16.8
2003 1158 24.5% 105.4 46.9% 22.4% 19.8% 24.7

Barnes's career is not very useful as a guide to Stanford point guard development as he played more shooting guard early in his career and was a high possession, shoot-first point guard. Anthony Goods does fit this model of a shooting guard who plays some point guard. Last year, Goods did not use nearly as many possessions as Barnes and had an extremely low assist rate. 

Anthony Goods' 2006-07 Season as a Part-Time Point Guard 

Year Min Pct POSS OFF Rtg EFG AST Rate TO Rate FT Rate
2007 841 20.6% 100.7 48.3% 13.8% 20.8% 22.9

"The Marksmen" (Average-Possession Point Guards) 

Most Stanford point guards do not account for a disproportionate number of possessions. Point guards who accounted for about 20% of possessions include Arthur Lee and Chris Hernandez. They did not have high assist rates, but were good shooters and got to the line a lot where there were excellent (FT Rate = "Made Free Throws per 100 Possessions"). Hernandez was the better field goal shooter (EFG = "Effective Field Goal Percentage"), but Lee took better care of the ball. Both also took over starting point guard duties from high possession players (Knight and Barnes respectively) and with the help of more experienced teammates, their teams improved their offensive efficiency over the previous year. 

Arthur Lee's Career 

Year Min Pct POSS OFF Rtg EFG AST Rate TO Rate FT Rate
1996 210 17.7% 94.9 53.0% 20.1% 30.3% 13.7
1997 625 18.7% 101.3 47.0% 15.4% 22.2% 24.1
1998 1105 21.3% 121.5 52.3% 26.6% 19.6% 38.8
1999 1076 20.8% 117.0 48.5% 27.5% 19.0% 37.4

Chris Hernandez' Career 

Year Min Pct POSS OFF Rtg EFG AST Rate TO Rate FT Rate
2002 395 14.3% 98.5 44.3% 18.3% 30.6% 31.6
2004 918 17.4% 117.8 58.0% 24.5% 26.1% 35.7
2005 1043 22.2% 110.6 53.3% 23.4% 20.8% 30.1
2006 993 21.6% 110.7 57.8% 22.8% 25.1% 31.1

Neither player saw much action as freshmen and posted sub-100 ratings like every other Stanford freshman point guard, although Hernandez's rating was the highest of any freshman point guard since 1994. As sophomores, they were able to ease into their roles. Lee ended up the season as the starting shooting guard alongside Knight and probably only played about 250 minutes at point guard. He then stepped in and led an experienced group with two years of very high offensive ratings. Hernandez had a very high rating during his redshirt sophomore year, but was not asked to use a high number of possessions. He lost Josh Childress and Matt Lottich as offensive threats and Mike Montgomery as his coach after that year. His offensive responsibility grew the following two years and his offensive efficiency ratings declined. 

"The Complements" (Low-Possession Point Guards) 

Some Stanford point guards use less-than-average percentage of possessions. This is a function of both the point guard's ability and the ability of his teammates. The two examples of these point guards, Mike McDonald and Tony Giovacchini, played with the most talented group of Stanford players in its history. This group included six NBA draft picks. As a result, these two would not need to use more than 16% of possessions in the three years they started at point guard. At that low possession usage, it should have been easier to have higher ratings, but that was not the case in two of the three years. 

Mike McDonald's Career 

Year Min Pct POSS OFF Rtg EFG AST Rate TO Rate FT Rate
1998 263 19.4% 87.0 34.6% 31.2% 29.4% 12.0
1999 212 19.4% 88.8 34.5% 29.4% 31.9% 29.0
2000 788 16.1% 94.0 45.6% 31.2% 35.4% 10.0
2001 917 15.8% 125.0 63.1% 30.2% 21.8% 8.3

Tony Giovacchini's Career

Year Min Pct POSS OFF Rtg EFG AST Rate TO Rate FT Rate
1999 175 18.3% 85.0 31.4% 23.9% 29.8% 29.8
2000 399 13.1% 109.5 46.1% 20.1% 21.0% 21.0
2001 418 15.4% 91.0 37.9% 24.2% 31.3% 21.5
2002 649 13.0% 90.4 41.3% 18.0% 31.8% 15.9

In their first years as starters, both McDonald and Giovacchini struggled with turnovers, which prevented them from having decent offensive efficiency. In his second year as a starter, McDonald improved tremendously. His turnover rate went way down and his shooting went way up. His individual improvement was a big reason for Stanford's seven-point offensive efficiency improvement from 2000 to 2001. How does this apply to the current Stanford point guard, Mitch Johnson? Here are his statistics from his first two years. 

Mitch Johnson's Freshman and Sophomore Years

Year Min Pct POSS OFF Rtg EFG AST Rate TO Rate FT Rate
2006 674 16.1% 78.3 37.6% 28.7% 37.9% 13.6
2007 730 15.5% 87.7 42.0% 25.6% 32.9% 12.2

First of all, except for Brevin Knight, no Stanford point guard discussed here played as many minutes as Johnson his freshman and sophomore years. Most point guards have the benefit of being an apprentice at this difficult position. Johnson did have the benefit of playing alongside Chris Hernandez during his freshman year, but he was not up to the task. Looking at other freshman seasons for Stanford point guards, it would have been pretty unreasonable to expect offensively efficient play from Johnson his first year out of high school. Johnson did show improvement his sophomore year, but his turnover rate was still too high. With a good assist rate, but a high turnover rate and low free throw rate, his first two years look a lot like Mike McDonald's first three years. McDonald's senior year improvement was exceptional and is unlikely to be repeated. However, Johnson has had enough experience that it is reasonable to expect him to grow into being an effective "complementary" point guard for the Stanford offense. Lot of people focus on his ability to knock down the outside shot, which no doubt is important, but I think turnover rate is probably the key statistic. If the point guard is a low-possession user, turnovers are especially harmful, because he probably was not going to produce points out of that play anyways. At least a missed shot gets the ball on the backboard, where Stanford's big men have been retrieving over 40% of the team's misses. In the early part of this year, Johnson has shown improvement in his offensive efficiency statistics. According to kenpom.com , this has come against opposition that ranks 301st out of 341 Division I teams in defensive efficiency. Therefore, the noted improvement might not hold up against higher quality Pac-10 competition. However, if Johnson had not shown much improvement against this schedule, that would be a serious red flag. 

Mitch Johnson This Season Year 

Year Min Pct POSS OFF Rtg EFG AST Rate TO Rate FT Rate
2008 239 15.9% 118.2 53.3% 26.7% 23.7% 20.5

The other point guard on Stanford this year is University of San Francisco transfer Drew Shiller. In his freshman year at USF, Shiller was a low-possession guard who played alongside a high- possession point guard in Armondo Surratt. Shiller's strength was shooting and he had a low assist rate. That has continued so far this year, with a scorching hot effective field goal percentage and a very low assist rate. Somebody with Shiller's ability to shoot should be able to contribute to the offense, but his low assist rate may make him less of a good fit with scorers such as Anthony Goods, Lawrence Hill, and Brook Lopez on the floor. It may make more sense to pair Shiller with non-shooting, play-making forward Fred Washington (23% assist rate last year). 

Drew Shiller's Season at USF and This Season

Year Min Pct POSS OFF Rtg EFG AST Rate TO Rate FT Rate
USF 2006 456 15.5% 101.5 51.9% 15.3% 25.2% 12.6
SU 2008 128 17.0% 135.2 80.4% 13.3% 24.8% 35.8

Are you fully subscribed to The Bootleg? If not, then you are missing out on all the top Cardinal coverage we provide daily on our website, as well as our full-length feature articles in our glossy magazine. Sign up today for the biggest and best in Stanford sports coverage with TheBootleg.com (sign-up) and The Bootleg Magazine (sign-up)!


The Bootleg Top Stories